Need helping making a term paper about Chinese American Culture

DGarcia08

Article1.pdf

Home >Nursing homework help >Need helping making a term paper about Chinese American Culture

Exploring Factors Influencing Chinese American Older Adults’ Intentions to

Plan for End-of-Life Care

Yanqin Liu

A Dissertation Presented in Partial Fulfillment

of the Requirements for the Degree

Doctor of Philosophy

Approved April 2018 by the

Graduate Supervisory Committee:

Anthony Roberto, Chair

Paul Mongeau

Marilyn Thompson

ARIZONA STATE UNIVERSITY

May 2018

ProQuest Number:

INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed,

a note will indicate the deletion.

ProQuest

Published by ProQuest LLC ( ). Copyright of the Dissertation is held by the Author.

Microform Edition © ProQuest LLC.

ProQuest LLC. 789 East Eisenhower Parkway

P.O. Box 1346 Ann Arbor, MI 48106 - 1346

10809417

2018

ABSTRACT

This study aimed to understand the factors that influence Chinese American older

adults’ advance care planning (ACP) on end-of-life care. The Theory of Planned

Behavior (TPB) and Health Belief Model (HBM) were primarily applied to explain

Chinese American older adults’ intentions toward two behaviors: 1) discussion of end-of-

life care plans with family members and 2) completion of an advance directive (AD).

Additionally, acculturation and family cohesion were considered to examine their

impacts on the TPB and HBM. A cross-sectional survey was conducted through face-to-

face interviews on a sample of 298 community-dwelling Chinese-American adults aged

55 and older living in the metropolitan Phoenix area of Arizona. Based upon random

assignment, 161 participants answered questions regarding discussing end-of-life care

plans with family members, while 137 participants answered questions related to the

completion of an AD. Hierarchical multiple regression analysis was used to focus on the

influence of TPB and HBM measures on behavioral intentions toward the two behaviors.

Results indicated that both the TPB and HBM had predictive power to explain the target

population’s intentions. However, the predictability of TPB and HBM measures varied

across the two behaviors. Acculturation moderated the relationship between attitudes and

intentions to complete an AD negatively. Family cohesion moderated the relationship

between perceived benefits and intentions to discuss end-of-life care plans with family

members negatively. These findings would help inform future interventions for

improving the target population’s ACP awareness and engagement.

ACKNOWLEDGMENTS

I gratefully acknowledge all who have given me support and encouragement for

my Ph.D. program.

I would like to thank my advisor, Dr. Tony Roberto, for providing tremendous

support for my academic path and inspiring me to develop my research interests to

accomplish my goals. Thank you for guiding me to be a better person, researcher, and

teacher. Also, I really appreciate Drs. Paul Mongeau and Marilyn Thompson’s help for

serving on my dissertation committee and contributing their time and insights to my

dissertation and multiple drafts. I feel so lucky to have studied in the Hugh Downs School

of Human Communication at the Arizona State University and have worked with many

great faculty members, staff, and colleagues in my life. I am truly grateful to learn from

them in my life.

Also, I would like to thank Drs. Fei Sun and Steve Corman. Thanks to them for

giving me the valuable opportunities to work on grants-funded research projects. Special

thanks go to Dr. Fei Sun, who helped me collect my dissertation data in Chinese

American senior communities in Phoenix, Arizona.

My journey would not have been possible without my caring families and friends.

I would like to thank my partner, Chase Carolan for listening to my concerns and offering

me a lot of emotional support. More importantly, my deepest appreciation goes to my

parents, Weidong Liu and Yuping Lu. I thank them for giving me unconditional love

since I was born and encouraging me to pursue my dreams in the U.S.

iii

TABLE OF CONTENTS

Page

LISTOF TABLES............................................................................................................ vii

LIST OF FIGURES........................................................................................................... ix

CHAPTER

1 INTRODUCTION .......................................................................................................1

Background ............................................................................................................. 1

Statement and Significance of the Problem ............................................................ 6

The Purpose of Study .............................................................................................. 9

2 LITERATURE REVIEW ..........................................................................................11

Advance Care Planning ........................................................................................ 11

Advance Care Planning in Chinese Communities ................................................ 14

Advance Care Planning Awareness .......................................................... 15

Factors Affecting Advance Care Planning ............................................... 16

Acculturation and Advance Care Planning ........................................................... 18

Family Influence and Advance Care Planning ..................................................... 19

Theory of Planned Behavior ................................................................................. 22

Theory of Planned Behavior in Health Topics ......................................... 28

Theory of Planned Behavior in Advance Care Planning .......................... 29

Health Belief Model .............................................................................................. 33

Health Belief Model in Health Topics ...................................................... 36

Health Belief Model in Advance Care Planning....................................... 38

Comparing Theory of Planned Behavior and Health Belief Model ..................... 41

CHAPTER Page

3 METHODOLOGY ....................................................................................................43

Overview ............................................................................................................... 43

Research Design ................................................................................................... 43

Participants ............................................................................................................ 44

Procedures ............................................................................................................. 45

Instrumentation ..................................................................................................... 48

Attitude ..................................................................................................... 48

Subjective Norms ...................................................................................... 49

Self-efficacy .............................................................................................. 49

Behavioral Intention.................................................................................. 50

Perceived Susceptibility ............................................................................ 50

Perceived Severity .................................................................................... 51

Perceived Benefits .................................................................................... 51

Perceived Barriers ..................................................................................... 52

Acculturation............................................................................................. 52

Family Cohesion ....................................................................................... 53

Demographic Variables ............................................................................ 53

Data Analysis Plan ................................................................................................ 54

4 RESULTS ..................................................................................................................56

Preliminary Statistics ............................................................................................ 56

Demographic Characteristics ................................................................................ 60

Theory of Planned Behavior Results .................................................................... 63

CHAPTER Page

RQ1a-b: Descriptive Statistics .................................................................. 63

RQ2: Comparing Theory of Planned Behavior Measures ........................ 63

H1a: Discussing End-of-Life Care Plans with Family Members ............. 64

H1b: Completing an Advance Directive ................................................... 70

Health Belief Model Results ................................................................................. 77

RQ5a-b: Descriptive Statistics .................................................................. 77

RQ6: Comparing Health Belief Model Measures ..................................... 77

H2a: Discussing End-of-Life Care Plans with Family Members ............. 77

H2b: Completing an Advance Directive ................................................... 83

Comparing Theory of Planned Behavior and Health Belief Model ..................... 87

5 DISCUSSION ............................................................................................................89

Overview ............................................................................................................... 89

Theory of Planned Behavior and Health Belief Model ........................................ 89

Theory of Planned Behavior ..................................................................... 89

Health Belief Model .................................................................................. 91

Comparing Theory of Planned Behavior and Health Belief Model ......... 92

Acculturation and Family Cohesion of Chinese American Older Adults ............ 93

Implications .......................................................................................................... 95

Advance Care Planning Interventions ...................................................... 95

Challenge of Using Timeline for the Recommended Behaviors .............. 98

Strengths ............................................................................................................... 99

Limitations .......................................................................................................... 100

CHAPTER Page

Conclusion .......................................................................................................... 101

REFERENCES ................................................................................................................103

APPENDIX

I IRB APPROVAL ............................................................................................. 103

II THEORY OF PLANNED BEHAVIOR AND HEALTH BELIEF MODEL

QUESTIONS ENGLISH AND MANDARIN VERSION 1 .......................... 119

III THEORY OF PLANNED BEHAVIOR AND HEALTH BELIEF MODEL

QUESTIONS ENGLISH AND MANDARIN VERSION 2 ......................... 127

IV SURVEY QUESTIONS ABOUT DEMOGRAPHICS, ACCULTURATION,

AND FAMILY COHESION IN ENGLISH AND MANDARIN ................. 135

vii

LIST OF TABLES

Table Page

1. Factor Loadings for Perceived Benefits and Barriers Toward Discussing

End-of-Life Care Plans with Family Members..................................................... 58

2. Factor Loadings for Perceived Benefits and Barriers Toward Completing

an Advance Directive............................................................................................ 59

3. Demographic Characteristics of Participants........................................................ 61

4. t-test Results of TPB and HBM Variables Comparison........................................ 64

5. Descriptive Information and Zero-order Correlation Coefficients Toward

Discussing End-of-Life Care Plans with Family Members (N = 161).................. 66

6. Results of Regression Analysis for TPB Toward Discussing End-of-Life

Care Plans (Acculturation as a Moderator) .......................................................... 67

7. Results of Regression Analysis for TPB Toward Discussing End-of-Life

Care Plans (Family Cohesion as Moderator) ....................................................... 68

8. Descriptive Information and Zero-order Correlation Coefficients Toward

Completing an Advance Directive (N = 137) ....................................................... 71

9. Results of Regression Analysis for TPB Toward Completing an Advance

Directive (Acculturation as a Moderator) ............................................................ 73

10. Results of Regression Analysis for TPB Toward Completing an Advance

Directive (Family Cohesion as a Moderator)......................................................... 74

11. Results of Regression Analysis for HBM Toward Discussing

End-of-Life Care Plans (Acculturation as a Moderator)....................................... 79

viii

Table Page

12. Results of Regression Analysis for HBM Toward Discussing

End-of-Life Care Plans (Family Cohesion as a Moderator) ................................. 80

13. Results of Regression Analysis for HBM Toward Completing an Advance

Directive (Acculturation as a Moderator) ............................................................. 85

14. Results of Regression Analysis for HBM Toward Completing an Advance

Directive (Family Cohesion as a Moderator) ....................................................... 86

LIST OF FIGURES

Figure Page

1. The Theory of Planned Behavior........................................................................... 22

2. The Health Belief Model....................................................................................... 34

3. Acculturation as a Moderator between Attitude and Behavioral

Intention toward Completing an Advance Directive............................................. 76

4. Family Cohesion as a Moderator between Perceived Benefits

and Behavioral Intention toward Discussing End-of-Life Care

Plans with Family Members………………………………………….…............. 83

CHAPTER 1: INTRODUCTION

Background

The growth of aging population has become an important demographic trend in

the U.S. (Ortman, Velkoff, & Hogan, 2014; Shrestha & Heisler, 2011). According to the

U.S. Census Bureau (2017c), residents aged 65 and older increased from 35 million in

2000 to 49.2 million in 2016, accounting for 12.4% and 15.2 % of the total population

respectively. The baby-boom generation is mainly responsible for this demographic

trend. Baby boomers began turning 65 in 2011 and would continue to do so for many

years to come. The projected population of people aged 65 and older will reach to 98.2

million in 2060 (U.S. Census Bureau, 2017b).

The increase of aging groups is accompanied by challenges to the U.S. health care

system due to this population’s complex health conditions. Given that aging groups

utilize health care system more frequently than other age categories (Nussbaum & Fisher,

2009), it is crucial for elders to plan for health care decisions in advance to ensure that

they will receive medical care that can reflect their values, wishes, and preferences.

Advance care planning (ACP) serves as an important component in end-of-life

care. It is conceptualized as a decision-making process regarding considering what care

people would like to receive in the future if they become unable to speak for themselves

due to a life-threatening event (National Hospice and Palliative Care Organization, 2016).

ACP involves ongoing processes including discussions about goals of care, resuscitation

and life support, palliative care options, surrogate decision making, and advance

directives (Houben, Spruit, Groenen, Wouters, & Janssen, 2014). The main goal of ACP

on end-of-life care is to let others know about a person’s medical treatment preferences in

advance and selecting a health care proxy when a person does not have the capacity to

make decisions on his or her own. A growing body of research suggests that ACP is

helpful for doctors and family members to know about patients’ medical treatment

preferences when patients are not able to speak for themselves (National Institute of

Aging, 2016; Sudore & Fried, 2010).

The previous literature has demonstrated the benefits of successful ACP

(Brinkman-Stoppelenburg, Rietjens, & van der Heide, 2014; Houben et al., 2014;

Kononovas & McGee, 2017). ACP has been associated with the improvement of quality

of life for patients and their families (Bischoff, Sudore, Miao, Boscardin, & Smith, 2013;

Heyland et al., 2013; Wright et al., 2008). Wright et al. (2008) found that end-of-life care

discussions were related to lower rates of ventilation, resuscitation, ICU admission, and

earlier hospice enrollment. More aggressive medical care was associated with worse

quality of life in patients and higher risk of major depressive disorder in bereaved

caregivers, while longer hospice stays were associated with better quality of life in

patients. Better patients’ quality of life was associated with better bereaved caregivers’

quality of life. Also, ACP was found to have positive impacts on lowering health care

costs in patients’ last week of life (Zhang et al., 2009), increasing patients’ satisfaction

with overall care in the hospital and reducing surviving relatives’ stress, anxiety, and

depression (Detering, Hancock, Reade, & Silvester, 2010).

It is recommended that people discuss with their family members and health

providers and then document their ACP preferences through advance directive

(Enguidanos & Ailshire, 2017). Advance directive (AD) refers to a legal health care

document that provides written directions about people’s medical treatment preferences

related to end-of-life care and goes into effect only when people are unable to speak for

themselves (Durbin, Fish, Bachman, & Smith, 2010; National Institute of Aging, 2016).

ADs become legally effective in the U.S. once people sign them in front of the required

witnesses (National Institute of Aging, 2016). It is implied that ADs remain in effect until

people would like to complete a new AD and invalidate their previous ones.

An AD includes two primary elements (i.e., a living will and durable power of

attorney for health care) and other documents like a do not resuscitate (DNR) order and

the Five Wishes. A living will is a written document that states how people want to be

treated when they become unable to speak for themselves. This document can guide

health providers to withhold specific life-sustaining treatments such as cardiopulmonary

resuscitation (CPR), mechanical ventilation for breathing, and artificial feeding. Durable

power of attorney for health care is a legal document appointing a health care proxy to

make medical decisions on behalf of people when they become unable to make decisions

on their own. A DNR order provide directions for physicians whether or not to perform

cardiopulmonary resuscitation. Created by a non-profit organization, Aging with Dignity,

the Five Wishes is an advance directive document includes:

wishes for the person I want to make care decisions for me when I can’t, the kind

of medical treatment I want or don’t want, how comfortable I want to be, how I

want people to treat me, and what I want my loved ones to know. (Aging with

Dignity, 2011)

These documents provide guidance for physicians and family members when a patient

does not have decisional capacity to communicate their decisions.

The AD development began in the U.S. in the late 1960s for end-of-life care

planning (Wilkinson, Wenger, & Shugarman, 2007). One of the primary milestone events

in the history of ADs was the enactment of the Patient Self-Determination Act (PSDA) in

1990 (Sabatini, 2010). To inform patients of their rights regarding decisions toward their

medical care, the PSDA was designed to ensure that patients are provided information

about ADs and can accept or refuse medical treatments (Brown, 2003; Sabatino, 2010).

The PSDA requires health care providers in hospitals, skilled nursing facilities, and other

health care settings to take the following steps, including informing patients’ rights to

participate in and directing their own medical care decisions, developing written policies

regarding ADs, asking new patients whether they have had an AD and having this

information in the patients’ records, providing patients written information regarding the

facility’s policies on ADs and patients’ rights to prepare these documents, and educating

staff and communities about ADs (Greco, Schulman, Lavizzo-Mourey, & Hansen-

Flaschen, 1991; Miller, 2017).

Although the PSDA protects patients’ rights to make their end-of-life care

decisions, this legislation relies heavily on inpatient facilities and does not encourage

people to prepare for their ADs before the need for hospitalization or long-term care

arises (Greco et al., 1991). Also, it does not specify that doctors must discuss ADs with

patients and their family members. Furthermore, a wide variety of medical and legal

literature have criticized the PSDA’s failure to meet the needs of patients with limited

English proficiency (Pope, 2013). These limitations have driven policy makers and health

professionals to explore the ways how ACP should be promoted. Since January 1, 2016,

Medicare has begun to pay health care providers for face-to-face conversations with

Medicare patients and/or their surrogates regarding their ACP (Centers for Medicare &

Medicaid Services, 2016). This updated policy acknowledges the value of ACP

discussions to improve the quality of delivering end-of-life care.

Although the previous literature has considered the AD completion as a primary

objective of ACP (Sudore & Fried, 2010), having ACP conversations regarding end-of-

life care between patients, family members, and doctors has become increasingly

important (Fried, Bullock, Iannone, & O’Leary, 2009). Both the AD completion and

ongoing discussions are viewed as important components for the design of effective ACP

interventions (Houben et al., 2014). ACP should be considered as a multifaceted

decision-making process that involves discussions in which patients, family members,

and doctors explore care goals under current and hypothetical illness conditions, discuss

treatment options in the context of these care goals, and finally articulate and document

treatment and care preferences (Morrison & Meier, 2004).

Medicare has begun reimbursing health providers for ACP discussions since

January 1, 2016. However, it does not necessarily indicate that health providers will talk

to their patients. A national survey of physicians who regularly treat patients aged 65 and

older found that 95% participants supported this new Medicare benefit that reimburses

health providers for ACP discussions, but only 14% participants who have fee-for-service

patients had actually billed Medicare for this conversation (PerryUndem

Research/Communication, 2016). Volandes (2015) suggested that patients start the

conversation on their own with family members instead of waiting for their doctors to

start the conversation. He explained that physicians’ medical trainings focus on medical

technology rather than communication skills and they do not have sufficient structural

supports to be encouraged to start the ACP process. At the same time, when patients are

not able to communicate their medical treatment preferences due to illnesses, doctors

typically seek guidance from patients’ family members. It is possible that family

members make choices that a patient would have disagreed, without knowing a patient’s

thoughts in advance.

Statement and Significance of the Problem

Asian Americans grew faster than any other ethnic population over the last two

decades in the U.S. (U.S. Census Bureau, 2012; U.S. Census Bureau, 2017c). The

estimated number of Asian alone or in combination residents in the U.S. in 2015 was 21

million (U.S. Census Bureau, 2017a). By 2060 this population is projected to grow to

10% of the total U.S. population (Colby & Ortman, 2015). As the largest groups of Asian

Americans, there are approximately five million Chinese Americans in the U.S. (U.S.

Census Bureau, 2016). According to Pew Research Center (2017), people aged 50–64

and 65 and older accounted for 19% and 12% of the Chinese population in the U.S.

Given the fast growth of Chinese American aging population in the U.S., it would

increasingly become common for health providers to work with their Chinese American

patients toward their ACP decisions on end-of-life care. However, it is rare to find the

literature specifically examining Chinese Americans aging population’s use of ACP on

their end-of-life care.

The previous literature indicated the 26.3% (n = 2093) AD completion rate among

U.S. adults in a national survey, and non-Hispanic Whites (30.7%; n = 1605) accounted

for most of those who completed one AD than African Americans (17%; n = 169) and

Latinos (16.7%; n = 175) in this survey (Rao, Anderson, Lin, & Laux, 2014). Similarly,

Huang, Neuhaus, & Chiong (2016) showed that non-Hispanic White participants were

significantly more likely to possess ADs (44%; n = 758) than African older Americans

(24%; n = 48) and Hispanic older Americans (29%; n = 30). Ethnic minority populations

were found to be less likely to complete an AD than their White counterparts.

Chinese Americans are not active in planning for end-of-life care. Gao, Sun, Ko,

Kwak, and Shen (2015) surveyed 385 Chinese Americans aged 55 and older and found

that 80 participants had heard about AD and only 38 of them had completed one AD. In

another study focusing on Chinese Americans, Hsiung (2011) found that approximately

67% of 206 Chinese Americans aged 45 years and older were not aware of ACP

importance. These findings indicated low ACP awareness and engagement.

The barriers that influence the ACP promotion among Chinese Americans are

multidimensional. For health providers, they may not be willing to initiate ACP processes

(Blackford & Street, 2016). De Vleminck et al. (2013) identified the potential barriers to

health providers’ unwillingness in a systematic review, including a lack of knowledge

and communication skills, difficulties in defining the right moment for initiating

conversations and advising patients to express their wishes, concern about causing

patients and family members’ emotional reactions when raising this topic, doubt about

pragmatic availability of ADs, and thinking patients should initiate discussions.

Meanwhile, De Vleminck et al. examined the barriers to patients’ ACP involvement. For

example, patients are reluctant to think about future health care problems. They may lack

knowledge about ACP processes and be afraid of upsetting their family members. Also,

an AD document can be challenging for them to understand and complete due to its

complexity and length.

For Chinese Americans, ACP is an unfamiliar concept in public discourse (Lee,

Hinderer, & Kehl, 2014). A large amount of them may lack sufficient ACP knowledge or

language proficiency to understand the importance of ACP. Furthermore, under the

influence of traditional Chinese culture, people rarely plan for their future illness

conditions and talk about death and dying (Lee, Cheng, Dai, Chang, & Hu, 2016). It is

common that discussions about a Chinese patient’s medical treatment is postponed until

the occurrence of a medical crisis. Without informing others in advance, patients may

receive medical treatments that do not reflect their treatment preferences during their end-

of-life periods. Also, crisis-oriented decision-making processes may cause emotional

distress in loved ones.

A low ACP awareness may not be directly associated with having aggressive care

treatments. However, when patients do not properly articulate their medical treatment

preferences, they are more likely to be overtreated than undertreated, contributing to the

high costs of medical care in their last months of life (Boerner, Carr, & Moorman, 2013).

The previous literature has shown that patients with ADs completed in the last months of

life had higher rates of election of aggressive care, compared to those who completed

earlier (Enguidanos & Ailshire, 2017). It may also cause health care proxies emotional

distress when important health decisions are made without knowing about patients’

medical treatment preferences (Detering et al., 2010). More studies are needed to focus

on Chinese Americans’ ACP behaviors to inform future educational interventions to

encourage them to take responsibilities for their health and improve their ACP awareness

and engagement.

In terms of involving in the ACP processes, the previous studies showed that

Chinese older adults feel more comfortable to discuss their end-of-life care with family

members. In a study by Zhang et al. (2015), Chinese elders living in Beijing reported

being comfortable with initiating the topics of end-of-life care with their family members

(70.7%) than physicians (62.9%). In another study, Gao et al. (2015) found that Chinese

American elders tended to discuss their preferences for the use of life-sustaining

treatments with their family members (23%) more than their physicians (6%). These

studies implied the importance of having ACP discussions with family members, when

Chinese older adults can communicate their wishes and values. These findings would

guide this dissertation to partially focus on participants’ behavior of discussing end-of-

life care with family members.

The Purpose of Study

This dissertation will focus on both the AD completion and ongoing

conversations and understand two relevant behaviors (i.e., discussing end-of-life care

plans with family members and completing an AD) to provide a comprehensive analysis

of the importance of ACP on end-of-life care. Although both are related to ACP on end-

of-life care and share similar characteristics, people may have different understandings of

these two behaviors. This comparison would drive us to think more about how we can

identify behavioral recommendations for the future ACP interventions among Chinese

American elderly. Given Chinese American aging groups’ low engagement in planning

for end-of-life care, it is challenging to measure actual behaviors that participants may

not perform until later in their lives. Therefore, the primary outcomes in this dissertation

are behavioral intentions instead.

The purpose of this study is to understand the factors that influence community-

dwelling (i.e., living in the community independently) Chinese American older adults’

intentions to plan for end-of-life care through the Theory of Planned Behavior (TPB) and

the Health Belief Model (HBM) in the metropolitan Phoenix areas. To achieve the

research goal, this study includes in-depth interviews as formative research and cross-

sectional survey as primary data collection technique. In formative research, participants

were asked to answer open-ended questions during the in-depth interviews and their

responses were used to develop the items in a cross-sectional survey. Later, participants

were asked to complete a survey consisting of questions regarding the TPB and HBM, as

well as their demographic information. It is noted that both theories apply an individual-

level approach to predict health behaviors. However, factors from interpersonal,

community, and societal levels may influence the predictability of the two theories.

Considering the collective and multicultural characteristics of Chinese populations in the

U.S., my dissertation examines the moderating impacts of acculturation and family

cohesion on the TPB and HBM.

CHAPTER 2: LITERATURE REVIEW

Advance Care Planning

The previous systematic reviews have evaluated the advance care planning (ACP)

effectiveness in different health conditions. Improving the advance directive (AD)

completion rate has been considered as the main goal for ACP-related research

(Brinkman-Stoppelenburg et al., 2014; Durbin et al., 2010; Houben et al., 2014). Given

that ACP has been considered as a communication process, the previous systematic

reviews have also evaluated the role of communication on ACP processes and

communicating about end-of-life care has been considered as an important component

(Durbin et al., 2010; Houben et al., 2014; Sharp, Moran, Kuhn, & Barclay, 2013).

Durbin et al. (2010) searched 2,000 potential studies published in CINAHL,

EBSCO, Medline, and Science Direct between 1991 and 2009 and selected 12

randomized and four nonrandomized studies published from the nursing, medical, and

social work literature. The analysis included two inpatient hospital-based studies, nine

outpatient hospital-based studies, and one community-based study. The authors

systematically analyzed evidence about one outcome, the percentage of newly completed

ADs (i.e., number of completed ADs postintervention minus number of completed ADs

at baseline divided by number of participants per group).

The results showed that most studies examined combined written and verbal

educational interventions (i.e., giving written materials to subjects with verbal

reinforcement of the material either simultaneously or over specific time periods). They

did not find sufficient evidence to support the effectiveness of a single written or a single

verbal educational intervention in significantly increasing the percentage of newly

completed ADs. However, three randomized studies consistently showed that combined

written and verbal educational interventions were significantly more effective than single

written interventions in increasing the percentage of newly completed ADs.

Later, Brinkman-Stoppelenburg et al. (2014) systematically searched PubMed,

EMBASE, and PsycINFO databases for experimental and observational studies on the

effects of ACP published January 2000 until December 2012 and hand searched the

Journal of the American Geriatric Society and the Journal of Palliative Medicine from

2009 to December 2012. They incorporated 113 papers in the review, including 95%

observational design and 5% experimental design. There were 48% studies conducted in

hospital, 32% in nursing home, 11% in a mixed setting, 8% in community, and 1% in

outpatient clinic. This review encompassed 52 studies on do-not-resuscitate (DNR)

orders, 45 studies on the completion of ADs (i.e., living wills and durable powers of

attorney), 16 studies on do-not-hospitalize (DNH) orders and 20 studies on complex ACP

interventions (communication components included).

In the review of Brinkman-Stoppelenburg et al. (2014), most studies on the effects

of DNR orders showed that these were associated with a decreased use of CPR (four of

five studies) and an increased use of hospice and/or palliative care (six out of six studies).

Most studies showed a decreased use of life-sustaining treatments (12 of 21 studies).

Studies on DNH orders (16 studies) showed a decrease in hospitalization (eight of nine

studies), a decrease in life-sustaining treatments (three of three studies) and an increase in

hospice and/or palliative care (five of five studies). Among 45 studies on ADs, life-

sustaining treatment use was the outcome in 22 studies and 10 of them reported that ADs

were associated with a decrease in the use of life-sustaining treatments. In five of seven

studies, patients with the completion of ADs were more frequently enrolled in hospice

care and/or palliative care service than those without ADs. Four of six studies found that

ADs were associated with an increase in the use of comfort plans.

Furthermore, a total of 20 studies on the effects of complex ACP interventions

showed that three of four studies reported increased compliance with patients’ end-of-life

wishes. Three of five studies reported a decrease in the use of life-sustaining approach

and four of eight studies reported an increase in participants’ satisfaction or quality of

life. Brinkman-Stoppelenburg et al. (2014) suggested that extensive ACP interventions

may be more effective to result in an increased frequency of out-of-hospital and out-of-

ICU care and in increased compliance with patients’ care satisfaction than written

documents alone.

Houben et al. (2014) conducted a literature search including Medline/PubMed and

Cochrane Central Register of Controlled Trials from 1966 through September 2013. They

systematically reviewed 26 trials focusing on ADs and 30 trials focusing on both ADs

and communication to identify the efficacy of ACP in different adult populations. These

studies were published between 1992 and 2012, including 15 studies in an inpatient

setting, 37 studies in an outpatient setting, and four studies in both settings.

It is found that patients in the intervention groups completed an AD more often in

comparison with control groups (odds radio = 3.26; 95% CI = 2.00–5.32; p < .001). The

results also showed a more likelihood for the occurrence of discussions about end-of-life

preferences between patients and health professionals in the intervention groups than

control groups (odds ratio = 2.82; 95% CI = 2.09–3.79; p < .001). In other words,

interventions focusing on ADs, as well as interventions that included both ADs and end-

of-life care communication, increased the AD completion and the occurrence of end-of-

life care discussions between patients and health care professionals.

Additionally, patients in the intervention groups had an increased likelihood of

delivered end-of-life care consistent with their end-of-life care preferences compared

with control groups (odds ratio = 4.66; 95% CI=1.20–18.08; p = .03). However, this

study did not find the supporting evidence showing whether the interventions that

included both ADs and communication worked more effectively that the interventions

that included only ADs.

These reviews have showed that both end-of-life care communication and AD

completion are important components for ACP promotion. This implication would guide

this dissertation to include both end-of-life care communication and AD completion and

examine the factors that influence these two recommended behaviors. In addition, most

reviewed studies were conducted in clinical or nursing home settings. More community-

based studies would be needed to help more senior community members understand the

importance of ACP awareness and engagement. This dissertation will address this

research gap by focusing on community-dwelling participants.

Advance Care Planning in Chinese Communities

Researchers have become increasingly interested in ACP among Chinese

communities during the past decades (Lee et al., 2014). The topics include ACP

awareness (Gao et al., 2015; Yap, Chen, Detering, & Fraser, 2017), end-of-life care

preferences (Ni et al., 2014; Zhang et al., 2015), factors affecting AD and ACP (Gao et

al., 2015; Tang, Lam, and Chiu, 2007; Zhang et al., 2015), palliative and hospice care

(Enguidanos, Yonashiro-Cho, & Cote, 2013; Kang et al., 2012), Chinese cultural

consideration of ACP (Chan & Yau, 2009; Lee, Cheng, Dai, Chang, & Hu, 2016), and

ACP interventions (Cheng, Lo, Chan, & Woo, 2010; Ho et al., 2016). ACP awareness

and factors affecting AD and ACP preferences were frequently examined.

Advance Care Planning Awareness

Chinese respondents commonly indicate low ACP awareness and knowledge.

Most people do not know about the definitions of ACP and AD or have not heard of AD

in both Eastern (Chu et al., 2011; Low, Ng, Yap, & Chan, 2000; Ni et al., 2014; Ting &

Mok, 2011; Zhang et al., 2015) and Western countries (Gao et al., 2015; Yap et al.,

2017). Gao et al. (2015) surveyed 385 Chinese Americans aged 55 and older living in the

metropolitan Phoenix areas and found that 79% had not heard of AD before. Only 10% of

them had completed one. Among those who had heard of AD, they tended to know more

about the role of an AD in medical treatment decision (95.1% accuracy rate) and less

about its role in financial affairs (59.3% accuracy rate).

Hsiung (2011) applied the Transtheoretical Model to study Chinese American

adults’ readiness for advance care planning. She identified six stages of change for the

target population, including precontemplation (non-believers; show no interest),

precontemplation (believers; unaware of ACP and not take changing seriously),

contemplation (seriously consider changing within the next six months), preparation

(seriously consider changing within the next month), action (have given oral directives or

made legal ADs within the six months), and maintenance (have completed an AD more

than six months and communicate with others continuously).

The results showed that among the 206 participants, 68% of the participants were

classified as believers at the stage of precontemplation and intended to initiate advance

care planning, followed by 12.13% at the action stage, 9.7% at the preparation stage, 5%

at the maintenance stage, 2.9% showing no interest in and willingness in ACP, and 1.94%

at the contemplation stage. People at the precontemplation-believer stage thought that it

was necessary and enthusiastic to receive additional information about ACP but they

were unclear about the most appropriate time to do it. Compared with precontemplators

and contemplators, participants at the action and maintenance staged seemed to be

relatively less traditional in cultural beliefs, better in English, and more knowledgeable

about ACP. Different from actioners showing no interest in updating their ADs,

maintainers have either given a copy of their completed ADs to family members,

renewed their ADs at least once, or promoted the AD completion with relatives and

friends. Overall, this study showed that most participants had low ACP awareness and

were still in the precontemplation stage.

Factors Affecting Advance Care Planning

To improve low ACP awareness, the researchers focused on the factors

influencing Chinese individuals’ AD and ACP preferences in different regions and

countries (Chu et al., 2011; Yap et al., 2017; Zhang et al., 2015). Zhang et al. (2015)

conducted a study among elders in Beijing and suggested education levels and age served

as predictors for ACP preferences. In other words, participants with higher education

levels were more likely to have heard of ACP and would prefer to document their ACP

decisions than those with lower education. Those aged less than 70 years were more

likely to have heard of ACP and refuse life-sustaining treatments than those aged 70

years and older.

In contrast, Chu et al. (2011) examined factors that may influence participants’

preferences for ADs and for community-based end-of-life care in Hong Kong

respectively, adjusting for the influence of age and sex. Significant predictors of the AD

preference included asking for relatives’ advice in medical decisions, wishing to be

informed of their terminal diagnoses, absence of stroke, and having no problems in self-

care in European Quality of Life-5 Dimensions. For the preferences for community-based

end-of-life care, the independent predictors included older age, not having siblings in

Hong Kong, Catholic religion, nonbeliever of traditional Chinese religion, not receiving

any old age allowance, lower Geriatric Depression Scale score, and being residents of

government-subsidized nursing homes.

For Chinese elders living in Western countries, having high-level language

proficiency becomes important. Yap et al. (2017) found that in-language materials, key

support networks (i.e., general practitioners, families, and Chinese community groups)

were useful tools for ACP promotion because older Chinese-speaking community

members have language barriers and rely on families, general practitioners, community

friends and volunteers, and medical interpreters to access health care.

In addition to language proficiency, it is necessary to consider the influence of

cultural adjustment among Chinese elders living in Western countries. Gao et al. (2015)

conducted a binary logistic regression to examine the impact of acculturation on AD

awareness. Gender, age, education, monthly income, self-rated health, and previous

experiences of end-of-life care were entered into the first block. Those with higher

education levels, higher monthly incomes, and experience with ventilators were more

likely to know about ADs. In the second model, the acculturation levels and years of U.S.

residency were entered. Education levels and experiences with ventilators remained

significant. Participants with higher acculturation levels and those residing more than 20

years in the U.S. were more likely to have AD awareness. It is also found that concerns

about causing family burdens was the most important factor that influenced Chinese

American elderly’s preference for end-of-life care, followed by pain relief, best interests

in the eyes of family members, the possibility of being cured, and financial cost.

It is implied that many ACP-related studies focusing on Chinese populations are

not theory-driven and factors that predict the target population’ AD and ACP preferences

vary across different settings. Lacking theory as the ground can limit the generalizability

of these studies in Chinese communities. This limitation would guide this dissertation to

be theory-based to better inform future ACP interventions for the target population.

Acculturation and Advance Care Planning

Acculturation is conceptualized as “the dual process of cultural and psychological

change that takes place as a result of contact between two or more cultural groups and

their individual members” (Berry, 2005, p. 698). It is commonly measured by migrants’

language use, media preferences, social affiliations, cultural customs/manners, belief

systems associated with a specific context or group, attachments to cultural groups, and

the positive esteem drawn from these attachments (Schwartz, Unger, Zamboanga, &

Szapocznik, 2010). A higher level of English proficiency is frequently associated with a

higher level of acculturation (Hsiung, 2011).

The previous literature has indicated the relationship between acculturation and

ACP among immigrant elders (Bito et al., 2007; Gao et al., 2015; Matsumura et al., 2002;

Wittenberg-Lyles, Villagran, & Hajek, 2008). Wittenberg-Lyles et al. (2008) revealed

that Latinos with higher levels of acculturation were significantly more likely to recall

hearing about ADs and significantly more likely to have completed an AD.

For Asian Americans, Matsumura et al. (2002) surveyed 539 English-speaking

Japanese Americans, 340 Japanese-speaking Japanese Americans, and 304 Japanese

living in Japan and found that acculturation was associated with a greater preference for

respondents to participate in decision making. Furthermore, English-speaking Japanese

Americans, who had higher acculturation levels, expressed more positive attitudes toward

ACP than Japanese-speaking Japanese Americans and respondents living in Japan. Gao et

al. (2015) found that Chinese American elders with higher acculturation levels and those

residing more than 20 years in the U.S. were more likely to have heard of AD after

controlling for the effects of demographics, health, and experiences of end-of-life care.

Given the influence of acculturation level on multicultural populations, this dissertation

will examine the role of acculturation in the theoretical frameworks. The research

questions will be asked after theoretical frameworks are discussed.

Family Influence and Advance Care Planning

The quality of the relationships within the family can affect the effectiveness of

ACP (Blackford & Street, 2016; Boerner et al., 2013; Kramer, Boelk, & Auer, 2006).

Boerner et al. (2013) surveyed 293 participants aged 55 and older and found that better

overall family functioning (e.g., sharing thoughts and feelings with one another and

collaborative problem solving) increased the odds of discussions about end-of-life care.

Furthermore, this study found a stronger effect of family functioning on discussions

about end-of-life care (odds ratio = 2.79) compared with the two-pronged approach (i.e.,

having both formal ADs and informal ACP discussions). Emotional support from spouse

increased the odds of having discussions about end-of-life care (odds ratio = 1.88).

Different from Boerner et al. (2013) focusing on older adults in the U.S., Lee et

al. (2014) systematically reviewed the 15 empirical studies (published from 1996 to

2012) regarding ACP and AD among Chinese population in Eastern (e.g., Hong Kong)

and Western cultures (e.g., U.S. and Canada). This review suggested that family was an

important topic for Chinese individuals and a family decision-making model may be

more appropriate for discussions with patients and families rather than focusing on

individuality, autonomy, and self-determination.

The importance of family on Chinese depends on the influence of collectivism

(Sun, Gao, & Coon, 2015). This cultural orientation encourages people to prioritize

family responsibilities over individual independence. China has been a predominantly

agricultural country, which emphasizes working labor groups on which individuals

depend. Individuals are organized to cooperate and support each other to survive. De

Bary (1998) considered a family as “the predominant social and economic institution in

an agricultural society and in many aspects it furnished the theoretical model for other

institutions such as the patriarchal dynastic state” (p. 17). Take family responsibilities has

been embedded in Chinese individuals’ value system. Because individual autonomy is

not given a priority in Chinese culture, it is common that older adults would like to rely

on their family members for health decision making. In this regard, it is helpful to

consider family dynamics when we examine older adults’ ACP behaviors.

As an important indicator of family functioning, family cohesion is defined as

shared affection, support, helpfulness, and caring among family members (Barber &

Buehler, 1996). It has been developed to examine the relationship with psychological

adjustments and the previous literature showed that family cohesion may serve as a

buffer to psychosocial stressors among multicultural populations in the U.S. such as

Latinos (Baer & Schmitz, 2007; Dillon, De La Rosa, & Ibañez, 2013; Guo, Li, Liu, &

Sun, 2014; Rivera et al., 2008; Ta, Holck, & Gee, 2010). For example, Dillon et al.

(2013) showed that more acculturative stress had a significantly greater decline in family

cohesion among Latinos and implied that high levels of cohesion may help protect

participants from acculturative stress. Rivera et al. (2008) found that higher family

cohesion was significantly associated with lower psychological distress among Latinos.

Furthermore, there was a significant interaction between family cohesion and family

cultural conflict, which suggests that although higher-level family cohesion was

associated with lower psychological distress, having strong family cohesion in the face of

family cultural conflict relates to greater psychological distress.

Although the values of family cohesion (e.g., loyalty and solidarity) are favored

among Chinese, it is scarce to apply family cohesion to examine Chinese American older

adults’ health-related behaviors. To my knowledge, the only existing study with family

cohesion as a variable for Chinese American older adults was used to examine the

influence of cohesion level on perceived threat of Alzheimer’s Disease (Sun et al., 2015).

Sun et al. showed that family cohesion served as a nonsignificant predictor of perceived

threat of Alzheimer’s Disease, but their findings indicated that family cohesion (r = −.14,

p < .01) was negatively associated with perceived threat of Alzheimer’s Disease among

385 Chinese Americans aged 55 and older. Given that the importance of family cohesion

on Chinese communities, this study will consider the influence of family cohesion on the

target population’s ACP behaviors. The research questions would be asked after

theoretical frameworks are discussed.

Theory of Planned Behavior

The Theory of Planed Behavior (TPB) follows a reasoned action approach to

focus on individual determinants that influence behavioral performance. Fishbein and

Ajzen (2010) did not assume that people are rational. Instead, the TPB includes both

deliberate and spontaneous decision-making process. They assumed that people’s

attitude, subjective norms, and perceptions of control follow in a reasonable and

consistent ways from their beliefs.

Figure 1. The Theory of Planned Behavior.

The TPB emphasize that individual’s behavioral intention serves as the most

immediate predictor for behavior. The previous meta-analyses showed the positive

correlation between behavioral intention and actual behavior (Alberracian, Johnson,

Fishbein, & Muellerleile, 2001). Behavioral intention is conceptualized as “indication of

a person’s readiness to perform the behavior” (Fishbein & Ajzen, 2010, p. 39).

Researchers measure behavioral intention by asking participants to estimate how likely

they will perform a behavior. To test the relationship between behavioral intention and

behavior, whether participants perform a specific behavior should be measured sometime

after behavioral intention is measured. However, a variety of the TPB studies did not

measure behavior prospectively. Instead, these studies excluded behavior from this model

or measured behavior retrospectively because past behavior is highly correlated to future

behavior (Fishbein & Ajzen, 2010).

The TPB was based upon the development of Theory of Reasoned Action (TRA).

The TRA postulates that attitudes and subjective norms jointly predict an individual’s

behavioral intention that may lead to behavioral performance (Fishbein & Ajzen, 1975;

Fishbein & Ajzen, 2010). Attitude is defined as “a latent disposition or tendency to

respond with some degree of favorableness or unfavorableness to a psychological object”

(Fishbein & Ajzen, 2010, p. 76). It is measured through the overall evaluation that reflect

the instrumental (i.e., the usefulness of the recommended behavior) and experiential (i.e.,

how enjoyable the behavior is) aspects toward a specific behavior (Yzer, 2013).

Behavioral beliefs (i.e., perceived consequences of performing the behavior) weighted by

outcome evaluations (i.e., evaluations of those consequences) are determinants of

individuals’ attitude. When individuals consider the importance of performing a behavior

on positive outcomes, it is likely for them to have a positive attitude toward specific

behaviors. In this regard, attitudes can be measured indirectly through behavioral beliefs

and outcome evaluations.

Subjective norms are another predictor of people’s behavioral intention in the

TRA. They are conceptualized as a person’s perception that most people who are

important to him or her think he or she should or should not perform a behavior (Fishbein

& Ajzen, 2010). Subjective norms refer to “a specific behavioral prescription or

proscription attributed to a generalized social agent” (Fishbein & Ajzen, 2010, p. 131).

They are used to deal with the influence of social environment on individuals’ behavior

(Fishbein & Ajzen, 1975). Two types of norms, injunctive (i.e., how likely important

others think I should perform a recommended behavior) and descriptive norms (i.e., how

likely significant others perform this recommended behavior themselves) are included in

the measurement of normative influence. The original use of subjective norms

emphasizes the injunctive nature. However, as well as the injunctive normative influence,

we may also experience normative pressure when we know important others perform a

behavior or not. Subjective norms are determined by normative beliefs weighted by

motivation to comply. Normative beliefs illustrate the perceptions certain important

others have about a person’s behavioral performance. Motivation to comply refers to the

extent to which people want to behave as important others prescribe. Knowing about

people’s normative beliefs may not be sufficient to understand the perceived norms,

because people may ignore what important others prescribe. Fishbein and Ajzen (2010)

believed that when people care about important others’ approval or disapproval of their

behavioral performance, they would be likely to intend to perform a behavior. Therefore,

Fishbein and Ajzen (2010) argued that it was necessary to measure normative beliefs

weighted by the motivation to comply. However, the previous literature suggested that

multiplying normative beliefs by motivation to comply added little or nothing to the

prediction of perceived norms (Fishbein & Ajzen, 2010).

The TRA helps explain the motivational factors that predict intentions and

behaviors under volitional control (i.e., the degree to which a person can exercise control

over the behavior; Ajzen, 1991). However, sometimes individuals’ intention to perform a

behavior are thwarted by a lack of perceived capability of performing the behavior (Stiff

& Mongeau, 2016). To increase the predictive power of behavioral intentions and

behavior, Ajzen (1985) developed the TRA into the TPB through the inclusion of

perceived behavioral control to accommodate the nonvolitional nature of behaviors.

The concept of perceived behavioral control is based on Bandura’s concept of

self-efficacy. Fishbein and Ajzen (2010) said that perceived behavioral control and self-

efficacy are conceptualized similarly. Perceived behavioral control is defined as

“people’s perceptions of the degree to which they are capable of, or have control over,

performing a given behavior” (Fishbein & Ajzen, 2010, p. 64), while self-efficacy is

conceptualized as “people’s beliefs about their capabilities to exercise control over their

own level of functioning and over events that affect their lives” (Bandura, 1991, p. 257).

Both concepts are concerned with perceived ability to perform a behavior.

TPB is based upon the assumption that people’s confidence level in their

capability of performing a behavior has a positive influence on individuals’ intention to

perform a behavior. In addition to behavioral intention, perceived behavioral control is

used to predict behavior directly in the TPB (Ajzen, 1991). Holding intention equal, the

more perceived behavioral control people have, the more likely it is that people will

perform a behavior. According to Fishbein and Ajzen (2010), perceived behavioral

control consists of capacity and autonomy. Capacity represents people’s perceptions of

their ability to perform a behavior and autonomy illustrates people’s perceptions of their

control over performing a behavior. Perceived behavioral control is determined by

control beliefs (i.e., people’s perceptions of having resources available to perform a

behavior) weighted by power beliefs (i.e., the extents to which having resources available

to perform a behavior is sufficient to overcome barriers to perform the behavior). It is

assumed that when people are confident that they have resources to overcome challenges

and perform a behavior, they have perceived behavior control towards behavioral

performance. In TPB, attitude, as well as subjective norms, and perceived behavioral

control work as a set of predictors for behavioral intention that may cause actual

behavior. Adding perceived behavioral control helps explain individuals’ behavioral

intention and behavior significantly better than the TRA (Cooke & French, 2008).

Perceived behavioral control and self-efficacy have been used interchangeably

(Armitage & Conner, 2001). Despite the conceptual similarities between perceived

behavioral control and self-efficacy, Ajzen (2002) explained perceived behavioral control

as a combination of perceived self-efficacy (i.e., ease or difficulty of performing a

behavior) and control (i.e., beliefs about the extent to which performing the behavior is

up to the actor) and mentioned that several studies provided consistent support for the

distinction between perceived self-efficacy and control. He also found that whereas the

addition of perceived self-efficacy improved the prediction of intentions, perceived

control had no significant effects on intentions. Perceived control may predict intentions

only when combined with self-efficacy items. In other words, perceived behavioral

control and self-efficacy can have different predictive power of intentions and behaviors.

Distinguishing perceived behavioral control and self-efficacy, Downs and

Hausenblas (2005) conducted a meta-analytic review of 111 TRA/TPB studies and found

that the association for perceived behavioral control–behavior (effect size = 0.67,

standard deviation = 0.07, n = 92) was not significantly different than self-efficacy–

behavior (effect size = 0.49, standard deviation = 0.04, n = 33) [QB (2) = 10,206.51, p <

0.01]. The association for self-efficacy–intention (effect size = 1.17, standard deviation =

0.05, n = 25) was significantly greater than perceived behavioral control–intention (effect

size = 1.04, standard deviation = 0.05, n = 103) and perceived-barriers intention (effect

size = -0.36, standard deviation = 0.06, n = 17) [QB (2) = 43,410.11, p < 0.01]. That is,

self-efficacy serves as a stronger predictor for intentions.

The current study would use self-efficacy in the TPB instead of perceived

behavioral control. Firstly, it is quite challenging to translate perceived control measures

from English to Mandarin. In Mandarin self-efficacy focuses on people’s abilities, while

perceived control items seemed to be abstract, Westernized, and hard to understand.

Secondly, Fishbein and Ajzen (2010) conceptualized perceived behavioral control as

consisting of capacity and autonomy. Given the interdependent nature of the Chinese

culture, older adults’ health promotive behaviors are frequently performed beyond

individual autonomy. It would be helpful to focus on capacity rather than autonomy. In

this dissertation, attitudes, subjective norms, and self-efficacy1 would be used to predict

participants’ intentions to plan for end-of-life care. The hypotheses would be addressed

after the review of TPB literature.

Theory of Planned Behavior in Health Topics

The TPB have been widely used in the correlational studies to understand and

predict human behavior across various health-related topics such as smoking, alcohol

consumption, healthy eating, physical activity, condom use, and sun screening

(Alberracian et al., 2001; Cooke & French, 2008; Cooke, Dahdah, Norman, & French,

2014; McEachan, Conner, Taylor, & Lawton, 2011; Riebl et al., 2015; Topa & Moriano,

2010). The TPB is considered a useful theoretical framework to help people understand

the importance of promoting health and well-being.

Armitage and Conner (2001) reviewed 161 articles that included 185 independent

empirical tests of the TPB. They found that the average multiple correlation of intention

and perceived behavioral control with behavior was .52, accounting for 27% of the

variance. The average multiple correlation of attitude, subjective norms, and perceived

behavioral control with intention was .63, which accounts for 39% of the variance. The

subjective norms-intention correlation was significantly weaker than attitude-intention

and perceived behavioral control-intention correlations. Perceived behavioral control

added around 6% to the prediction of intention above attitude and subjective norms and

this implied the unique contribution of perceived behavioral control to the TPB.

McEachan et al. (2011) reviewed 206 articles that included 237 prospective tests

of health-related behaviors. They found that intention had the strongest relationship with

prospective behaviors (mean ρ = .43), and this represented a medium-large effect size.

Direct measures of attitude and perceived behavioral control showed medium-sized

relationships with behavior (both mean ρ = .31) as well. In terms of predicting intention,

direct attitude showed the strongest correlation with mean ρ of .57, followed by perceived

behavioral control (direct mean ρ = .54) and subjective norms (direct mean ρ = .40). It is

also found that intention and perceived behavioral control accounted for 19.3% of the

variance in behavior and intention was the main predictor of behavior contributing three

times more to the final equation (B =.37) than perceived behavioral control (B = .11).

With regard to the prediction of intention, attitude (B = .35) served as the strongest

predictor, followed by perceived behavioral control (B =.34) and subjective norms (B

= .15). These three predictors accounted for 44.3% of the variance in intention. The

findings were consistent with the previous literature showing that attitude served as a

strong predictor. The inclusion of perceived behavioral control increased a significant

amount of variation in intentions and behaviors in the TPB. Intention worked as the

primary predictor of behavior.

Theory of Planned Behavior in Advance Care Planning

Although the TPB has been widely used across different health contexts, only a

few ACP applied the TPB qualitatively (Kataoka-Yahiro, Yancura, Page, & Inouye,

2011; Lee, Byon, Hinderer, & Alexander, 2017) and quantitatively (Hong, Casado, &

Lee, 2018; Nahapetyan, Orpinas, Glass, & Song, 2017). Qualitative studies focused on

the examination of behavioral, normative and control beliefs, while quantitative studies

used cross-sectional survey to study the relationships among TPB measures. In general,

the TPB was found to provide strong support to guide future educational interventions in

ACP-related topics.

Kataoka-Yancura et al. (2011) conducted four focus group sessions (field notes of

focus groups included) among 14 Asian Pacific Islander family caregivers of patients

receiving hemodialysis for Stage 4 to 5 chronic kidney disease. Attitude, subjective

norms, and perceived behavioral control were the three major measures with linkages to

categories and subcategories. In their descriptive analysis, attitude included three major

categories: benefits of completing ACP and AD (e.g., minimizing burden and family

disputes, and foreseeing problems ahead of time), barriers (e.g., unwillingness to think of

death, denial of negative consequences, and lack of knowledge), and triggers (e.g.,

diagnosis with a terminal illness). Subjective norms included information related to

participants’ perception of the social prevalence and desirability of AD completion and

had two major categories: people outside the family (e.g., health professionals) and social

opinion. The participants believed that family was primary in decision making followed

by the opinions of others based on different viewpoints of health care professionals.

Social opinion included social norms expressed by the participants such as “Those who

have ACP are cared for well” and “Most people have AD.” Perceived behavioral control

depends on family dynamics and was linked to family member’s role and family

member’s communication style (e.g., lack of consensus building was associated with

inability to complete an AD).

Similarly, Lee et al. (2017) applied a qualitative descriptive design to examine

behavioral, normative, and control beliefs in the ACP discussion among 60 community-

dwelling Chinese Americans. The participants were divided into two groups by age. The

authors grouped 30 participants aged 65 years and older in the older group and another 30

participants in the younger group. Through focus groups, observation of group

interaction, and the non-verbal communication, Lee et al. focused on the similarities and

differences of beliefs and cultural implications in ACP among different generations.

This study found that most of the participants in both groups had positive attitudes

towards ACP and believed that ACP produced good outcomes for patients and their

families and lessened the burden of others in making end-of-life care decisions. In terms

of normative beliefs, participants in both groups believed that the discussing death and

dying and planning for end-of-life are taboos in Chinese society. The younger group

perceived that the discussion of ACP would be unpleasant and difficult and would upset

the senior members in their families, while the older group participants perceived that the

discussion of ACP would be unwelcome and upsetting to their children. Interestingly,

participants in the younger group found it easier to have ACP conversations among their

generation or with their children. When it comes to control beliefs, both groups expressed

that the biggest barrier to ACP discussions is lack of an appropriate opportunity. The

older group participants expressed lack of knowledge about ACP. Some of them had

never heard of ACP and felt confused this with making a will, euthanasia, or making a

funeral arrangement. After understanding the definition of ACP participants in this group

believed that it was difficult to discuss ACP because they lacked personal support and

necessary materials such as specific ACP information, Chinese language support,

appropriate translated forms, and counseling services. For the younger group, it is quite

challenging to be the surrogate because they did not know their parents’ wishes.

To complement qualitative TPB-based studies, Hong et al. (2018) applied the

TPB and prior research to examine the relationships between acculturation, attitudes,

subjective norms, perceived control, and intention to discuss ACP for a family member

with Alzheimer’s disease. The authors conducted path analyses by using a cross-sectional

convenience sample of 261 Korean Americans aged 40 and older. Age, gender,

education, and knowledge about Alzheimer’s disease and ACP were included as

covariates. It is found that attitude (β = .271, p < .001) and subjective norms (β = .412, p

< .001) were associated with intention for ACP discussion for a family member with

Alzheimer’s disease. However, perceived control was not associated with intention for

ACP discussion either directly or indirectly. Acculturation was not associated with any of

the three determinants of the TPB or intention for ACP discussion for a family member

with Alzheimer’s disease. Among covariates, only knowledge about Alzheimer’s disease

(β = .137, p = .010) was associated with intention for ACP discussion.

Different from Hong et al. (2018) focusing on discussing ACP for a family

member with Alzheimer’s disease, Nahapetyan et al. (2017) surveyed 146 Caucasian

Americans aged 60 and older and found that intentions to use hospice was significantly

correlated with hospice knowledge, attitude, subjective norms, perceived control to use

hospice, and preferences for comfort care. The multiple regression analyses showed that

higher hospice knowledge (β = .23, p < .001), higher subjective norms that support

hospice utilization (β = .21, p = .004), higher perceived control to use hospice (β = .41, p

< .001), and preferences for end-of-life care (β = .15, p = .019) were significantly

associated with intentions to use hospice. Together, these variables explained 54% of the

variance in intentions to use hospice.

These studies have demonstrated the effectiveness of the TPB regarding ACP

behaviors, but none of them addressed the needs for Chinese American older adults. To

provide recommendations to design educational interventions to increase Chinese

American older adults’ ACP awareness and engagement, it is crucial to examine how the

TPB influences Chinese Americans’ intentions to discuss end-of-life care plans with

family members and complete an AD respectively:

RQ1a-b: What are participants’ attitudes, subjective norms, self-efficacy, and

behavioral intentions toward (a) discussing end-of-life care plans with family members

and (b) completing an advance directive?

RQ2: Are participants’ attitudes, subjective norms, self-efficacy, and behavioral

intentions toward (a) discussing end-of-life care plans with family members and (b)

completing an advance directive different from each other?

H1a-b: Participants’ attitudes, subjective norms, and self-efficacy will positively

predict behavioral intentions toward (a) discussing end-of-life care plans with family

members and (b) completing an advance directive.

RQ3a-b: Does acculturation moderate the relationships between TPB measures

(i.e., attitude, subjective norm, and self-efficacy) and behavioral intentions toward (a)

discussing end-of-life care plans with family members and (b) completing an advance

directive.

RQ4a-b: Does family cohesion moderate the relationships between TPB measures

and behavioral intentions toward (a) discussing end-of-life care plans with family

members and (b) completing an advance directive.

Health Belief Model

The Health Belief Model (HBM) has been extensively used in health behavior

research to explain preventive health behavior and provide recommendations for health

behavioral intervention (Champion & Skinner, 2008). The HBM provides a connection

between beliefs and behaviors and explains what beliefs should be considered for the

targeted population in health interventions. The original HBM model emphasizes that

some factors can explain why people behave to prevent and control health problems,

including perceived susceptibility, perceived severity, perceived benefits, perceived

barriers, self-efficacy, and cues to action.

Figure 2. The Health Behavior Model.

The original HBM includes the four components: perceived susceptibility,

perceived severity, perceived benefits, and perceived barriers (Bartholomew, Parcel, Kok,

& Gottlieb, 2006). It emphasizes that people are likely to take action to reduce the threats

they have, if the following conditions are met: 1) they believe that they are susceptible to

an illness condition (perceived susceptibility), 2) they believe that this condition may

have serious negative impacts on their life (perceived severity), 3) they believe that

taking certain actions may help them reduce the susceptibility or severity of the condition

or produce other positive outcomes (perceived benefits), 4) they believe that the

anticipated benefits of taking action outweigh the potential costs to take action (perceived

barriers). Here perceived susceptibility refers to people’s perceived likelihood of getting

an illness condition. Perceived severity refers to people’s perceived seriousness of the

illness condition as well as the sequential negative consequences. Perceived benefits refer

to people’s perceived benefits of taking actions for the threat reduction. Perceived

barriers refer to people’s perceived negative aspects of taking actions.

The original HBM highlights that whether people decide to take actions is

influenced by their perceived susceptibility to and perceived severity of an illness

condition as well as perceived benefits and barriers. This decision-making process is

driven by cues to action (Rosenstock, 1974). Here cues to action refer to the strategies or

reminders to trigger the readiness to take actions internally (e.g., body pain and

symptoms of a disease) and externally (e.g. media exposure and a friend’s experience

with the illness condition). Although it is necessary to consider cues to action in the

HBM, this variable has been rarely studied due to the fleeting nature and cues to action

vary across different contexts (Champion & Skinner, 2008).

In the earlier stage of the HBM development, researchers focused on

circumscribed preventive actions, such as accepting immunizations (Rosenstock,

Strecher, & Becker, 1988). It was common that participants had sufficient capabilities to

perform the recommended behavior. However, when researchers considered more

complicated problems associated with certain behaviors like healthy eating and exercise,

an appropriate amount of efficacious influence might be required for people to take

actions. Later, Rosenstock et al. (1988) added self-efficacy in the original HBM to

account for initiation and maintenance of behavioral change. As addressed earlier in this

chapter, self-efficacy is defined as “people’s beliefs about their capabilities to exercise

control over their own level of functioning and over events that affect their lives”

(Bandura, 1991, p. 257). That is, people must feel self-efficacious to take actions to

achieve behavior change, when they perceive susceptibility and severity and believe that

performing a recommended behavior can lead to a positive outcome.

Health Belief Model in Health Topics

Researchers have conducted several systematic reviews to examine the

effectiveness of the HBM to predict behaviors (Carpenter, 2010; Harrison, Mullen, &

Green, 1992; Janz & Becker, 1984; Zimmerman & Vernberg, 1994). The previous studies

have been found to provide substantial empirical evidence to support the HBM to explain

and predict health-related behaviors.

Janz and Becker (1984) found studies with prospective design yielded

significance ratios as good as or better than those with retrospective design. Janz and

Becker categorized the included studies into three topics, preventive health behaviors

(action taken to prevent illness), sick-role behaviors (action taken after the medical

diagnosis to prevent further illness progress), and clinic visits (clinic utilization for a

variety of reasons). Overall, perceived susceptibility, benefits, and barriers were good

predictors of behavior, while severity was weak. The results suggested that perceived

barriers served as the most powerful variable in the HBM across different behaviors and

designs. Perceived susceptibility was a stronger predictor to understand preventive health

behaviors rather than sick-role behaviors, while perceived benefits remained a stronger

predictor to sick-role behaviors rather than preventive health behaviors. This review

focused mainly on statistical significance test and failed to provide specific estimates of

the strength of the relations between HBM dimensions.

To estimate the strength of the relationship between HBM dimensions and health-

related behaviors, Harrison et al. (1992) reviewed 16 studies that included reliability

measures, all the four major dimensions (i.e., susceptibility, severity, benefits, and costs)

in the original model, and a behavioral dependent variable to evaluate the predictive

validity of the HBM. They found that 22 of the 24 mean effect sizes were positively

significant and the effect sizes for the four dimensions varied across different studies. The

results yielded low to moderate effects of participants’ susceptibility (r = .15), severity (r

= .08), benefits (r = .13), and costs (r = -.21). The results indicated that retrospective

studies had significantly larger effect sizes for perceived benefits and costs and smaller

effect sizes for severity than prospective studies.

Different from the previous reviews, Carpenter (2010) exclusively incorporated

studies that measured HBM variables at time one and measured health-related behaviors

associated with those variables at time two to decide whether HBM variables could

predict behaviors longitudinally. The results yielded low to moderate relationships

between participants’ perceived severity, benefits and barriers, and likelihood of

performing the target behavior. Benefits and barriers worked as stronger predictors of

behavior, while severity provided a low estimate for behavior and susceptibility served as

the weakest predictor in this meta-analysis. However, the author did not explain why

susceptibility–behavior relationship was so small. In terms of time between measures as a

moderator, the longer periods of time were associated with the weaker effects of HBM

variables except barriers. In other words, HBM variables were more likely to be

positively related to health-related behaviors when these behaviors were measured shortly

after HBM variables were measured. Barriers were not likely to be influenced by the time

length as many of them may change little over time. When it comes to the type of

outcome (treatment vs. prevention), benefits and barriers were better predictors of

prevention behaviors rather than treatment behaviors.

Health Belief Model in Advance Care Planning

The HBM may provide a reasonable explanation regarding what cognitive beliefs

influence the ACP-related behaviors (Hamel, Guse, Hawranik, & Bond, 2002; Kent,

1996; VandeCreek & Frankowski, 1996). Ko (2008) examined the ethnic differences in

ACP in terms of the completion of AD and end-of-life communication and compared 112

Korean American and 105 non-Hispanic White adults aged 65 and older. Hierarchical

multiple regression and logistic regression analysis were conducted to test the

relationship among the primary HBM variables, knowledge, completion of an AD, and

end-of-life communication.

The results indicated that Korean Americans were significantly less likely than

non-Hispanic Whites to complete an AD and have end-of-life care discussions. Non-

Hispanic Whites were more likely to perceive susceptibility, severity, and benefits about

ACP than Korean Americans, while Korean Americans were more likely to perceive

barriers about ACP than non-Hispanic Whites. After controlling for the effects of

demographic variables, knowledge had a direct positive effect and perceived barriers had

a direct negative effect on the completion of an AD, while perceived susceptibility,

severity, and benefits did not predict the completion of an AD. Also, knowledge and

perceived severity had a direct positive effect, and perceived barriers had a direct

negative on end-of-life communication after controlling for the effects of social-

demographic variables, while perceived susceptibility and benefits did not predict end-of-

life communication. Ethnicity had a significant indirect effect on the completion of an

AD and end-of-life communication through ACP knowledge, but not through any of

HBM variables. This study indicated the important role of ethnicity on ACP and its effect

on knowledge as a mediator. However, the results of this study did not fully support the

HBM and this suggested a combination of HBM and other models to explain ACP-

related behaviors.

Szalai (2015) developed a difficult conversation model that included individual

and relational predictors of communicative behaviors to identify college students’

engagement to have ACP conversations about their own preferences through formative

research and additional two studies. Although college students and the target population

of the current study have different demographic characteristics, Szalai’s study would

inform the current study in terms of theoretical framework and research topic.

In formative stage, Szalai conducted six focus groups among college students and

asked about their ACP knowledge, personal experience, and willingness to participate to

ACP as well as perceived self-efficacy, perceived severity, perceived susceptibility,

perceived barriers, perceived benefits, anticipated response (e.g., individuals are more

likely to discuss ACP with a family member when they anticipate this family member’s

positive reaction), and relational closeness regarding ACP conversations. Here perceived

severity was evaluated based upon an imaginary condition regarding how serious a

concern it was to be in a situation where the participant could not communicate his or her

treatment wishes, but he or she had not done ACP ahead of time. The benefits of having

ACP conversations included allowing individuals’ wishes to be respected, decreasing

familial burden, and preventing conflict. The barriers to having ACP conversations

included families’ negative reactions, unawareness of ACP conversations, lacking

knowledge of having ACP conversations, and unimportance of ACP conversations for

healthy young people. The results showed that the majority of participants were not

aware of ACP, but they were willing to have ACP conversations. People with greater

perceived susceptibility, greater perceived severity, more frequent cues to action, greater

self-efficacy, fewer perceived barriers, more positive anticipated responses, and greater

relational closeness will be more likely to have ACP conversations.

Based on the results of focus group in informative research, Szalai tested HBM

variables and relational closeness to predict college students’ intentions to have ACP

conversations with a family member through hierarchical regression analysis in Study

One. The results indicated that participants with greater perceived susceptibility, greater

perceived self-efficacy, and greater perceived benefits were more likely to have ACP

conversations. Study Two integrated the HBM, the TPB, and the disclosure decision-

making model (DD-MM) constructs and conducted path analysis to evaluate the

effectiveness of the difficult conversation model. Based on the results, this study

speculated a model suggesting that a combination of the HBM, TPB, and DD-MM

variables can predict participants’ intent to have ACP conversations. The findings

provided empirical support for the contribution of relational variables to ACP

conversations. However, given the potential unimportance of ACP conversations for

healthy young people, the findings from this study may not be generalized to nonstudent

populations. Research focusing nonstudent populations is needed to examine the

effectiveness of the HBM and TPB.

Guided by the previous studies, the following questions and hypotheses are

proposed:

RQ5a-b: What are participants’ perceived susceptibility, perceived severity,

perceived barriers, and perceived benefits toward (a) discussing end-of-life care plans

with family members and (b) completing an advance directive?

RQ6: Are participants’ perceived susceptibility, perceived severity, perceived

barriers, and perceived benefits toward (a) discussing end-of-life care plans with family

members and (b) completing an advance directive different from each other?

H2a-b: Participants’ perceived susceptibility, perceived severity, perceived

barriers, perceived benefits, and self-efficacy will predict behavioral intentions toward (a)

discussing end-of-life care plans with family members and (b) completing an advance

directive.

RQ7a-b: Does acculturation moderate the relationships between HBM measures

(i.e., perceived susceptibility, perceived severity, perceived barriers, perceived benefits,

and self-efficacy) and behavioral intentions toward (a) discussing end-of-life care plans

with family members and (b) completing an advance directive?

RQ8a-b: Does family cohesion moderates the relationships between HBM

measures and behavioral intentions toward (a) discussing end-of-life care plans with

family members and (b) completing an advance directive?

Comparing Theory of Planned Behavior and Health Belief Model

Both the TPB and HBM have been extensively used to explain and predict health

behavior and provide recommendations for health behavioral interventions (Champion &

Skinner, 2008). Several previous studies tested both theories simultaneously and

compared them against one another and the TPB has been found to demonstrate a

stronger fit to the data or account for more variance than the HBM (Gerend & Shepherd,

2012; McClenahan, Shevlin, Adamson, Bennett, & O’Neill, 2007; Montanaro & Bryan,

2014; Şimşekoğlu, & Lajunen, 2008; Thornton & Calam, 2010; Yang, 2015).

However, most of these studies tested both theories among student samples. As

the previous literature addressed, the differences between student and nonstudent samples

vary across different contexts (Hanel & Vione, 2016; Peterson & Merunka, 2014). For

example, when Notani (1998) reviewed the TPB-based studies systematically and found

that nonstudent samples should provide better predictions of behavior from perceived

behavioral control compared to student samples, while the relationship between

perceived behavioral control and intentions is significant only for the student samples.

Therefore, it can be problematic to generalize findings from student to nonstudent

samples. It is worthwhile to compare the TPB and HBM among a nonstudent sample to

guide future interventions tailored for the target population. Furthermore, to my

knowledge, both theories have not been tested simultaneously regarding ACP. My

dissertation would address this research gap by comparing the utility of the two theories

in predicting behavioral intentions critical to end-of-life care among Chinese American

older adults:

H3a-b: The Theory of Planned Behavior has more predictive power for

participants’ intentions toward (a) discussing end-of-life care plans with family members

and (b) completing an advance directive than the Health Belief Model.

CHAPTER 3: METHODOLOGY

Overview

The purpose of this research was to understand the factors that influence Chinese

American older adults’ intentions to plan for end-of-life care through the Theory of

Planned Behavior (TPB) and the Health Belief Model (HBM). Two recommended

behaviors related to planning for end-of-life care were studied: discussing end-of-life care

plans with family members and completing an advance directive (AD). The target

population was Chinese Americans aged 55 and older living in the metropolitan Phoenix,

Arizona. Participant recruitment occurred through Chinese senior community settings.

Participants were randomly assigned to answer a set of questions deriving from the TPB

and HBM regarding either one of the two recommended behaviors. Dependent variables

were participants’ intentions to discuss end-of-life care plans with family members and

complete an AD respectively.

Research Design

Quantitative survey methodology building upon qualitative interviews as

formative research was applied to examine the factors that influence Chinese American

elders’ intentions to plan for end-of-life care. Researchers frequently use qualitative

interviews to “verify, validate, or comment on information obtained from other sources

[and] achieve efficiency in data collection” (Lindlof & Taylor, 2011, p. 175). Therefore,

interviews were conducted to inform the development of data collection instrument in the

cross-sectional survey. The objectives of qualitative interviews were to: 1) explore the

target population’s basic understandings of the advantages and disadvantages of planning

for end-of-life care, 2) understand participants’ preference for the two recommended

behaviors associated with planning for end-of-life care, and 3) identify when participants

would like to plan for end-of-life care if they have not done yet. Participants’ responses

were used to inform the design of the cross-sectional survey.

After formative research, quantitative survey methodology was used because it

offers researchers a tool to provide statistical estimates of the characteristics of the

targeted population and then generalize the findings to a larger population to help

alleviate social problems (Andres, 2012; Fowler, 2014). More specifically, the better we

understand the influence of predictors on Chinese American older adults’ intentions to

plan for end-of-life care, the more likely it is that we will be able to provide

recommendations to improve ACP awareness and engagement. The objectives of the

cross-sectional survey were to: 1) compare two theories of health behavior, the TPB and

HBM, in their prediction of Chinese American older adults’ intentions to discuss end-of-

life care plans with family members and complete an AD respectively; 2) identify the

roles of acculturation and family cohesion on the target population’s intentions to discuss

end-of-life care plans with family members and complete an AD respectively.

Participants

To be considered eligible in this research, participants should be aged 55 and

older, speak English or Chinese, have U.S. citizenship or permanent residency, and living

in the metropolitan Phoenix areas. Although many developed countries accepted 65 years

of age or older as the older population (World Health Organization, 2000), people whose

age is 55 and older are included in this study to target a broader population.

Procedures

After approval from the university institutional review board, participants were

recruited from a beginner-level English class in one Chinese American subsidized senior

housing center in Phoenix. I volunteered to teach this eight-week, beginner-level English

class. At the end of the last class, students were informed of the participation opportunity

for this formative research. A consent letter was given to people who would like to

participate in this study. Participants were asked to have a semi-structured interview.

They can choose either English or Chinese speaking to respond to the interview

questions. All the interviews were audio recorded and transcribed. In return for their

participation, each participant was given a $10 gift card.

Participants’ responses were coded and used to inform the development of

quantitative questions regarding the TPB and HBM. Firstly, participants were given the

TPB and HBM scales (translated from English to Chinese) to provide feedback whether

they were able to understand each scale item to help make the cross-sectional survey

understandable. Secondly, their responses guided the identification of the recommended

behaviors to inform the survey. Some participants preferred to complete an AD over ACP

discussions, because it is more reliable to document their medical preferences in a written

legal document, while some others believed that they do not need an AD and having ACP

discussions are appropriate enough. Because it was difficult to select which of the two

recommended behaviors associated with planning for end-of-life care would be more

helpful to promote ACP on end-of-life care among Chinese American older adults, this

dissertation focused on these two recommended behaviors instead of one of them.

Thirdly, many participants gave ambiguous answers (e.g., “when I am sick enough” and

“when I need”) regarding when they would like to plan for end-of-life care, while only

one participant said he would like to discuss his end-of-life care plans within three years.

Given that the target population may have low-level awareness of planning for future

care plans, having the time at which a behavior is performed may not help explain

behaviors, although Fishbein and Ajzen (2010) suggested that a specific behavior should

include four elements, “the action performed, the target at which the action is directed,

the context in which it is performed, and the time at which it is performed” (p. 29).

After formative research, a cross-sectional survey was conducted among the

target population in community-based settings, including but not limited to senior

centers, subsidized senior housing apartments, religious sites, senior social clubs, and

community events. The recruitment strategies included delivering printed materials such

as flyers and posters in Mandarin, Cantonese, and English in a wide variety of

community settings, contacting potential participants face-to-face or on the telephone,

providing referrals (e.g., from community leaders, health professionals, past participants,

and friends), and developing partnerships with community organizations (e.g., offering

members workshops and classes and volunteering in community events), and offering

Chinese grocery market gift cards as research incentive. Developing partnerships with

community organizations and providing referrals were found to be the most effective

strategies for recruiting participants in this study.

Several of our participants’ recruitment requests were declined when potential

participants: 1) were emotionally resistant to questions related to death and dying

process, 2) were still concerned that their private information would be disclosed to

others after that knowing their responses would be anonymous and confidential, 3)

believed that the survey was too long to complete, 4) were under medical treatments and

did not feel comfortable to participate, 5) did not have time to participate due to busy

schedules, or 6) had low literacy levels in Chinese or English.

A Chinese (either Mandarin or Cantonese) or English-written questionnaire

packet was given to people who agreed to participate. Although the questionnaire was

designed to be self-administered, three student workers were available to help potential

respondents complete the questionnaire if needed. I provided approximately three-hour

training to the student workers. This training included research ethics, study aims and

contributions, questionnaire information, and possible challenges during data collection.

Most questionnaires were self-administered. Student workers read survey questions to a

few participants with vision problems.

This study focused on the two behaviors related to planning for end-of-life care,

but participants were not asked to answer all the questions related to both behaviors. Due

to participants’ age, they might experience fatigue easily when they answered too many

questions. Therefore, I asked each participant to answer questions regarding one of the

two behaviors to avoid compromising reliability of instrumentation. Two versions of the

survey questions related to the TPB and HBM were created. One version included a set of

questions regarding participants’ understandings of discussing end-of-life care plans with

family members, while the other version included questions regarding completing an AD.

The TPB and HBM measures are listed in Appendix II and III. Each participant was

randomly given one of the two versions to complete the survey anonymously. The survey

took them approximately 30 minutes to one hour. In return for their participation, each

participant was given a $10 gift card.

To develop a Chinese-written questionnaire packet, all of the materials including

survey questions, consent letter, and recruitment script were created in English, translated

into Chinese, then back-translated into English by two Chinese scholars with doctoral

degree who can speak both English and Chinese fluently and were not involved in this

dissertation. Further, both Chinese and English-written surveys were pilot tested among

10 Chinese individuals who can speak English fluently to receive feedback about the

survey questions and translations. The pilot study was used to modify the survey

questions before it was administered on a larger scale and excluded for data analysis.

Instrumentation

The independent variables include attitude, subjective norms, perceived

behavioral control, perceived susceptibility, perceived severity, perceived benefits,

perceived barriers, and self-efficacy. Behavioral intentions toward the two behaviors

were the dependent variables. Acculturation and family cohesion were measured to

evaluate whether they would moderate the relationship between TPB and HBM measures

and dependent variables.

Attitude

Attitude was measured through four 5-point Likert items (1=completely disagree,

2=mostly disagree, 3=neither agree nor disagree, 4=mostly agree, and 5=completely

agree). These items were drawn from Fishbein and Ajzen’s (2010) attitude scale (i.e.,

bad-good, harmful-beneficial, useless-useful, worthless-valuable) and modified to fit the

study context. Cronbach’s alpha for attitudes toward discussing end-of-life care plans

with family members was .88. Cronbach’s alpha for attitudes toward completing an AD

was .90. Attitude scores were calculated based on the mean across all the four items for

both behaviors.

Subjective Norms

Subjective norm were measured through three 5-point Likert items (1=completely

disagree, 5=completely agree). These items were drawn from Fishbein and Ajzen’s

(2010) subjective norm scale and modified to fit the study context. Sample items

included: “Most people who are important to me think that I should discuss my end-of-

life care plans with family members/Most people who are important to me think that I

should complete an advance directive.” Cronbach’s alpha for subjective norms toward

discussing end-of-life care plans with family members was .92. Cronbach’s alpha for

subjective norms toward completing an AD was .95. Subjective norm scores were

calculated based on the mean across all the three items for both behaviors.

Self-Efficacy

Self-efficacy was measured through four 5-point Likert items (1=completely

disagree, 5=completely agree). These items were drawn from Witte’s (2000) self-efficacy

scale and modified to fit the study context. Sample items included: “I am confident that I

can discuss my end-of-life care plans with family members/I am confident that I can

complete an advance directive.” Cronbach’s alpha for self-efficacy toward discussing

end-of-life care plans with family members was .88. Cronbach’s alpha for self-efficacy

toward completing an AD was .87. Self-efficacy scores were calculated based on the

mean across all the four items for both behaviors.

Behavioral Intention

Behavioral intention was measured through three 5-point Likert items

(1=completely disagree, 5=completely agree). These items were drawn from Fishbein

and Ajzen’s (2010) behavioral intention scale and modified to fit the study context.

Sample items included: “I intend to discuss my end-of-life care plans with family

members/I intend to complete an advance directive.” Cronbach’s alpha for behavioral

intentions toward discussing end-of-life care plans with family members was .92.

Cronbach’s alpha for behavioral intentions toward completing an AD was .94. Intention

scores were calculated based on the mean across all the three items for both behaviors.

Perceived Susceptibility

Perceived susceptibility toward participants’ individual-level threat, receiving

unwanted medical treatments, was measured through four 5-point Likert items

(1=completely disagree, 5=completely agree). These items were drawn from Witte’s

(2000) susceptibility scale and modified to fit the study context. Sample items included:

“If I do not discuss my end-of-life care plans with family members I will be susceptible

to unwanted medical treatments/If I do not complete an advance directive I will be

susceptible to unwanted medical treatments.” Cronbach’s alpha for perceived

susceptibility toward discussing end-of-life care plans with family members was .92.

Cronbach’s alpha for perceived susceptibility toward completing an AD was .92.

Perceived susceptibility scores were calculated based on the mean across all the four

items for both behaviors.

Perceived Severity

Perceived severity toward participants’ individual-level threat, receiving

unwanted medical treatments, was measured through four 5-point Likert items

(1=completely disagree, 5=completely agree). These items were drawn from Witte’s

(2000) severity scale and modified to fit the study context. Sample items included:

“Receiving unwanted medical treatments would be harmful to me.” Cronbach’s alpha for

perceived severity was .84 for those who answered questions regarding discussing end-

of-life care plans with family members. Cronbach’s alpha for perceived severity was .83

for those who answered questions regarding completing an AD. Perceived severity scores

were calculated based on the mean across all the four items for both groups.

Perceived Benefits

Perceived benefits were measured through six 5-point Likert items (1=completely

disagree, 5=completely agree). These items were adapted from VandeCreek and

Frankowski’s (1996) living will benefits scale and participants’ responses in formative

research. Sample items included: “Discussing end-of-life care plans with family members

will help me get the wanted medical treatments in the future/Completing an advance

directive will help me get the wanted medical treatments in the future.” Cronbach’s alpha

for perceived benefits toward discussing end-of-life care plans with family members

was .92. Cronbach’s alpha for perceived benefits toward completing an AD was .91.

Perceived benefits scores were calculated based on the mean across all the six items for

both behaviors.

Perceived Barriers

Perceived barriers were measured through nine 5-point Likert items

(1=completely disagree, 5=completely agree) were used to measure toward the two

recommended behaviors. These items were adapted from VandeCreek and Frankowski’s

(1996) living will barrier scale and participants’ responses in formative research. Sample

items included: “Discussing end-of-life care plans with family members will cause my

death anxiety/Completing an advance directive will cause my death anxiety.” Cronbach’s

alpha for perceived barriers toward discussing end-of-life care plans with family

members was .91. Cronbach’s alpha for perceived barriers toward completing an AD

was .91. Perceived barriers scores were calculated based on the mean across all the nine

items for both behaviors.

Acculturation

Acculturation was measured with ten 5-point Likert items (1=completely

disagree, 5=completely agree). These items were drawn from Gupta and Yick’s (2001)

acculturation scale validated with Chinese Americans. This acculturation scale was used,

because it covered three different dimensions including language preference, social

customs, and social networks. Sample items included: “I write in English more often than

in Chinese.” A higher score indicated a higher level of acculturation. Cronbach’s alpha

for acculturation was .85 among participants who answered questions regarding

discussing end-of-life care plans with family members. Cronbach’s alpha for

acculturation was .87 among participants who answered questions regarding completing

an AD. Acculturation level scores were calculated based on the mean across all the ten

items for both groups.

Family Cohesion

Family cohesion was measured through five 4-point Likert items (1=strongly

agree, 2=somewhat agree, 3=somewhat disagree, 4=strongly disagree). These items were

drawn from the Alegria et al. (2004) family cohesion scale validated with ethnic minority

groups. Same items included: “Family members respect one another.” All these items

were reversely coded. A higher score indicated a higher level of family cohesion.

Cronbach’s alpha for family cohesion was .88 for participants who answered questions

regarding discussing end-of-life care plans with family members. Cronbach’s alpha for

acculturation was .83 among participants who answered questions regarding completing

an AD. Family cohesion scores were calculated based on the mean across all the five

items for both groups.

Demographic Variables

Participants provided demographic information, including age, gender (0=male,

1=female), education level (1=6th grade or lower, 2=9th grade, 3=12th grade,

4=vocational or trade school, 5=college, 6=postgraduate or higher), religion (0=does not

have a religious belief, 1= have a religious belief), residence length in the U.S. Given that

the previous literature showed the positive influence of older adults’ prior experiences of

ACP engagement on their future care plans (Fried et al., 2009), whether the participant

had discussed end-of-life care plans with family members (0=I had not had discussed

end-of-life care plans with family members, 1=had discussed end-of-life care plans with

family members), and whether the participant had completed an AD (0=I had not had

completed an AD, 1=I had completed an AD) were considered as control variables.

Data Analysis Plan

To examine whether there were significant differences between participants who

answered questions regarding the two recommended behaviors, independent-sample

t-tests and chi-square tests were conducted to compare demographic variables and the

TPB/HBM measures through SPSS 24. Hierarchical regression analyses were performed

to test whether the TPB and HBM can explain and predict participants’ behavioral

intentions. Before regression models were analyzed, all the continuous independent

variables were mean-centered to avoid multicollinearity (Cohen, Cohen, West, & Aiken,

2003). Meanwhile, the associations between all the predictors and the dependent

variables were examined closely with Pearson correlation. To explore the role of

acculturation and family cohesion in explaining and predicting behavioral intentions, they

were included as additional predictors beyond TPB/HBM measures in hierarchical

regression analyses. Their interaction effects with TPB/HBM measures were tested in

SPSS 24 to examine whether acculturation and family cohesion moderated the

relationships between TPB/HBM measures and behavioral intentions.

Given that completing an AD is considered as a one-time behavior by the target

population, participants who reported the AD completion but still answered questions

about their intentions to complete an AD were excluded from the main analysis, because

they may not complete another AD unless an update is needed. In the main analysis, there

were 12 subjects with missing data either on age or residence length in the U.S. For the

variable of age, there was a case with missing value. For the variable of residence length

in the U.S., there were 11 cases with missing values.

Pairwise deletion was used to handle missing data. Pairwise deletion occurs when

subjects relating to each pair of variables with missing data involved in an analysis are

deleted (Bryman, 2004). This approach does not include a case when it has a missing

value on a particular variable, but this case is useable when researchers analyze its other

variables with non-missing values. Pairwise deletion was used to handle missing data in

this dissertation, because it helps mitigate the loss of data by discarding the data for a

case with one or more missing values (Enders, 2010).

CHAPTER 4: RESULTS

Preliminary Statistics

Firstly, given the complexity of perceived benefits and barriers across health

contexts, the measured items were adapted from VandeCreek and Frankowski’s (1996)

living will benefit and barrier scale and participants’ responses in formative research.

An exploratory factor analysis was conducted as a preliminary step to investigate the

factor structures of perceived benefits and barriers for participants who completed this

survey toward the two behaviors respectively.

The Kaiser’s rule of eigenvalues greater than one, the scree plot and parallel

analysis were employed as criteria to determine the maximum number of factors to be

retained (Pett, Lackey, & Sullivan, 2003). O’Connor’s (2000) SPSS program was used to

conduct parallel analysis. Given that perceived benefits and barriers are related to each

other, principal axis factoring with direct oblimin rotation was conducted to identify the

underlying relationships between the items from the existing scale and emerging items

from participants’ answers in the formative research through SPSS 24.

The scree plot and parallel analysis with 95th percentile criterion yielded a three-

factor solution, perceived benefits, perceived barriers 1 (i.e., care planning concern) and

perceived barriers 2 (i.e., future care unpreparedness). All items loaded strongly on one

factor and at least three items loaded substantially on each factor. The factor loadings are

presented in Table 1 and 2. For perceived benefits, all the six items loaded on the

intended factor. For perceived barriers, Item 3, “It is difficult to discuss my end of life

care plans because I do not know what my medical treatment preferences will be in the

future,” Item 6, “I am not used to considering my end of life care plans in advance,” and

Item 7, “I am not sick enough to discuss end of life care plans with family members,”

loaded on one factor focusing on participants’ future care unpreparedness, while other six

items loaded on another factor highlighting participants’ concerns about the potential

negative consequences of planning for end-of-life care in advance.

Table 1

Factor Loadings for Perceived Benefits and Barriers Toward Discussing End-of-Life

Care Plans with Family Members

Note. Extraction method = principal axis factoring; Rotation method = direct oblimin.

Benefit Barrier1 Barrier2

1. Discussing my end of life care plans with family

members will help my family know about my medical

treatment preferences in advance.

.835 -.177 .168

2. Discussing my end of life care plans with family

members will help my doctor know about my medical

treatment references in advance.

.829 -.151 .057

3. Discussing my end of life care plans with family

members will help me get the wanted medical

treatments in the future.

.868 -.149 .136

4. Discussing my end of life care plans with family

members will help me relieve family burdens.

.801 .174 -.122

5. Discussing my end of life care plans with family

members will help me reduce family conflicts.

.743 .217 -.311

6. Discussing my end of life care plans with family

members will increase the quality of my life in my last

days.

.802 .000 .004

1. It makes me sad to discuss my end of life care plans

with family members.

.038 .781 .070

2. Discussing my end of life care plans with family

members will increase my family conflicts.

-.096 .745 -.113

4. I feel uncomfortable to discuss my end of life care

plans.

-.026 .730 .228

5. Discussing my end of life care plans with family

members will cause my death anxiety.

.006 .769 .073

8. Discussing my end of life care plans with my family

members is bad luck.

-.026 .660 .136

9. It will make my family members sad if I discuss my

end of life care plans with them.

.025 .715 .072

3. It is difficult to discuss my end of life care plans

because I do not know what my medical treatment

preferences will be in the future.

-.037 .333 .571

6. I am not used to considering my end of life care plans

in advance.

.035 .295 .730

7. I am not sick enough to discuss end of life care plans

with family members.

-.027 .082 .623

Table 2

Factor Loadings for Perceived Benefits and Barriers Toward Completing an Advance

Directive

Note. Extraction method = principal axis factoring; Rotation method = direct oblimin.

Benefit Barrier

Barrier

1. Completing an advance directive will help my family know

about my medical treatment preferences in advance.

.781 -.281 -.301

2. Completing an advance directive will help my doctor know

about my medical treatment references in advance.

.774 -.235 -.288

3. Completing an advance directive will help me get the wanted

medical treatments in the future.

.811 -.171 -.126

4. Completing an advance directive will help me relieve family

burdens.

.886 .102 .122

5. Completing an advance directive will help me reduce family

conflicts.

.766 .122 .201

6. Completing an advance directive will increase the quality of

my life in my last days.

.730 .109 .114

1. It makes me sad to complete an advance directive. .005 .768 -.093

2. Completing an advance directive will increase my family

conflicts.

-.146 .578 -.079

4. I feel uncomfortable to complete an advance directive. .020 .841 -.100

5. Completing an advance directive will cause my death anxiety. -.012 .941 .062

8. Completing an advance directive is bad luck. -.083 .723 -.029

9. It will make my family members sad if I complete an advance

directive.

.099 .518 -.049

3. It is difficult to complete an advance directive because I do

not know what my medical treatment preferences will be in the

future.

-.029 .207 -.612

6. I am not used to considering my end of life care plans in

advance.

-.084 .467 -.515

7. I am not sick enough to complete an advance directive. -.032 .214 -.629

Demographic Characteristics

A total of 319 participants living in the metropolitan Phoenix area participated in

this study. As described in the previous chapter, two versions of questionnaires were

randomly delivered to participants. Based on the random assignment, 161 participants

answered questions regarding discussing end-of-life care plans with family members,

while 158 answered another set of questions regarding completing an advance directive

(AD). Independent sample t-tests and chi-square tests were conducted to compare

participants’ age, residence length in the U.S., gender, religion, whether they had

discussed end-of-life care plans with family members or not, and whether they had

completed an AD. There were no significant differences between the groups except

education. The effect of education was controlled for when multiple regression analyses

were performed. All the participants’ demographic information is presented in Table 3.

To examine intentions to plan for end-of-life care, participants who reported that

they had already completed an AD before the survey and answered questions regarding

completing an AD in the survey (n = 21) were excluded from the main analysis because

they may not complete another AD after completing an AD unless an update is needed.

Table 3

Demographic Characteristics of Participants

Group 1

(N = 161)

Group 2-In

(N = 137)

Group 2-Out

(N = 21)

Group 1 vs

Group 2-In

Group 2-In vs

Group 2-Out

M(SD)/

n(%)

M(SD)/

n(%)

M(SD)/

n(%)

t-value(df)/

χ2(df)

t-value(df)/

χ2(df)

Age 73.04(9.16) 74.40(7.87) 76.62(9.80) -1.36(294) -1.16(156)

Residence in

the U.S.

24.19(19.00) 20.10(18.62) 51.71(26.29) 1.86(292) -.5.31(23)**

Gender .85(1) .003(1)

Male 54(33.5%) 53(38.7%) 8(38.1%)

Female 107(66.5%) 84(61.3%) 13(61.9%)

Religion 2.09(1) .87(1)

Yes 102(64.2%) 76(55.9%) 14(66.7%)

No 57(35.8%) 60(44.1%) 7(33.3%)

Education 10.07(5) 5.11(5)

6th grade or

lower

23(14.3%) 18(13.1%) 2(9.5%)

9th grade 23(14.3%) 22(16.1%) 0(.0%)

12th grade 23(14.3%) 25(18.2%) 5(23.8%)

Vocational or

trade school

35(21.7%) 22(16.1%) 3(14.3%)

College 31(19.3%) 40(29.2%) 9(42.9%)

Postgraduate or

higher

26(16.1%) 10(7.3%) 2(9.5%)

I had

discussed end-

of-life care

with family

members

7.16(1)* 44.19(1)**

Yes 54(33.5%) 27(19.7%) 19(90.5%)

No 107(66.5%) 110(80.3%) 2(9.5%)

I had

completed an

Yes 26(16.1%) 21(13.3%) 21(100.0%)

No 135(83.9%) 137(85.3%) 0(.0%)

Note. Group 1 = Participants who answered questions regarding discussing end-of-life care plans

with family members. Group 2-In = Participants who did not complete an advance directive

before the survey and answered questions regarding completing an advance directive. Group 2-

Out = Participants who reported completing an advance directive before the survey and answered

questions regarding completing an advance directive. ** = p < .001. * = p < .05.

The final sample (N = 298) included 107 males and 191 females. The age of

participants ranged from 55 to 101 years (M = 73.67, SD = 8.60). There were 267

participants who completed the survey in Chinese (i.e., Mandarin or Cantonese), while 31

completed the survey in English. There were 178 participants who reported that they had

a religious belief, while 117 reported that they did not have one and three did not report

whether they had religious beliefs. In terms of participants’ education levels, there were

41 participants with 6th grade or lower, 45 with 9th grade, 48 with 12th grade, 57 with

vocational or trade school certificate, 71 with college degrees, and 36 with graduate

degrees. Participants’ residence length (i.e., how long they have lived in the U.S.) ranged

from less than one year to 89 years (M = 22.30, SD = 18.90). Only 18 participants were

U.S. born. Most of the participants were foreign-born in mainland China (n = 200),

Taiwan (n = 48), Hong Kong (n = 11), and other areas outside the U.S. (n = 21).

Among the 298 participants, 161 participants answered questions regarding

discussing end-of-life care plans with family members and 137 answered questions

regarding completing an AD. Independent sample t-tests and chi-square tests were

conducted to compare the two groups on demographic variables. The results are

presented in Table 3. It is found that there were no significant differences in gender, age,

religion, education levels, and residence in the U.S. between participants who completed

the survey regarding discussing end-of-life care plans with family members and others

who completed the survey regarding completing an AD. However, there were significant

differences in terms of whether participants had discussed end-of-life care plans with

family members or not, and whether participants had completed an AD. Their effects

were controlled for when multiple regression analyses were performed.

Theory of Planned Behavior Results

RQ1a-b: Descriptive Statistics

RQ1a-b asked what participants’ attitudes, subjective norms, self-efficacy, and

behavioral intentions are. Table 4 shows means, standard deviations of TPB measures

toward the two recommended behaviors and t-test results. Given that participants rated

these measures near the middle of the five-point scales, they appeared to have neutral

attitudes, subjective norms, self-efficacy, and intentions toward the two behaviors.

RQ2: Comparing Theory of Planned Behavior Measures

RQ2 asked whether there were group differences in participants’ attitudes,

subjective norms, self-efficacy, and behavioral intentions toward the two behaviors. The

t-test results of comparing group differences are presented in Table 4. There were group

differences in terms of subjective norms (t(296) = 3.02, p < .05), self-efficacy (t(296) =

2.28, p < .05), and behavioral intentions (t(296) = 2.52, p < .001). Participants scored

significantly higher in subjective norms, self-efficacy, and behavioral intentions toward

discussing end-of-life care plans with family members than completing an AD.

Table 4

t-test Results of TPB and HBM Measures Comparison

Group 1

(N = 161)

Group 2

(N = 137)

t-test

M(SD) M(SD) t-value df p-value

1. Attitude 3.73(.69) 3.57(.75) 1.87 296 .06

2. Subjective norms 3.79(1.04) 3.42(1.08) 3.02 296 .003*

3. Self-efficacy 3.90(.95) 3.64(1.00) 2.28 296 .02*

4. Perceived

susceptibility

3.23(1.18) 3.27(1.17) -.29 296 .78

5. Perceived severity 3.48(1.08) 3.52(1.03) -.30 296 .77

6. Perceived benefits 4.19 (.92) 4.06(.92) 1.20 296 .23

7. Perceived barriers 2.31(1.08) 2.49(1.07) -1.43 296 .16

8. Intention 3.95(1.14) 3.60(1.25) 2.52 296 .01*

9. Acculturation 2.22(.79) 2.05(.81) 1.81 293 .07

10. Family cohesion 3.57(.46) 3.63(.44) -1.18 296 .24

Note. Group 1=Participants who answered questions regarding discussing end-of-life care plans

with family members. Group 2=Participants who answered questions regarding completing an

advance directive. Variables 1–10 measured on a 5-point scale. Variable 11 measured on a 4-

point scale. ** = p < .001, * = p < .05.

H1a: Discussing End-of-Life Care Plans with Family Members

H1a predicted that participants’ attitudes, subjective norms, and self-efficacy

would positively predict behavioral intentions toward discussing end-of-life care plans

with family members positively. RQ3a and RQ4a asked whether acculturation and family

cohesion would separately moderate the relationships between TPB measures and

behavioral intentions toward discussing end-of-life care plans with family members.

Table 5 shows means, standard deviations, and zero-order correlations among attitudes,

subjective norms, self-efficacy, and intentions to discuss end-of-life care plans with

family members.

Two hierarchical multiple regression analyses were performed with behavioral

intentions toward discussing end-of-life care plans as the dependent variable, TPB

measures as the independent variables, and acculturation or family cohesion as a

moderator. Block 1 contained control variables, including age, residence length in the

U.S., gender, religion, education levels, whether the participant had discussed end-of-life

care plans with family members, and whether the participant had completed an AD.

Block 2 included attitudes, subjective norms, and self-efficacy as the predictors. Block 3

added acculturation/family cohesion as an additional predictor. Block 4 added the

interaction effects between acculturation/family cohesion and TPB measures. The results

of these regression analyses are presented in Table 6 and 7.

T a b le

D e sc

ri p ti

v e I

n fo

rm a ti

o n a

n d Z

e ro

-o rd

e r

C o rr

e la

ti o n C

o e ff

ic ie

n ts

T o w

a rd

D is

c u ss

in g E

n d

-o f-

L if

e C

a re

P la

n s

w it

h F

a m

il y

M e m

b e rs

( N

= 1

6 1

)

V a ri

a b

M (

S D

) 1

1 . A

tt it

u d

3 .7

3 (.

6 9 )

–

2 . S

u b je

c ti

v e n

o rm

s 3 .7

9 (1

.0 4

) .5

9 * *

–

3 . S

e lf

-e ff

ic a c y

3 .9

0 (.

9 5 )

.6 1 * *

.6 3 * *

–

4 . P

e rc

e iv

e d

su sc

e p

ti b

il it

3 .2

3 (1

.1 8 )

.3 2 * *

.3 9 * *

.2 4 *

–

5 . P

e rc

e iv

e d

s e v e ri

3 .4

8 (1

.0 8 )

.2 2 * *

.2 7 * *

.1 5

0 * *

–

6 . P

e rc

e iv

e d

b e n

e fi

4 .1

9 (

.9 2 )

.3 8 * *

.3 4 * *

.3 8 * *

.3 4 * *

.3 8 * *

–

7 . P

e rc

e iv

e d

b a rr

ie rs

.3 1 (1

.0 8 )

-. 4 0 * *

3 2 * *

4 1 * *

0 5

-. 2

0 *

–

8 . In

te n

ti o

3 .9

5 (1

.1 4 )

.6 4 * *

.6 5 * *

.6 6 * *

.3 1 * *

.2 6 *

.5 2

* *

-. 4

6 *

–

9 . A

c c u

lt u ra

ti o

2 .2

2 (.

7 9 )

.1 2

.1 7 *

.0 4

0 4

-. 0

.2 0

–

1 0

. F

a m

il y c

o h

e si

o n

3 .5

7 (.

4 6 )

.0 4

-. 0 2

0 3

.0 9

-. 1

7 *

-. 0

N o

te . V

a ri

a b

le s

1 – 1 0

m e a su

re d o

n a

5 -p

o in

t sc

a le

. V

a ri

a b le

1 1 m

e a su

re d o

n a

4 -p

o in

t sc

a le

. *

* =

p <

. 0

0 1

. *

= p

< .

0 5

T a b le

R e su

lt s

o f

R e g

re ss

io n

A n

a ly

si s

fo r

T P

B T

o w

a rd

D is

c u ss

in g E

n d

-o f-

L if

e C

a re

P la

n s

(A c c u lt

u ra

ti o

n a

s a

M o

d e ra

to r)

V a ri

a b

le s

M o d e l

M o

d e l

M o

d e l

(S E

) β

B (S

E )

sr 2

B (S

E )

sr 2

B (S

E )

A g e

-. 0

1 (.

0 1 )

-. 1 0

-. 0 0 2 (.

0 1 )

-. 0 1

.0 0 (.

0 1 )

-. 0

0 4

.0 0

(. 0

1 )

-. 0

0 2

R e si

d e n

c e l

e n

g th

0 0

1 (.

0 1 )

-. 0 2

.0 0 1 (.

0 0 4 )

.0 2

.0 0

-. 0 0 1 (.

0 1 )

-. 0

.0 0

-. 0

0 1

(. 0

1 )

-. 0

.0 0

G e n

d e r

-. 3

2 (.

1 9 )

-. 1 3

-. 1 5 (.

1 3 )

-. 0 6

0 3

-. 1 7 (.

1 4 )

-. 0

.0 0

-. 1

6 (

.1 4

) -.

0 7

0 3

R e li

g io

.5 3

(. 1 8 )

.2 3 *

.0 5

.2 8 (.

1 3 )

.1 2 *

.0 1

.2 9 (.

1 3 )

.1 2

.0 1

8 (.

1 3 )

.1 2

.0 1

E d

u c a ti

o n

.0 7

(. 0 6 )

.1 1

.0 4 (.

0 4 )

.0 6

.0 0 3

.0 4 (.

0 4 )

.0 5

0 2

.0 4

(. 0

4 )

.0 5

.0 0

W h

e th

e r

I h

a d

d is

c u ss

e d

e n

d -o

f- li

c a re

w it

h f

a m

il y

m e m

b e rs

.8 9

(. 2 0 )

.3 7 *

.1 0

.4 0 (.

1 5 )

.1 7 *

.0 2

.4 1 (.

1 5 )

.1 7

.0 2

2 (.

1 5 )

.1 8

.0 2

W h

e th

e r

I h

a d

c o

m p

le te

d a

n A

-. 2

8 (.

2 8 )

-. 0 9

-. 3 6 (.

2 0 )

-. 1 2

-. 3 6 (.

2 0 )

-. 1

.0 1

3 3

(. 2

1 )

-. 1

.0 1

A tt

it u

d e

2 (.

1 1 )

.2 6 * *

.0 4

.4 2 (.

1 1 )

.2 6

* *

.0 4

3 (.

1 2 )

.2 6

* *

.0 4

S u

b je

c ti

v e n

o rm

(S B

N )

6 (.

0 8 )

.2 4 *

.0 3

.2 7 (.

0 8 )

.2 4

.0 3

5 (.

0 8 )

.2 3

.0 2

S e lf

-e ff

ic a c y (

S E

)

.3 6 (.

0 9 )

.3 0 * *

.0 4

5 (.

0 9 )

.2 9

* *

.0 4

.3 3

(. 0

9 )

.2 8

* *

.0 3

A c c u

lt u

ra ti

o n

.0 7 (.

1 1 )

.0 5

0 1

.0 7

(. 1

2 )

.0 5

.0 0 1

A tt

it u

d e *

A c c u lt

u ra

ti o

8 (.

1 3 )

.0 4

.0 0 1

S B

N *

A c c u

lt u

ra ti

o n

0 6

(. 1

1 )

-. 0

.0 0 1

S E

* A

c c u

lt u

ra ti

o n

0 4

(. 1

3 )

-. 0

.0 0

Δ R

.2 3

.3 9

.0 0

R 2 (A

d ju

st e d

R 2 )

.2 3 (.

2 0 )

.6 2 (.

5 9 )

.6 2

(. 5

9 )

.6 2

(. 5

9 )

F -c

h a

n g

6 .4

1 * *

4 9 .3

4 * *

.3 5

.2 1

N o te

. *

* =

p <

. 0

0 1

. *

= p

< . 0

5 .

T a b le

R e su

lt s

o f

R e g

re ss

io n

A n

a ly

si s

fo r

T P

B T

o w

a rd

D is

c u ss

in g E

n d

-o f-

L if

e C

a re

P la

n s

(F a m

il y C

o h

e si

o n

a s

a M

o d

e ra

to r)

V a ri

a b

le s

M o d e l

M o

d e l

(S E

) β

B (S

E )

sr 2

B (S

E )

sr 2

B (S

E )

A g e

-. 0

1 (.

0 1

) -.

1 0

-. 0 0 2 (.

0 1 )

-. 0 1

-. 0 0 2 (.

0 1 )

-. 0

.0 0

0 0

3 (.

0 1

) -.

0 3

R e si

d e n

c e l

e n

g th

0 0

1 (.

0 1

) -.

0 2

.0 0 1 (.

0 0 4 )

.0 2

.0 0 1 (.

0 0 4 )

-. 0

.0 0

0 (.

0 0

4 )

.0 0

G e n

d e r

-. 3

2 (.

1 9

) -.

1 3

-. 1 5 (.

1 3 )

-. 0 6

0 3

-. 1 5 (.

1 3 )

-. 0

.0 0

-. 1

5 (.

1 3

) -.

0 6

0 3

R e li

g io

.5 3

(. 1

8 )

.2 3 *

.0 5

.2 8 (.

1 3 )

.1 2

.0 1

8 (.

1 3 )

.1 2

.0 1

9 (.

1 3 )

.1 2

.0 1

E d u c a ti

o n

.0 7 (.

0 6 )

.1 1

.0 1

.0 4

(. 0

4 )

.0 6

.0 0 3

.0 4

(. 0

4 )

.0 6

.0 0 3

.0 4 (.

0 4 )

.0 6

.0 0 3

W h

e th

e r

I h

a d

d is

c u ss

e d

e n

d -o

li fe

c a re

w it

fa m

il y m

e m

b e rs

.8 9

(. 2

0 )

.3 7 *

.1 0

.4 0 (.

1 5 )

.1 7 *

.0 2

0 (.

1 5 )

.1 7

.0 2

6 (.

1 5 )

.1 9

.0 2

W h

e th

e r

I h

a d

c o

m p

le te

d a

n A

-. 2

8 (.

2 8

) -.

0 9

-. 3 6 (.

2 0 )

-. 1 2

-. 3 6 (.

2 0 )

-. 1

.0 1

3 6

(. 2

0 )

-. 1

.0 1

A tt

it u

d e

2 (.

1 1 )

.2 6 * *

.0 4

2 (.

1 1 )

.2 6

* *

.0 4

1 (.

1 2 )

.2 5

.0 3

S u

b je

c ti

v e n

o rm

(S B

N )

6 (.

0 8 )

.2 4 *

.0 3

6 (.

0 8 )

.2 4

.0 3

7 (.

0 8 )

.2 4

.0 3

S e lf

-e ff

ic a c y

(S E

)

6 (.

0 9 )

.3 0 * *

.0 4

.3 6

(. 0

9 )

.3 0 * *

.0 4

.3 4 (.

0 9 )

.2 8 * *

.0 4

F a m

il y c

o h

e si

o n

(F C

)

.0 0 (.

1 3 )

.0 0

0 3

(. 1

3 )

-. 0

.0 0

A tt

it u

d e *

F C

-. 1

6 (.

2 6

) -.

0 4

0 1

S B

N *

F C

.3 5

(. 1

8 )

.1 3

.0 1

S E

* F

2 3

(. 2

1 )

-. 0

.0 0

Δ R

.2 3

.3 9

.0 0

.0 1

R 2 (A

d ju

st e d

R 2 )

.2 3 (.

2 0 )

.6 2 (.

5 9 )

.6 2 (.

5 9 )

.6 3

(. 6

0 )

F -c

h a

n g

6 .4

1 * *

4 9 .3

4 * *

.0 0

1 .5

N o

te . * * =

p <

. 0 0 1 .

* =

p <

. 0 5 .

At Block 1, the seven control variables explained a significant amount of

variation, R2-change = .23, F-change(7, 148) = 6.41, p < .001. Religion and whether the

participant had discussed end-of-life care plans with family members were the significant

positive predictors for behavioral intentions. That is, having a religious belief and having

discussed end-of-life care plans with family members can help explain participants’

intentions toward discussing end-life-life care plans with family members.

At Block 2, the inclusion of attitude, subjective norms, and self-efficacy

accounted for a substantial amount of variation in intentions beyond control variables,

R2-change = .39, F-change(3,145) = 49.34, p < .001. The proportion of variation in

behavioral intentions explained by the set of predictors was statistically significant,

R2 = .62, adjusted R2 = .59, F(10, 145) = 23.69 , p < .001. Attitudes (β = .26, sr2 = .04,

p < .001), subjective norms (β = .24, sr2 = .03, p < .05), self-efficacy (β = .30, sr2 = .04,

p < .001), religion (β = -.06, sr2 = .01, p < .05), and whether the participant had discussed

end-of-life care plans with family members (β = .17, sr2 = .02, p < .05) were significant

predictors for behavioral intentions. H1a was supported.

RQ3a: Acculturation as a Moderator. At Block 3, acculturation was added as

an additional predictor and did not increase a statistically proportion of variation in

behavioral intentions. At Block 4, all the three interaction effects between TPB measures

and acculturation were included as predictors together. They did not increase a

statistically proportion of variation in behavioral intentions either. The results are

presented in Table 6. Acculturation did not predict participants’ intentions to discuss end-

of-life care plans with family members after the inclusion of attitudes, subjective norms,

and self-efficacy in the regression analysis. Acculturation did not moderate the

relationships between TPB measures and behavioral intentions.

RQ4a: Family Cohesion as a Moderator. Another hierarchical regression was

conducted with behavioral intentions toward discussing end-of-life care plans with family

members as the dependent variable, TPB measures as the independent variables, and

family cohesion as a moderator. The results are presented in Table 7. Family cohesion did

not improve the prediction of behavioral intentions beyond TPB measures. The

interaction effects between family cohesion and TPB measures did not increase a

significant amount of variation in behavioral intentions. Family cohesion did not

moderate the relationships between TPB measures and intentions to discuss end-of-life

care plans with family members.

H1b: Completing an Advance Directive

H1b predicted that participants’ attitudes, subjective norms, and self-efficacy

would positively predict behavioral intentions toward completing an AD. RQ3b and

RQ4b asked whether acculturation and family cohesion would separately moderate the

relationships between TPB measures and behavioral intentions toward completing an

AD. Table 8 shows means, standard deviations, and zero-order correlations among

attitudes, subjective norms, self-efficacy, and intentions to complete an AD.

T a b le

D e sc

ri p ti

v e I

n fo

rm a ti

o n a

n d Z

e ro

-o rd

e r

C o rr

e la

ti o n C

o e ff

ic ie

n ts

T o w

a rd

C o m

p le

ti n g a

n A

d v a n c e D

ir e c ti

v e (

N =

1 3 7

)

V a ri

a b

M (

S D

) 1

1 . A

tt it

u d

3 .5

7 (.

7 5 )

–

2 . S

u b je

c ti

v e

n o

rm s

3 .4

1 (1

.0 8

) .5

2 * *

–

3 . S

e lf

-e ff

ic a c y

3 .6

4 (1

.0 0

) .4

2 * *

.5 0 * *

–

4 .

P e rc

e iv

e d

su sc

e p

ti b

il it

3 .2

7 (1

.1 7 )

.3 5 * *

.4 8 * *

.3 0 * *

–

5 . P

e rc

e iv

e d

se v e ri

3 .5

2 (1

.0 3

) .1

9 *

.4 0 * *

.5 1 * *

–

6 . P

e rc

e iv

e d

b e n

e fi

4 .0

6 (.

9 2 )

.5 3 * *

.5 8 * *

.5 2 * *

.4 3 * *

.4 4 * *

–

7 . P

e rc

e iv

e d

b a rr

ie rs

2 .4

9 (1

.0 7

) -.

3 7 * *

2 4 *

2 2 *

1 5

1 0

2 7 *

–

8 . In

te n

ti o

3 .6

0 (1

.2 5

) .6

2 * *

.6 5 * *

.5 4 * *

.4 8 * *

.3 1 * *

.6 6 * *

-. 3

7 *

–

9 . A

c c u

lt u ra

ti o

2 .0

5 (.

8 1 )

.1 7

.0 2

.0 6

.0 4

-. 0 9

0 1

1 3

–

1 0

. F

a m

il y

c o

h e si

o n

3 .6

3 (.

4 4 )

.2 8 *

.2 5 *

.1 3

.0 9

4 *

-. 2

4 *

9 *

-. 1

N o

te :

V a ri

a b le

s 1 –

1 0

m e a su

re d o

n a

5 -p

o in

t sc

a le

. V

a ri

a b le

1 1 m

e a su

re d o

n a

4 -p

o in

t sc

a le

. *

* =

p <

. 0

0 1

. *

= p

< .

0 5

Two hierarchical multiple regression analyses were performed with behavioral

intentions toward completing an AD as the dependent variable, TPB measures as the

independent variables, and acculturation or family cohesion as a moderator. Block 1

contained control variables, including age, residence length in the U.S., gender, religion,

education levels, and whether the participant had discussed end-of-life care plans with

family members). Block 2 included attitudes, subjective norms, and self-efficacy as the

predictors. Block 3 added acculturation/family cohesion as an additional predictor. Block

4 added the interaction effects between acculturation/family cohesion and TPB measures.

The results of these regression analyses are presented in Table 9 and 10.

At Block 1, the six control variables explained a significant amount of variation,

R2-change = .19, F-change(6, 128) = 4.95, p < .001. Education levels and whether the

participant had discussed end-of-life care plans with family members were the significant

positive predictors for behavioral intentions, while age and residence length in the U.S.

worked as significant negative predictors for behavioral intentions.

At Block 2, attitudes, subjective norms, and self-efficacy accounted for a

substantial amount of variation in behavioral intentions beyond control variables,

R2-change = .42, F-change(3,125) = 45.3, p < .001. The proportion of variation in

behavioral intentions explained by the set of predictors was statistically significant,

R2 = .61, adjusted R2 = .58, F(9, 125) = 21.82 , p < .001. Attitudes (β = .31, sr2 = .07,

p < .001), subjective norms (β = .37, sr2 = .08, p < .001), self-efficacy (β = .19, sr2 = .02,

p < .05), and whether the participant had discussed end-of-life care plans with family

members (β = .15, sr2 = .02, p < .05) were significant predictors for behavioral intentions.

H1b was supported.

T a b le

R e su

lt s

o f

R e g

re ss

io n

A n

a ly

si s

fo r

T P

B T

o w

a rd

C o m

p le

ti n g a

n A

d v a n c e D

ir e c ti

v e (

A c c u lt

u ra

ti o

n a

s a

M o d

e ra

to r)

V a ri

a b

le s

M o d e l

M o

d e l

M o

d e l

(S E

) β

B (S

E )

sr 2

B (S

E )

sr 2

B (S

E )

A g e

-. 0

3 (.

0 1 )

-. 1 7 *

-. 0 1 (.

0 1 )

-. 0 8

-. 0 1 (.

0 1

) -.

0 8

-. 0

1 (.

0 1

) -.

0 9

R e si

d e n

c e l

e n

g th

0 1

(. 0 1 )

-. 1 9 *

-. 0 0 2 (.

0 0 4 )

-. 0 4

0 1

-. 0 0 3 (.

0 1

) -.

0 5

0 1

-. 0

0 3

(. 0

1 )

-. 0

.0 0 1

G e n

d e r

-. 1

7 (

.2 2 )

-. 0 7

0 4

.1 8 (.

1 6 )

.0 7

.0 0 4

.1 8 (.

1 6 )

.0 7

.0 0

.1 6

(. 1

6 )

.0 6

0 3

R e li

g io

.3 8

(. 2

1 )

.1 5

.0 2

5 (.

1 5 )

.0 6

.0 0 3

.1 5 (.

1 5 )

.0 6

.0 0

.1 5

(. 1

5 )

.0 6

0 3

E d

u c a ti

o n

.1 5

(. 0

7 )

.1 9 *

.0 3

7 (.

0 5 )

.0 8

.0 1

6 (.

0 5 )

.0 8

.0 0

.0 7

(. 0

5 )

.0 8

0 5

W h

e th

e r

I h

a d

d is

c u ss

e d

e n

d -o

f- li

c a re

w it

h f

a m

il y

m e m

b e rs

.8 9

(. 2

5 )

.2 8 *

.0 8

8 (.

1 8 )

.1 5 *

.0 2

8 (.

1 8 )

.1 5

.0 2

9 (.

1 8 )

.1 6

.0 2

A tt

it u

d e

2 (.

1 1 )

.3 1 * *

.0 7

2 (.

1 2 )

.3 1

* *

.0 6

.3 9

(. 1

2 )

.2 4

.0 3

S u

b je

c ti

v e n

o rm

(S B

N )

2 (.

0 8 )

.3 7 * *

.0 8

2 (.

0 8 )

.3 7

* *

.0 8

2 (.

0 8 )

.3 7

* *

.0 8

S e lf

-e ff

ic a c y (

S E

)

.2 3 (.

0 9 )

.1 9 *

.0 2

3 (.

0 9 )

.1 9

.0 2

6 (.

0 9 )

.2 1

.0 2

A c c u

lt u

ra ti

o n

.0 3 (.

1 5 )

.0 2

.0 0

.0 5

(. 1

5 )

.0 3

A tt

it u

d e *

A c c u lt

u ra

ti o

4 3

(. 1

5 )

-. 2

1 *

S B

N *

A c c u

lt u

ra ti

o n

3 (.

1 0 )

.0 2

S E

* A

c c u

lt u

ra ti

o n

4 (.

1 1 )

.0 3

Δ R

.1 9

.0 0

.0 3

R 2 (A

d ju

st e d

R 2 )

.1 9 (.

1 5 )

.6 1 (.

5 8 )

.6 1

(. 5

8 )

.6 4

(. 6

1 )

F -c

h a

n g

4 .9

5 * *

4 5 .3

0 * *

.0 4

3 .5

9 *

N o te

. *

* =

p <

. 0

0 1

. *

= p

< . 0

5 .

T a b le

1 0

R e su

lt s

o f

R e g

re ss

io n

A n

a ly

si s

fo r

T P

B T

o w

a rd

C o m

p le

ti n g a

n A

d v a n c e D

ir e c ti

v e (

F a m

il y C

o h e si

o n

a s

a M

o d

e ra

to r)

V a ri

a b

le s

M o d e l

M o

d e l

M o

d e l

(S E

) β

B (S

E )

sr 2

B (S

E )

sr 2

B (S

E )

A g e

-. 0

3 (.

0 1

) -.

1 7 *

-. 0 1 (.

0 1 )

-. 0 8

-. 0 1 (.

0 1 )

-. 0

.0 1

-. 0

1 (.

0 1

) -.

0 7

R e si

d e n

c e l

e n

g th

0 1

(. 0

1 )

-. 1 9 *

-. 0 0 2 (.

0 0 4 )

-. 0 4

0 1

-. 0 0 3 (.

0 0 4 )

-. 0

.0 0 1

-. 0

0 1

(. 0

0 4

) -.

0 2

G e n

d e r

-. 1

(. 2

2 )

-. 0 7

0 4

.1 8 (.

1 6 )

.0 7

.0 0 4

.1 9 (.

1 6 )

.0 7

.0 0

.1 6

(. 1

6 )

.0 6

.0 0

R e li

g io

.3 8 (.

2 1 )

.1 5

.0 2

.1 5

(. 1

5 )

.0 6

.0 0 3

.1 4

(. 1

5 )

.0 6

.0 0 3

.0 6 (.

1 5 )

.0 3

.0 0 1

E d

u c a ti

o n

.1 5

(. 0

7 )

.1 9 *

.0 3

7 (.

0 5 )

.0 8

.0 1

.0 7 (.

0 5 )

.0 9

.0 1

.0 5

(. 0

5 )

.0 6

.0 0

W h

e th

e r

I h

a d

d is

c u ss

e d

e n

d -

o f-

li fe

c a re

w it

fa m

il y m

e m

b e rs

.8 9

(. 2

5 )

.2 8 *

.0 8

8 (.

1 8 )

.1 5 *

.0 2

.4 9 (.

1 8 )

.1 6

.0 2

.4 7

(. 1

8 )

.1 5

.0 2

A tt

it u

d e

2 (.

1 1 )

.3 1 * *

.0 7

.5 3 (.

1 2 )

.3 2

* *

.0 7

.5 9

(. 1

2 )

.3 5

* *

.0 8

S u

b je

c ti

v e n

o rm

(S B

N )

2 (.

0 8 )

.3 7 * *

.0 8

.4 3 (.

0 8 )

.3 7

* *

.0 8

.3 9

(. 0

8 )

.3 4

* *

.0 7

S e lf

-e ff

ic a c y

(S E

)

3 (.

0 9 )

.1 9 *

.0 2

.2 3 (.

0 9 )

.1 8

.0 2

.2 8

(. 0

9 )

.2 2

.0 3

F a m

il y c

o h

e si

o n

(F C

)

-. 0 7 (.

1 8 )

-. 0

.0 0

-. 0

4 (.

1 8

) -.

0 2

A tt

it u

d e *

F C

.3 4

(. 2

3 )

.1 2

.0 1

S B

N *

F C

-. 3

3 (.

1 9

) -.

1 4

S E

* F

7 (.

2 1 )

.1 0

.0 1

Δ R

.1 9

.0 0

R 2 (A

d ju

st e d

R 2 )

.1 9 (.

1 5 )

.6 1 (.

5 8 )

.6 1

(. 5

8 )

.6 3

(. 5

9 )

F -c

h a

n g

4 .9

5 * *

4 5 .3

0 * *

.1 5

.1 6

N o te

. * * =

p <

. 0 0 1 .

* =

p <

. 0

5 .

RQ3b: Acculturation as a Moderator. At Block 3 acculturation was added as an

additional predictor and did not increase a statistically proportion of variation in

behavioral intentions. At Block 4 all the three interaction effects between TPB measures

and acculturation were included as predictors together. They increased a statistically

proportion of variation in behavioral intentions, R2-change = .03, F-change(3, 121) =

3.59, p < .05. The interaction between attitudes and acculturation was a significant

predictor for behavioral intentions, β = -.21, sr2 = .03, p < .05. As acculturation increased,

attitudes had a smaller influence on intentions to complete an AD. Furthermore, adding

the interaction effect between attitudes and acculturation reduced the size of the

significant effects of attitudes on intentions. The results are presented in Table 9.

The significant interaction effect was probed with simple slopes analysis. Results

of the simple slopes analysis is presented in Figure 3. “High,” “moderate,” and “low”

levels of responses to participants’ acculturation were created by computing one standard

deviation above and one standard deviation below the mean. This analysis indicated that

the relationship between attitudes and behavioral intentions was larger when participants

were less acculturated.

Figure 3. Acculturation as a Moderator between Attitude and Behavioral Intention toward

Completing an Advance Directive.

RQ4b: Family Cohesion as a Moderator. Another hierarchical regression was

conducted with behavioral intentions toward completing an AD as the dependent variable

and TPB measures as the independent variables and family cohesion as a moderator. The

results are presented in Table 10. Family cohesion did not improve the proportion of

variation in behavioral intentions. The interaction effects between family cohesion and

TPB measures did not increase a significant amount of variation in behavioral intentions.

Family cohesion did not moderate the relationships between TPB measures and

intentions to complete an AD.

0.5

1.5

2.5

3.5

4.5

Low Moderate High

In te

n ti

o n

Attitude

Low

Moderate

High

Health Belief Model Results

RQ5a-b: Descriptive Statistics

RQ5a-b asked what participants’ perceived susceptibility, perceived severity,

perceived benefits, and perceived barriers are. Table 4 shows means, standard deviations

of HBM measures toward the two recommended behaviors and t-test results. The results

indicated that participants tended to score neutral in perceived susceptibility and severity

and high in perceived benefits and low in perceived barriers.

RQ6: Comparing Health Belief Model Measures

RQ6 asked whether there were group differences in participants’ perceived

susceptibility, perceived severity, perceived benefits, and perceived barriers toward the

two behaviors. The results are presented in Table 4. There were no significant group

differences in terms of HBM measures.

H2a: Discussing End-of-Life Care Plans with Family Members

H2a predicted that participants’ perceived susceptibility, perceived severity,

perceived barriers, perceived benefits, and self-efficacy would predict behavioral

intentions toward discussing end-of-life care plans with family members. RQ7a and

RQ8a asked whether acculturation and family cohesion would separately moderate the

relationships between HBM measures and behavioral intentions toward discussing end-

of-life care plans with family members.

Two hierarchical multiple regression analyses were performed with behavioral

intentions toward discussing end-of-life care plans as the dependent variable, HBM

measures as the independent variables, and acculturation or family cohesion as a

moderator. Block 1 contained control variables, including age, residence length in the

U.S., gender, religion, education levels, whether the participant had discussed my end-of-

life care plans with family members, and whether the participant had completed an AD.

Block 2 included perceived susceptibility, perceived severity, perceived benefits,

perceived barriers, and self-efficacy. Block 3 added acculturation/family cohesion as an

additional predictor. Block 4 added the interaction effects between acculturation/family

cohesion and HBM measures. The results of these regression analyses are presented in

Table 11 and 12. Table 5 shows means, standard deviations, and zero-order correlations

among perceived susceptibility, perceived severity, perceived benefits, perceived barriers,

self-efficacy, and intentions to discuss end-of-life care plans with family members.

T a b

le 1

R e su

lt s

o f

R e g re

ss io

n A

n a ly

si s

fo r

H B

M T

o w

a rd

D is

c u ss

in g E

n d -o

f- L

if e C

a re

P la

n s

(A c c u lt

u ra

ti o n a

s a M

o d e ra

to r)

V a ri

a b

le s

M o d e l

M o

d e l

M o

d e l

(S E

) β

(S E

) β

(S E

) β

B (S

E )

A g e

-. 0

1 (.

0 1 )

-. 1 0

-. 0 1 (.

0 1 )

-. 0 7

0 4

-. 0 1 (.

0 1

) -.

0 5

0 2

-. 0

1 (.

0 1

) -.

0 6

0 3

R e si

d e n

c e l

e n

g th

0 0 1 (.

0 1 )

-. 0 2

.0 0 2 (.

0 0 4 )

.0 3

.0 0

-. 0 0 1 (.

0 1

) -.

0 2

.0 0

(. 0

1 )

-. 0

0 3

G e n

d e r

-. 3

2 (.

1 9 )

-. 1 3

-. 1 6 (.

1 4 )

-. 0 7

0 4

-. 1 9 (.

1 4

) -.

0 8

-. 1

8 (.

1 5

) -.

0 7

0 4

R e li

g io

.5 3

(. 1 8 )

.2 3 *

.0 5

.2 3 (.

1 3 )

.1 0

.0 1

.2 4 (.

1 3 )

.1 0

.0 1

.2 3

(. 1

4 )

.1 0

E d u c a ti

o n

.0 7

(. 0 6 )

.1 1

.0 1

.0 3 (.

0 4 )

.0 5

.0 0

.0 2 (.

0 4 )

.0 2

.0 0

.0 3

(. 0

5 )

.0 4

0 1

W h e th

e r

I h a d d

is c u ss

e d

e n d

-o f-

li fe

c a re

w it

fa m

il y m

e m

b e rs

.8 9 (.

2 0 )

.3 7 *

.1 0

.2 3

(. 1

7 )

.1 0

.0 1

.2 2

(. 1

7 )

.0 9

.0 1

.2 0 (.

1 7 )

.0 8

.0 0 4

W h e th

e r

I h

a d

c o

m p

le te

a n A

-. 2

8 (.

2 8 )

-. 0 9

-. 3 6 (.

2 0 )

-. 1 2

-. 3 6 (.

2 0

) -.

1 2

-. 3

8 (.

2 1

) -.

1 2

P e rc

e iv

e d

s u

sc e p ti

b il

it y

(S U

S )

6 (.

0 6 )

.0 6

.0 0 2

.0 6 (.

0 6 )

.0 7

.0 0

.0 6

(. 0

6 )

.0 6

0 3

P e rc

e iv

e d

s e v e ri

ty (

S E

V )

9 (.

0 7 )

.0 8

.0 0 4

.0 9 (.

0 7 )

.0 9

.0 1

.1 0

(. 0

7 )

.0 9

P e rc

e iv

e d

b e n e fi

ts (

B E

N )

7 (.

0 8 )

.2 2 *

.0 3

.2 7 (.

0 8 )

.2 2

.0 3

.2 5

(0 8 )

.2 0

.0 3

P e rc

e iv

e d

b a rr

ie rs

( B

A R

)

-. 2 1 (.

0 7 )

-. 2 0 *

-. 2 2 (.

0 7

) -.

2 1

.0 3

-. 2

5 (.

0 8

) -.

2 3

.0 3

S e lf

-e ff

ic a c y (

S E

)

.5 2

(. 0

8 )

.4 4 * *

.1 3

.5 1

(. 0

8 )

.4 3 * *

.1 3

.4 8 (.

0 8 )

.4 0 * *

.1 0

A c c u

lt u

ra ti

o n

.1 2 (.

1 2 )

.0 8

.0 0

.1 2

(. 1

2 )

.0 9

0 3

S U

S *

A c c u

lt u

ra ti

o n

0 4

(. 0

8 )

-. 0

.0 0

S E

V *

A c c u

lt u

ra ti

o n

0 (.

1 1 )

.0 7

0 3

B E

N *

A c c u

lt u

ra ti

o n

0 9

(. 1

1 )

-. 0

.0 0

B A

R *

A c c u

lt u ra

ti o n

1 0

(. 1

1 )

-. 0

.0 0

S E

* A

c c u

lt u

ra ti

o n

0 8

(. 1

3 )

-. 0

.0 0

Δ R

.2 3

.3 8

.0 0

R 2 (A

d ju

st e d

R 2 )

.2 3 (.

2 0 )

.6 1 (.

5 8 )

.6 1

(. 5

8 )

.6 2

(. 5

7 )

F -c

h a

n g

6 .4

1 * *

2 7 .9

4 * *

.9 4

.3 7

N o te

: *

* =

p <

. 0

0 1

. *

= p

< . 0

5 .

T a b le

1 2

R e su

lt s

o f

R e g re

ss io

n A

n a ly

si s

fo r

H B

M T

o w

a rd

D is

c u ss

in g E

n d -o

f- L

if e C

a re

P la

n s

(F a m

il y C

o h e si

o n a

s a M

o d e ra

to r)

V a ri

a b

le s

M o d e l

M o

d e l

M o

d e l

(S E

) β

B (S

E )

sr 2

B (S

E )

B (S

E )

sr 2

A g e

-. 0

1 (.

0 1 )

-. 1 0

-. 0 1 (.

0 1 )

-. 0 7

0 4

-. 0 1 (.

0 1

) -.

0 7

0 4

-. 0

1 (.

0 1

) -.

0 8

R e si

d e n

c e l

e n

g th

0 0

1 (.

0 1 )

-. 0 2

.0 0 2 (.

0 0 4 )

.0 3

0 1

-. 0 0 2 (.

0 0

4 )

-. 0

.0 0

(. 0

0 4 )

. 0

.0 0

G e n

d e r

-. 3

2 (.

1 9 )

-. 1 3

-. 1 6 (.

1 4 )

-. 0 7

0 4

-. 1 6 (.

1 4

) -.

0 7

0 4

-. 2

0 (.

1 4

) -.

0 9

R e li

g io

.5 3 (.

1 8 )

.2 3 *

.0 5

3 (.

1 3 )

.1 0

.2 3 (.

1 3 )

.1 0

.0 1

1 (.

1 3 )

. 0

.0 1

E d u

c a ti

o n

.0 7 (.

0 6 )

.1 1

.0 3 (.

0 4 )

.0 5

0 2

.0 3 (.

0 4 )

.0 5

.0 0

.0 3

(. 0

4 )

. 0

.0 0

W h

e th

e r

I h

a d

d is

c u ss

e d

e n d

-o f-

li fe

c a re

w it

fa m

il y m

e m

b e rs

.8 9 (.

2 0 )

.3 7 *

.1 0

3 (.

1 7 )

.1 0

.2 3 (.

1 7 )

.0 9

.0 1

8 (.

1 6 )

. 1

.0 1

W h

e th

e r

I h

a d

c o

m p

le te

a n A

-. 2

8 (.

2 8 )

-. 0 9

-. 3 6 (.

2 0 )

-. 1 2

-. 3 7 (.

2 0

) -.

1 2

-. 2

8 (.

2 0

) -.

0 9

P e rc

e iv

e d

s u

sc e p ti

b il

it y

(S U

S )

6 (.

0 6 )

.0 6

0 2

.0 5 (.

0 6 )

.0 6

.0 0

.0 6

(. 0

6 )

. 0

.0 0

P e rc

e iv

e d

s e v e ri

ty (

S E

V )

9 (.

0 7 )

.0 8

0 4

.0 9 (.

0 7 )

.0 8

.0 1

5 (.

0 7 )

. 0

.0 0 1

P e rc

e iv

e d

b e n e fi

ts (

B E

N )

7 (.

0 8 )

.2 2 *

.0 3

.2 8 (.

0 8 )

.2 2

.0 3

7 (0

8 )

. 2

2 *

.0 3

P e rc

e iv

e d

b a rr

ie rs

( B

A R

)

-. 2 1 (.

0 7 )

-. 2 0 *

-. 2 2 (.

0 7

) -.

2 0

.0 3

1 9

(. 0

7 )

-. 1

8 *

S e lf

-e ff

ic a c y (

S E

)

.5 2 (.

0 8 )

.4 4 * *

.1 3

.5 2 (.

0 8 )

.4 3

* *

.1 3

8 (.

0 8 )

.4 0

* *

.1 0

F a m

il y c

o h

e si

o n (

F C

)

0 9 (.

1 4

) -.

0 3

0 1

-. 0

5 (.

1 4

) -.

0 2

S U

S *

F C

.2 1

(. 1

4 )

. 1

.0 1

S E

V * F

2 0 (.

1 6 )

-. 0 8

0 4

B E

N *

F C

-. 4

4 (.

1 9

) -.

1 7

.0 1

B A

R *

F C

-. 1

5 (.

1 3

) -.

0 7

0 4

S E

* F

6 (.

1 9 )

.0 0

Δ R

.2 3

.0 0

.0 3

R 2 (A

d ju

st e d

R 2 )

.2 3 (.

2 0 )

.6 1 (.

5 8 )

.6 1

(. 5

8 )

.6 5

(. 6

0 )

F -c

h a

n g

6 .4

1 * *

2 7 .9

4 * *

.4 0

.6 0

N o te

. *

* =

p <

. 0

0 1

. *

= p

< . 0

5 .

At Block 1, the seven control variables explained a significant amount of

variation, R2-change = .23, F-change(7, 148) = 6.41, p < .001. Religion and whether the

participant had discussed end-of-life care plans with family members were the significant

positive predictors for behavioral intentions.

At Block 2, perceived susceptibility, perceived severity, perceived benefits,

perceived barriers, and self-efficacy accounted for a statistically significant amount of

variance in behavioral intentions beyond control variables, R2-change = .38,

F-change(5,143) = 27.94, p < .001. The proportion of variation in behavioral intentions

explained by the set of predictors was statistically significant, R2 = .61, adjusted R2 = .58,

F(12, 143) = 18.78, p < .001. Perceived benefits (β = .22, sr2 = .03, p < .001), perceived

barriers, (β = -.20, sr2 = .02, p < .05), and self-efficacy (β = .44, sr2 = .13, p < .001) were

significant predictors for behavioral intentions. H2a was supported.

RQ7a: Acculturation as a Moderator. At Block 3 acculturation was added as an

additional predictor and did not increase a statistically proportion of variation in

behavioral intentions. At Block 4 all the five interaction effects between HBM measures

and acculturation were included as predictors together. They did not increase a

statistically proportion of variation in behavioral intentions. The results are presented in

Table 11. That is, acculturation did not predict participants’ intentions to discuss end-of-

life care plans with family members after the inclusion of perceived susceptibility,

perceived severity, perceived benefits, perceived barriers, and self-efficacy. Acculturation

did not moderate the relationships between HBM measures and behavioral intentions.

RQ8a: Family Cohesion as a Moderator. Another hierarchical regression was

conducted with behavioral intentions toward discussing end-of-life care plans with family

members as the dependent variable, HBM measures as the independent variables, and

family cohesion as a moderator. The results are presented in Table 12. Family cohesion

did not increase a statistically proportion of variation in behavioral intentions. However,

adding the interaction effects between HBM measures and family cohesion increased a

statistically proportion of variation in behavioral intentions, R2-change = .03, F-change(5,

137) = 2.60, p < .05. Perceived benefits (β = .22, sr2 = .03, p < .05), perceived barriers (β

= -.18, sr2 = .02, p < .05), and self-efficacy (β = .40, sr2 = .10, p < .001) remained

significant, but the sizes of the significant effects of perceived barriers and self-efficacy

decreased. The interaction between family cohesion and perceived benefits worked as an

additional significant predictor for behavioral intention, β = -.17, sr2 = .01, p < .05. When

family cohesion increased, perceived benefits had a smaller impact on intentions to

discuss end-of-life care plans with family members.

The significant interaction effect was probed with simple slopes analysis. Results

of the simple slopes analysis is presented in Figure 4. “High,” “moderate,” and “low”

levels of responses to participants’ family cohesion were created by computing one

standard deviation above and one standard deviation below the mean. This analysis

indicated that the relationship between perceived benefits and behavioral intentions was

larger when participants had lower-level family cohesion.

Figure 4. Family Cohesion as a Moderator between Perceived Benefits and Behavioral Intention

toward Discussing End-of-Life Care Plans with Family Members.

H2b: Completing an Advance Directive

H2b predicted that participants’ perceived susceptibility, perceived severity,

perceived barriers, perceived benefits, and self-efficacy would predict behavioral

intentions toward completing an AD. RQ7b and RQ8b asked whether acculturation and

family cohesion would separately moderate the relationships between HBM measures

and behavioral intentions toward completing an AD.

Two hierarchical multiple regression analyses were performed with behavioral

intentions toward completing an AD as the dependent variable, HBM measures as the

independent variables, and acculturation or family cohesion as a moderator. Block 1

contained control variables, including age, residence length in the U.S., gender, religion,

0.5

1.5

2.5

3.5

4.5

Low Moderate High

In te

n ti

o n

Perceived benefits

Low

Moderate

High

education levels, and whether the participant had discussed end-of-life care plans with

family members. Block 2 included perceived susceptibility, perceived severity, perceived

benefits, perceived barriers, and self-efficacy. Block 3 added acculturation/family

cohesion as an additional predictor. Block 4 added the interaction effects between

acculturation/family cohesion and HBM measures. The results of these regression

analyses are presented in Table 13 and 14. Table 8 shows means, standard deviations, and

zero-order correlations among perceived susceptibility, perceived severity, perceived

benefits, perceived barriers, self-efficacy, and intentions to complete an AD.

At Block 1, the six control variables explained a significant amount of variation,

R2-change = .19, F-change(6, 128) = 4.95, p < .001. Education levels and whether the

participant had discussed end-of-life care plans with family members were the significant

positive predictors for behavioral intentions, while age and residence length in the U.S.

predicted intentions to complete an AD negatively.

At Block 2, perceived susceptibility, perceived severity, perceived benefits,

perceived barriers, and self-efficacy accounted for a substantial amount of variation in

intentions beyond control variables, R2-change = .40, F-change(5,123) = 24.05, p < .001.

The proportion of variation in intentions explained by the set of predictors was

statistically significant, R2 = .59, adjusted R2 = .55, F(11, 123) = 16.06, p < .001.

Perceived susceptibility (β = .25, sr2 = .04, p < .05), perceived benefits (β = .43, sr2 = .10,

p < .001), perceived barriers (β = -.16, sr2 = .02, p < .05), and self-efficacy (β = .25,

sr 2 = .04, p < .001) were significant predictors for intentions. H2b was supported.

T a b le

1 3

R e su

lt s

o f

R e g re

ss io

n A

n a ly

si s

fo r

H B

M T

o w

a rd

C o m

p le

ti n g a

n A

d v a n c e D

ir e c ti

v e (

A c c u

lt u

ra ti

o n

a s

a M

o d

e ra

to r)

V a ri

a b le

s M

o d e l

M o d e l

M o

d e l

M o

d e l

(S E

) β

B (S

E )

sr 2

B (S

E )

sr 2

B (S

E )

A g e

-. 0

3 (.

0 1 )

-. 1 7 *

-. 0 1 (.

0 1 )

-. 0 7

-. 0 1 (.

0 1

) -.

0 7

0 3

-. 0

1 (.

0 1

) -.

0 7

0 4

R e si

d e n

c e l

e n

g th

0 1

(. 0

1 )

-. 1 9 *

.0 0 3 (.

0 0 4 )

.0 4

0 1

.0 0 1 (.

0 1 )

.0 1

.0 0

1 (.

0 1 )

.0 1

.0 0

G e n d e r

-. 1

7 (

.2 2 )

-. 0 7

0 4

-. 0 5 (.

1 6 )

-. 0 2

-. 0 4 (.

1 6

) -.

0 2

-. 0

2 (.

1 7

) -.

0 1

R e li

g io

.3 8

(. 2 1 )

.1 5

.0 2

-. 0 6 (.

1 6 )

-. 0 2

-. 0 6 (.

1 6

) -.

0 3

0 1

-. 0

7 (.

1 7

) -.

0 3

0 1

E d u c a ti

o n

.1 5

(. 0 7 )

.1 9 *

.0 3

.0 1 (.

0 5 )

.0 1

-. 0 0 2 (.

0 6

) -.

0 0

.0 0

1 (.

0 6 )

.0 0

W h e th

e r

I h

a d

d is

c u ss

e d

e n d

-o f-

li fe

c a re

w it

fa m

il y m

e m

b e rs

.8 9

(. 2 5 )

.2 8 *

.0 8

.2 7 (.

2 0 )

.0 9

.2 7 (.

2 0 )

.0 9

.2 8

(. 2

1 )

.0 9

.0 1

P e rc

e iv

e d

s u

sc e p ti

b il

it y

(S U

S )

7 (.

0 8 )

.2 5 *

.0 4

6 (.

0 8 )

.2 5

.0 4

1 (.

1 0 )

.2 0

.0 2

P e rc

e iv

e d

s e v e ri

ty (

S E

V )

1 5 (.

0 9 )

-. 1 3

-. 1 5 (.

0 9

) -.

1 2

-. 0

9 (.

1 2

) -.

0 7

0 2

P e rc

e iv

e d

b e n e fi

ts (

B E

N )

8 (.

1 1 )

.4 3 * *

.1 0

8 (.

1 1 )

.4 3

* *

.1 0

1 (.

1 3 )

.3 7

* *

.0 5

P e rc

e iv

e d

b a rr

ie rs

( B

A R

)

-. 1 8 (.

0 8 )

-. 1 6 *

-. 1 8 (.

0 8

) -.

1 5

.0 2

1 9

(. 0

9 )

-. 1

7 *

S e lf

-e ff

ic a c y (

S E

)

.3 1 (.

0 9 )

.2 5 *

.0 4

1 (.

0 9 )

.2 5

.0 4

2 (.

1 1 )

.2 6

.0 3

A c c u lt

u ra

ti o

.0 7 (.

1 5 )

.0 4

0 1

.0 6

(. 1

6 )

.0 4

.0 0

S U

S * A

c c u

lt u

ra ti

o n

1 1

(. 1

2 )

-. 0

.0 0

S E

V * A

c c u lt

u ra

ti o n

2 (.

1 6 )

.0 8

.0 0 2

B E

N *

A c c u

lt u

ra ti

o n

1 8

(. 1

8 )

-. 0

.0 0

B A

R *

A c c u

lt u ra

ti o n

0 5

(. 1

3 )

-. 0

.0 0

S E

* A

c c u

lt u

ra ti

o n

0 1

(. 1

3 )

-. 0

.0 0

Δ R

.1 9

.0 0

.0 1

R 2 (A

d ju

st e d

R 2 )

.1 9 (.

1 5 )

.5 9 (.

5 5 )

.5 9

(. 5

5 )

.6 0

(. 5

4 )

F -c

h a n

g e

4 .9

5 * *

2 4 .0

5 * *

.1 9

N o te

: *

* =

p <

. 0

0 1

. *

= p

< . 0

5 .

T a b le

1 4

R e su

lt s

o f

R e g re

ss io

n A

n a ly

si s

fo r

H B

M T

o w

a rd

C o m

p le

ti n g a

n A

d v a n c e D

ir e c ti

v e (

F a

m il

y C

o h e si

o n a

s a M

o d e ra

to r)

V a ri

a b le

s M

o d e l

M o d e l

M o

d e l

M o

d e l

(S E

) β

B (S

E )

sr 2

B (S

E )

sr 2

B (S

E )

A g e

-. 0

3 (.

0 1 )

-. 1 7 *

-. 0 1 (.

0 1 )

-. 0 7

-. 0 1 (.

0 1 )

-. 0

.0 1

-. 0

1 (.

0 1

) -.

0 8

R e si

d e n

c e l

e n

g th

0 1

(. 0 1 )

-. 1 9 *

.0 0 3 (.

0 0 4 )

.0 4

0 1

.0 0 3 (.

0 0 4

) .0

.0 0 1

.0 0

2 (.

0 1 )

.0 3

.0 0

G e n d e r

-. 1

(. 2

2 )

-. 0 7

0 4

-. 0 5 (.

1 6 )

-. 0 2

-. 0 5 (.

1 6 )

-. 0

.0 0

-. 0

3 (.

1 7

) -.

0 1

R e li

g io

.3 8

(. 2 1 )

.1 5

.0 2

-. 0 6 (.

1 6 )

-. 0 2

-. 0 6 (.

1 6 )

-. 0

.0 0

-. 0

7 (.

1 7

) -.

0 3

0 1

E d u c a ti

o n

.1 5

(. 0 7 )

.1 9 *

.0 3

.0 1 (.

0 5 )

.0 1

.0 1 (.

0 5 )

.0 1

.0 0

-. 0

0 1

(. 0

6 )

-. 0

0 1

W h e th

e r

I h

a d

d is

c u ss

e d

e n d

-o f-

li fe

c a re

w it

fa m

il y m

e m

b e rs

.8 9

(. 2 5 )

.2 8 *

.0 8

.2 7 (.

2 0 )

.0 9

.2 7 (.

2 0 )

.0 9

.0 1

3 (.

2 1 )

.0 7

.0 0

P e rc

e iv

e d

s u

sc e p ti

b il

it y

(S U

S )

7 (.

0 8 )

.2 5 *

.0 4

.2 7 (.

0 8 )

.2 5

.0 4

4 (.

0 8 )

.2 3

.0 3

P e rc

e iv

e d

s e v e ri

ty (

S E

V )

1 5 (.

0 9 )

-. 1 3

-. 1 5 (.

0 9 )

-. 1

.0 1

1 7

(. 1

0 )

-. 1

.0 1

P e rc

e iv

e d

b e n e fi

ts (

B E

N )

8 (.

1 1 )

.4 3 * *

.1 0

.5 8 (.

1 1 )

.4 3

* *

.1 0

2 (.

1 2 )

.4 6

* *

.1 0

P e rc

e iv

e d

b a rr

ie rs

( B

A R

)

-. 1 8 (.

0 8 )

-. 1 6 *

-. 1 8 (.

0 8 )

-. 1

6 *

-. 2

2 (.

0 8

) -.

1 9

.0 3

S e lf

-e ff

ic a c y (

S E

)

.3 1 (.

0 9 )

.2 5 *

.0 4

.3 1 (.

0 9 )

.2 5

.0 4

2 (.

0 9 )

.2 6

.0 4

F a m

il y c

o h

e si

o n (

F C

)

0 1 (.

1 8 )

-. 0

0 2

-. 0

7 (.

2 0

) -.

0 2

S U

S *

F C

-. 1

3 (.

2 0

) -.

0 5

0 1

S E

V *

F C

.3 9

(. 2

5 )

.1 3

.0 1

B E

N *

F C

-. 1 2 (.

2 9 )

-. 0 4

0 1

B A

R * F

0 7

(. 1

9 )

-. 0

.0 0

S E

* F

7 (.

2 0 )

.0 3

.0 0

Δ R

.1 9

.4 0

.0 2

R 2 (A

d ju

st e d

R 2 )

.1 9 (.

1 5 )

.5 9 (.

5 5 )

.5 9

(. 5

5 )

.6 0

(. 5

5 )

F -c

h a n

g e

4 .9

5 * *

2 4 .0

5 * *

.0 0

.8 8

N o te

: *

* =

p <

. 0

0 1

. *

= p

< . 0

5 .

RQ7b: Acculturation as a Moderator. At Block 3, acculturation was added as

an additional predictor and did not increase a statistically proportion of variation in

behavioral intentions. At Block 4, all the five interaction effects between HBM measures

and acculturation were included as predictors together. They did not increase a

statistically proportion of variation in behavioral intentions. The results are presented in

Table 13. Acculturation did not predict participants’ intentions to complete an AD after

the inclusion of HBM measures. Acculturation did not moderate the relationships

between HBM measures and intentions to complete an AD.

RQ8b: Family Cohesion as a Moderator. Another hierarchical regression was

conducted with behavioral intentions toward completing an AD as the dependent variable

and HBM measures as the independent variables and family cohesion as a moderator.

The results are presented in Table 14. Family cohesion did not improve the proportion of

variation in behavioral intentions as an additional predictor. The interaction effects

between family cohesion and TPB measures did not increase a significant amount of

variation in participants’ intentions to complete an AD. Family cohesion did not

moderate the relationships between HBM measures and intentions to complete an AD.

Comparing Theory of Planned Behavior and Health Belief Model

H3a predicted that the TPB had more predictive power for participants’ intentions

to discuss end-of-life care plans with family members than the HBM. As addressed

earlier, the inclusion of TPB measures accounted for a statistically significant amount of

variation in intentions beyond control variables, R2-change = .39, F-change(3,145) =

49.34, p < .001. In contrast, HBM measures accounted for a statistically significant

amount of variation in intentions beyond control variables, R2-change = .38,

F-change(5,143) = 27.94, p < .001. TPB measures accounted for 1% more variance in

intentions than HBM measures. This finding suggested that the TPB had more predictive

power for participants’ intentions toward discussing end-of-life care plans with family

members than the HBM. H3a was supported.

H3b predicted that the TPB had more predictive power for participants’ intentions

toward completing an AD than the HBM. TPB measures accounted for a substantial

amount of variation in intentions beyond control variables, R2-change = .42,

F-change(3,125) = 45.3, p < .001. HBM measures accounted for a statistically significant

amount of variation in intentions beyond control variables, R2-change = .40,

F-change(5,123) = 24.05, p < .001. TPB measures accounted for 2% more variance in

intentions than HBM measures. This result implied that the TPB had more predictive

power for participants’ intentions toward completing an AD than the HBM. H3b was

supported.

CHAPTER 5: DISCUSSION

Overview

This study examined the factors that can predict community-dwelling Chinese

American older adults’ intentions to plan for end-of-life care toward the two

recommended behaviors in advance care planning (ACP), discussing end-of-life care

plans with family members and completing an advance directive (AD). The results

indicated that attitudes, subjective norms, and self-efficacy were significant predictors for

behavioral intentions toward the two behaviors in this study. In terms of Health Belief

Model (HBM), perceived benefits, perceived barriers, and self-efficacy were significant

predictors for behavioral intentions toward the two behaviors. Perceived susceptibility

only significantly increased variation in intentions to complete an AD. In addition,

acculturation moderated the relationship between attitudes and intentions to complete an

AD. Family cohesion moderate the relationship between perceived benefits and

intentions to discuss end-of-life care plans with family members.

Theory of Planned Behavior and Health Belief Model

Theory of Planned Behavior

The results indicated that there were significant differences of subjective norms,

self-efficacy, and behavioral intentions toward the two behaviors. That is, participants

scored higher in subjective norms, self-efficacy, and behavioral intentions toward

discussing end-of-life care plans with family members than completing an AD. This

finding suggested that it is possible for participants to perceive more family support and

capability for ACP conversations. They were more likely to discuss their end-of-life care

plans with family members compared to the completion of an AD.

This study indicated that attitudes, subjective norms, and self-efficacy had

important predictability for behavioral intentions toward the two behaviors but the

predictability of these measures toward these two behaviors were different. According to

the existing meta-analyses, attitude is the most important predictor (Albarracin et al.,

2001; Downs & Hausenblas, 2005). This contradicted the current study showing that

attitude did not work as the most influential determinant of behavioral intentions toward

the two behaviors. Instead, the results illustrated that self-efficacy played the most

significant role in predicting participants’ intentions to discuss end-of-life care plans and

subjective norms had the largest influence on intentions to complete an AD. Given the

changing influence of these factors on different behavioral intentions, it is helpful for

intervention designers to understand what factor is most likely to improve specific

behavioral intentions for ACP promotion.

The previous meta-analyses suggested that subjective norms did not have the

strongest relationship with intentions compared with attitudes (Cooke et al., 2014;

McEachan et al., 2011). However, subjective norms had the strongest relationship with

intentions to complete an AD. This finding showed that 8% variances in intentions

toward the completion of an AD was explained by subjective norms, beyond that

explained by the other predictors, while 3% variances in intentions to discuss end-of-life

care plans with family members was explained by subjective norms, beyond that

explained by the other predictors. In other words, participants’ understandings of their

significant others’ reactions toward the completion of an AD had a stronger impact on

their likelihood of behavioral performance than the other behavior. Both strong effects of

subjective norms toward the two behaviors reflect that due to the potential influence of

collectivism, having perceived social approval from the reference group is a critical

determinant for engaging in a behavior and can be more salient than other predictors.

Self-efficacy was a strong predictor of behavioral intentions toward the two

behaviors. Interestingly, 4% variances in intentions to discuss end-of-life care plans with

family members was explained by self-efficacy, beyond that explained by the other

predictors. However, 2% variances in intentions to complete an AD was explained self-

efficacy, beyond that explained by the other predictors. That is, participants’ self-efficacy

toward discussing end-of-life care plans with family members had a stronger impact on

their likelihood of behavioral performance compared with the other behavior. Because

completing an AD requires sufficient English proficiency or others’ assistance when a

person does not speak English, participants may feel more challenging to take actions

toward the completion of an AD. They were likely to perceive more abilities to have end-

of-life care conversations with their family members.

Health Belief Model

In the HBM, perceived benefits, perceived barriers, and self-efficacy were

significant predictors for participants’ intentions toward the two behaviors. Regarding

discussing end-of-life care plans with family members, self-efficacy was the strongest

predictor for behavioral intentions, followed by perceived benefits, and perceived

barriers. In terms of the completion of an AD, perceived benefits were the most important

factor predicting behavioral intentions, followed by perceived susceptibility and self-

efficacy, and perceived barriers. Self-efficacy explained more unique variation in

intentions to discuss end-of-life care plan with family members than intentions to

complete an AD. This finding was consistent with the TPB results of this study that self-

efficacy explained more unique variance in behavioral intentions toward discussing end-

of-life care plans with family members. This similarity provided strong evidence that

participants felt less capable of the completion of an AD than the other behavior.

The positive effects of perceived benefits and negative effects of perceived

barriers on behavioral outcomes were consistent with the previous literature (Carpenter,

2010; Harrison et al., 1992). However, the size of significant effects of perceived benefits

was larger than that of perceived barriers toward the two behaviors. It is crucial for

participants to perceive benefits of ACP to overcome relevant barriers. Although barriers

(e.g., I am not sick enough to discuss end of life care plans with family members/I am not

sick enough to complete an AD) inhibit the target population from ACP engagement,

addressing the benefits of ACP would help reduce the negative influence of barriers.

Severity was a weak factor predicting participants’ behavioral outcomes. This

finding was consistent with the previous HBM meta-analysis (Harrison et al., 1992).

However, a study of Korean American and Non-Hispanic White older adults found that

the likelihood of end-of-life communication would increase as perceived severity

increased. This inconsistency indicated that the influence of severity varied across

different health contexts and populations. Participants in the current study had low ACP

awareness. It is likely that many of them had not developed a solid understanding of the

seriousness of having unwanted medical treatments. Therefore, perceived severity did not

work as an important predictor in this study.

Comparing Theory of Planned Behavior and Health Belief Model

The inclusion of attitudes, subjective norms, self-efficacy significantly separately

increased 39% and 42% variation in behavioral intentions toward discussing end-of-life

care plans with family members and completing an AD. This finding was consistent with

the previous TPB meta-analyses showing that TPB measures can explain approximately

40%–49% of variances in intentions (Armitage & Conner, 2001; McEachan et al., 2011).

In contrast, adding perceived susceptibility, perceived severity, perceived benefits,

perceived barriers, and self-efficacy significantly increased 38% and 40% variances in

intentions toward discussing end-of-life care plans with family members and completing

an AD respectively. These findings indicated that both the TPB and HBM had predictive

power to explain participants’ intentions to plan for end-of-life care, but TPB measures

accounted for slightly more variation in intentions.

Acculturation and Family Cohesion of Chinese American Older Adults

In this study, participants had low acculturation levels but high family cohesion

levels. This finding was consistent with the existing literature showing Chinese American

elders had low-level acculturation and high-level family dependency (Dong, Bergren, &

Chang, 2015; Hsiung & Ferrans, 2007). In a population study of Chinese older adults in

Chicago, lower acculturation level was found to be associated with older age, more

offspring, lower income, fewer years living in the U.S., lower overall health status, and

lower quality of life among 3159 participants (Dong et al., 2015). Hsiung and Ferrans

(2007) considered Chinese American older immigrants as the most traditional and least

acculturated compared to other Chinese populations in the U.S. Most Chinese American

older immigrants were foreign-born and arrived in the U.S. at advanced ages. They

primarily came from mainland China, Hong Kong, Taiwan, Singapore, or some other

areas to be with their children and help take care of their grandchildren. They usually

speak Mandarin, Cantonese, or other Chinese dialects, know little or no English, socialize

with other Chinese American older immigrants, maintain traditional Chinese beliefs, and

favor family responsibility and support over individual independence. It is not surprising

that participants were found to have low acculturation and high family cohesion levels.

Acculturation and family cohesion did not explain the TPB and HBM as

additional predictors. Limited variations in these two variables may explain their

nonsignificant effects on the TPB and HBM. Also, both TPB and HBM measures

explained a substantial amount of variances in intentions. Therefore, the inclusion of

acculturation and family cohesion did not increase variances in intentions. Furthermore,

acculturation and family cohesion measurements used in this study did not capture

specific beliefs related to ACP and may result in the nonsignificant effects of

acculturation and family cohesions on intentions.

Interestingly, acculturation moderated the relationship between attitude and

intention to complete an AD negatively. In other words, when participants were less

acculturated, their attitudes had a larger impact on their intentions toward the completion

of an AD. In contrast, for participants in a higher acculturated group, their attitudes had a

smaller impact on their intentions toward the completion of an AD. It is likely that higher

acculturation levels made participants become more aware of the complexity of ACP

process. Having positive reactions toward the completion of an AD does not guarantee

their successful engagement in the completion of an AD. Also, there was no significant

relationship between acculturation and intentions to complete an AD, while acculturation

was positively associated with intentions to discuss end-of-life care plans with family

members. This implied that the role of acculturation can vary in different behaviors.

Additionally, family cohesion moderated the relationship between perceived

benefits and intention to discuss end-of-life care plans with family members negatively.

When participants had lower family cohesion levels, their perceived benefits had a larger

impact on their intentions toward the discussion of end-of-life care plans with family

members. When they had higher family cohesion levels, their perceived benefits had a

smaller impact on their intentions. It is possible that when Chinese American older adults

and their families are closely united and under the influence of filial piety, their children

and grandchildren may provide them strong assistance for their daily life, diminishing

their need to learn U.S. customs (Dong et al., 2015). Consequently, older adults with

higher-level family cohesion may not perceive more benefits of end-of-life care

discussions and wait for their highly acculturated family members for decision making.

Implications

Advance Care Planning Interventions

As behavior contributes to the cause of mortality and morbidity (Michie &

Johnson, 2012; Parkin, Boyd, & Walker, 2011), interventions to change behavior-related

components are essential. Social and behavioral science theories systematically explain

why individuals behave the ways they do. Therefore, health promotion interventions that

are based on social and behavioral science theories are more effective than those lacking

a theoretical base (Glanz & Bishop, 2010; Green, 2000; Noar, Benac, & Harris, 2007;

Sales, Smith, Curran, & Kochevar, 2006). It is important to apply theories to increase the

effectiveness of intervention design, implementation, and evaluation.

This theory-based study would provide guidance to inform the future

interventions for Chinese American older adults. To the best of my knowledge, only a

few ACP interventions have been created for Chinese elders in Hong Kong, including Let

Me Talk, Anticipatory Grief Therapy, and Dignity-Conserving End-of-Life Care Program

(Chan & Pang, 2010; Cheng et al., 2010; Ho et al., 2016). There were many ACP

interventions in the U.S., but none of them were particularly designed to improve Chinese

American older adults’ ACP awareness and engagement. Lee, Hinderer, and Friedmann

(2015) designed a one-hour nurse-led culturally-sensitive seminar for Chinese American

adults aged from 32 to 87 and offered a step-by-step guide of the AD completion process.

Although this seminar introduced participants ACP components, researchers and

intervention designers need to rely more on communication.

The current study finding suggested that participants reported higher intentions to

discuss end-of-life care plans with family members, compared with intentions to

complete an AD. It is likely that Chinese American older adults feel more comfortable to

communicate with their family members rather than the completion of an AD. This

implication is consistent with what the existing interventions suggested among Chinese

older adults in Hong Kong (Chan & Pang, 2010; Cheng, Lo, Chan, & Woo, 2010; Ho et

al., 2016). These interventions commonly valued the importance and complexity of

communication on planning for end-of-life care and included multiple sessions to invite

participants to share their memorable life stories and lessons, lessons they have learned

about life and death, and then guided them to explore their understandings of life-

sustaining treatments, their end-of-life care expectations, and their preferred health care

decision maker. Planning for end-of-life care is not a one-time commitment for Chinese

American older adults and their family members. Instead, it should be considered as an

ongoing process involving communication that matches older adults’ life values and

medical preferences. When an ACP-related intervention is designed for the target

population, it is helpful to consider changing people’s intentions to discuss end-of-life

care plans with family members first. Both TPB and HBM had strong predictive power

for the target population’s intentions to discuss end-of-life care plans with family

members. However, the TPB accounted for slightly more variation in intentions and can

be given a priority for the design of an intervention.

Given that self-efficacy is the strongest predictor for intentions to discuss end-of-

life care plans in both TPB and HBM, it can be a good component to focus on in an ACP-

related intervention. As Fishbein and Ajzen (2010) suggested, people’s perceptions of

their capabilities of performing a behavior are associated with their perceptions of having

resources available to perform a behavior and the extents to which having resources

available to perform a behavior is sufficient to overcome barriers to perform the behavior.

It can be helpful for intervention designers to get inspiration from the HBM measures,

perceived barriers and benefits. These measures can be included to educate the target

population to develop self-efficacy to overcome the barriers to discuss end-of-life care

plans with family members and maximize the values of perceived benefits.

Additionally, to design effective efficacy-related components for the intervention,

it is necessary to further explore the reasons why people have difficulty in making their

decisions whether they will discuss their end-of-life care plans with family members and

validate the perceived barriers scale among Chinese American older adults. Based on the

preliminary results, perceived barriers yielded a two-factor solution. One factor focused

on participants’ future care unpreparedness, while the other factor addressed participants’

concerns about the potential negative consequences of planning for end-of-life care in

advance. In the previous literature researchers commonly considered perceived barriers

as a composite variable in ACP contexts and did not specify perceived barriers in the

prediction of behavioral performance (Ko & Lee, 2009; Szalai, 2015). Without knowing

about specific barriers for the target population, intervention designers may not be able to

develop useful information to reach the goal of improving participants’ ACP awareness

and engagement. The multidimensional nature of perceived barriers should be examined

more specifically for the effectiveness of targeting self-efficacy among Chinese

American older adults.

Challenge of Using Timeline for the Recommended Behaviors

One challenge is that the two recommended behaviors identified in this

dissertation did not include the time at which they are performed. To design theory-based

studies to alleviate social problems, it is very important to firstly identify a recommended

behavior that researchers want participants to reinforce or change (Fishbein & Ajzen,

1975; Downs & Hausenblas, 2005). For the identification of a specific behavior, Fishbein

and Ajzen (2010) said that it is helpful to consider a behavior with four elements to

maximize the behavioral prediction, “the action performed, the target at which the action

is directed, the context in which it is performed, and the time at which it is performed” (p.

29). However, the recommended behaviors identified in this dissertation did not include

the time at which they are performed, because the formative research and previous

literature did not provide strong evidence to support the effectiveness of including time at

which behaviors are performed. Hsiung (2015) found that the cutoff points of 30 days

and six months that are frequently used in smoking cessation research may be arbitrary in

behavior change related to ACP. Having ACP on end-of-life care with significant others

were more complicated than a person’s decision to quit smoking. As this dissertation

addressed earlier, most participants in formative research could not predict when they

would engage in ACP processes. Participants had good intentions but might not be

decisive about timing. In this regard, using a clear cutoff point (e.g., 30 days) is not an

effective way to measure the target population’s behavioral components. More future

studies are needed to further explore this challenge to increase the intervention

effectiveness for the target population.

Strengths

One of the major strengths of the study was the sample. The target population was

community-dwelling Chinese American adults aged 55 and older living in the Phoenix

metropolitan areas. Considering the majority of ACP research focusing on non-Hispanic

Whites and increasing need for community-based research, this study can help improve

the sample diversity in terms of race, ethnicity and study location. Furthermore, given the

challenges of promoting ACP in Chinese American senior communities, the participants

of the study are an ideal choice for researchers and intervention designers to understand

this population.

Another strength is that this study was theory-driven and applied the TPB and

HBM to examine Chinese American older participants’ intentions to plan for their end-

of-life care. This study provided strong empirical support for the TPB and HBM for an

important topic in an ethnic minority population and had significant implications for

developing effective theory-driven interventions to improve the target population’s ACP

awareness and engagement.

100

Additionally, the measures for the TPB and HBM measures were guided by

validated scales, previous literature, and formative research in this dissertation. A

combination of qualitative and quantitative approaches were used to explore the

complexity of the target population’s understandings of ACP on end-of-life care and

ensure the effectiveness of measures written in multiple languages.

Limitations

There are several limitations in this dissertation. Firstly, most participants were

Chinese-speaking Chinese Americans and foreign-born immigrants. They were more

likely to have lower acculturation levels and higher family cohesion levels than English-

speaking and U.S. born Chinese Americans. In the future studies, researchers should

consider recruiting more English-speaking Chinese American older adults to further

explore the acculturation influence on ACP among Chinese American senior

communities to increase sample diversity.

Secondly, data collection was performed in the Phoenix metropolitan areas.

Although this study demonstrated ACP understandings of Chinese American older adults

living in the Phoenix metropolitan areas, the findings may not be generalized to other

Chinese aging populations in the U.S. Furthermore, participants from similar community

settings were more likely to see the same health care providers and receive the same

ACP-related information than samples drawn randomly from the nationwide population.

This may limit the generalizability of this study. To understand Chinese American older

adults’ ACP processes, it is helpful to recruit participants from different regions.

Thirdly, this dissertation applied individual-level theoretical frameworks and did

not primarily consider the determinants from other levels such as community-based

101

influence. Given the multilevel impacts of health promotive behaviors, it would be

worthwhile to consider the complex interaction between individual, relationship,

community, and societal factors in the future studies related to ACP.

Conclusion

This study examined the factors predicting community-dwelling Chinese

American older adults’ intentions toward discussing end-of-life care plans with family

members and completing an AD respectively. The results indicated that attitudes,

subjective norms, and self-efficacy worked as a significant set of predictors for

behavioral intentions toward the two behaviors in this study. In the HBM, perceived

benefits, perceived barriers, and self-efficacy were significant predictors for behavioral

intentions toward the two behaviors. Perceived susceptibility only significantly increased

variation in intention to complete an AD. In addition, acculturation moderated the

relationship between attitude an intention to complete an AD negatively. Family cohesion

moderate the relationship between perceived benefits and intention to discuss end-of-life

care plans with family members negatively. Overall, the TPB and HBM provided strong

theoretical foundations for us to understanding Chinese American older adults’ behaviors

related to ACP. The study findings would inform the future interventions to improve

Chinese American older adults’ ACP awareness and engagement.

102

Endnote

1In addition to testing self-efficacy, perceived control was measured through three

5-point Likert items (1=completely disagree, 5=completely agree). These items were

drawn from a Fishbein and Ajzen’s (2010) control belief scale and modified to fit the

study context. Sample items included: “Whether I discuss my end-of-life care plans with

family members is under my personal control/Whether I complete an advance directive is

under my personal control.” Cronbach’s alpha for perceived control toward discussing

end-of-life care plans with family members was .82. Cronbach’s alpha for perceived

control toward completing an advance directive was .91. Perceived control scores were

calculated based on the mean across the three items for both behaviors. Two hierarchical

multiple regression analyses were performed with behavioral intentions toward the two

behaviors, and attitudes, subjective norms, self-efficacy, and perceived control as

independent variables. Control for the effects of demographic variables, attitudes,

subjective norms, and self-efficacy were significant predictors except perceived control.

103

REFERENCES

Aging with Dignity. (2011). Five wishes. Retrieved from

https://agingwithdignity.org/docs/default-source/default-document-

library/product-samples/fwsample.pdf?sfvrsn=2

Ajzen, I. (1985). From intentions to action: A theory of planned behavior. In J. Kuhl & J.

Beckman (Eds.), Action control: From cognitions to behaviors (pp. 11–39). New

York: Springer.

Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human

Decision Processes, 50, 179–211.

Ajzen, I. (2002). Perceived behavioral control, self-efficacy, locus of control, and the

theory of planned behavior. Journal of Applied Social Psychology, 32, 665–683.

Albarracin, D., Johnson, B. T., Fishbein, M., & Muellerleile, P. A. (2001). Theories of

reasoned action and planned behavior as models of condom use: A meta-

analysis. Psychological Bulletin, 127, 142–161.

Alegria, M., Vila, D., Woo, M., Canino, G., Takeuchi, D., Vera, M.,…Shrout, P. (2004).

Cultural relevance and equivalence in the NLAAS instrument: Integrating etic and

emic in the development of cross-cultural measures for a psychiatric

epidemiology and services study of Latinos. International Journal of Methods in

Psychiatric Research, 13, 270–288.

Armitage, C. J., & Conner, M. (2001). Efficacy of the theory of planned behaviour: A

meta‐analytic review. British Journal of Social Psychology, 40, 471–499.

Andres, L. (2012). Designing and doing survey research. Thousand Oaks, CA: Sage

Publications Inc.

Baer, J. C., & Schmitz, M. F. (2007). Ethnic differences in trajectories of family cohesion

for Mexican American and Non-Hispanic White adolescents. Journal of Youth

and Adolescence, 36, 583–592.

Bandura, A. (1991). Social cognitive theory of self-regulation. Organizational Behavior

and Human Decision Processes, 50, 248–287.

Barber, B. K., & Buehler, C. (1996). Family cohesion and enmeshment: Different

constructs, different effects. Journal of Marriage and the Family, 58, 433–441.

Bartholomew, L. K., Parcel, G., Kok, G., & Gottlieb, N. H. (2006). Behavior-oriented

theories used in health promotion. In J. Allegrante & K. McLeroy (Eds.),

Planning health promotion programs (pp. 81–135). San Francisco: Jossey-Bass.

104

Berry, J. W. (2005). Acculturation: Living successfully in two cultures. International

Journal of Intercultural Relations, 29, 697–712.

Bischoff, K. E., Sudore, R., Miao, Y., Boscardin, W. J., & Smith, A. K. (2013). Advance

care planning and the quality of end-of-life care among older adults. Journal of

the American Geriatrics Society, 61, 209–214. doi:10.1111/jgs.12105

Bito, S., Matsumura, S., Singer, M. K., Meredith, L. S., Fukuhara, S., & Wenger, N. S.

(2007). Acculturation and end-of-life decision making: Comparison of Japanese

and Japanese American focus groups. Bioethics, 21, 251–262.

Blackford, J., & Street, A. F. (2016). Advance care planning. In E. Wittenberg, B. Ferrell,

J. Goldsmith, T. Smith, M. Glajchen, G. Handzo, & S. L. Ragan (Eds.), Textbook

of palliative care communication (pp. 285–293). New York, NY: Oxford

University Press.

Boerner, K., Carr, D., & Moorman, S. (2013). Family relationships and advance care

planning: Do supportive and critical relations encourage or hinder planning? The

Journals of Gerontology: Series B, 68, 246–256. doi:10.1093/geronb/gbs161

Brinkman-Stoppelenburg, A., Rietjens, J. A., & van der Heide, A. (2014). The effects of

advance care planning on end-of-life care: A systematic review. Palliative

Medicine, 28, 1000–1025.

Brown, B. A. (2003). The history of advance directives a literature review. Journal of

Gerontological Nursing, 29, 4–9. doi:10.3928/0098-9134-20030901-04

Bryman, A. (2004). Pairwise deletion. In M. S. Lewis-Beck, A. Bryman, & T. F. Liao

(Eds.), The SAGE encyclopedia of social science research methods (pp. 782).

Thousand Oaks, CA: SAGE Publications Inc. doi:10.4135/9781412950589.n682

Carpenter, C. J. (2010). A meta-analysis of the effectiveness of health belief model

variables in predicting behavior. Health Communication, 25, 661–669.

Centers for Medicare & Medicaid Services. (2016). Advance care planning. Retrieved

from https://www.cms.gov/Outreach-and-Education/Medicare-Learning-

Network-MLN/MLNProducts/Downloads/AdvanceCarePlanning.pdf

Champion, V. L., & Skinner, C. S. (2008). The health belief model. K. Glanz, B. K.

Rimer, & K. Viswanath (Eds.), Health behavior and health education: Theory,

research, and practice (4th ed., pp. 45–65). San Francisco, CA: Jossey-Bass.

Chan, C. K. L., & Yau, M. K. (2009). Death preparation among the ethnic Chinese well-

elderly in Singapore: An exploratory study. Journal of Death and Dying, 60, 225–

239. doi:10.2190/OM.60.3.b

105

Chan, H. Y., & Pang, S. M. (2010). Let me talk – an advance care planning programme

for frail nursing home residents. Journal of Clinical Nursing, 19, 3073–3084. doi:

10.1111/j.1365-2702.2010.03353.x

Cheng, J. O.-Y., Lo, R., Chan, F., & Woo, J. (2010). A pilot study on the effectiveness of

anticipatory grief therapy for elderly facing the end of life. Journal of Palliative

Care, 26, 261–269.

Chu, L., Luk, J. K., Hui, E., Chiu, P. K., Chan, C. S., Kwan, F.,…Woo, J. (2011).

Advance directive and end-of-life care preferences among Chinese nursing home

residents in Hong Kong. Journal of the American Medical Directors Association,

12, 143–152. doi:10.1016/j.jamda.2010.08.015

Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple

regression/correlation analysis for the behavioral sciences. Mahwah, New Jersey:

Lawrence Erlbaum Associates, Inc.

Colby, S. L., & Ortman, J. M. (2015). Projections of the size and composition of

the U.S. population: 2014 to 2060. Retrieved from

https://census.gov/content/dam/Census/library/publications/2015/demo/p25-

1143.pdf

Cooke, R., Dahdah, M., Norman, P., & French, D. P. (2014). How well does the theory of

planned behaviour predict alcohol consumption? A systematic review and meta-

analysis. Health Psychology Review, 10, 148–167.

Cooke, R., & French, D. P. (2008). How well do the theory of reasoned action and theory

of planned behaviour predict intentions and attendance at screening programmes?

A meta-analysis. Psychology and Health, 23, 745–765.

De Bary, W. T. (1998). Asian values and human rights: A Confucian communitarian

perspective. Cambridge, MA: Harvard University Press.

De Vleminck, A., Houttekier, D., Pardon, K., Deschepper, R., Van Audenhove, C.,

Vander Stichele, R., & Deliens, L. (2013). Barriers and facilitators for general

practitioners to engage in advance care planning: A systematic review.

Scandinavian Journal of Primary Health Care, 31, 215–226.

Detering, K. M., Hancock, A. D., Reade, M. C., & Silvester, W. (2010). The impact of

advance care planning on end of life care in elderly patients: Randomised

controlled trial. BMJ, 340. doi:10.1136/bmj.c1345

Dillon, F. R., De La Rosa, M., & Ibañez, G. E. (2013). Acculturative stress and

diminishing family cohesion among recent Latino immigrants. Journal of

Immigrant and Minority Health, 15, 484–491.

106

Dong, X., Bergren, S. M., & Chang, E. S. (2015). Levels of acculturation of Chinese

older adults in the greater Chicago area—the population study of Chinese elderly

in Chicago. Journal of the American Geriatrics Society, 63, 1931–1937.

doi: 10.1111/jgs.13604

Downs, D. S., & Hausenblas, H. A. (2005). The theories of reasoned action and planned

behavior applied to exercise: A meta-analytic update. Journal of Physical Activity

and Health, 2, 76–97.

Durbin, C. R., Fish, A. F., Bachman, J. A., & Smith, K. V. (2010). Systematic review of

educational interventions for improving advance directive completion. Journal of

Nursing Scholarship, 42, 234–241. doi: 10.1111/j.1547-5069.2010.01357.x

Enders, C. K. (2010). Applied missing data analysis. New York, NY: Guilford Press.

Enguidanos, S., & Ailshire, J. (2017). Timing of advance directive completion and

relationship to care preferences. Journal of Pain and Symptom Management, 53,

49–56. doi: 10.1016/j.jpainsymman.2016.08.008.

Enguidanos, S., Yonashiro-Cho, J., & Cote, S. (2013). Knowledge and perceptions of

hospice care of Chinese older adults. Journal of the American Geriatrics Society,

61, 993–998.

Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behavior: An introduction

to theory and research. Reading, MA: Addison-Wesley.

Fishbein, M., & Ajzen, I. (2010). Predicting and changing behavior: The reasoned action

approach. New York, NY: Psychology Press.

Fowler, F. J. (2014). Survey research methods (5th ed.). Thousand Oaks, CA: Sage

Publications Inc.

Fried, T. R., Bullock, K., Iannone, L., & O’Leary, J. R. (2009). Understanding advance

care planning as a process of health behavior change. Journal of the American

Geriatrics Society, 57, 1547–1555.

Gao, X., Sun, F., Ko, E., Kwak, J., & Shen, H. W. (2015). Knowledge of advance

directive and perceptions of end-of-life care in Chinese-American elders: The role

of acculturation. Palliative and Supportive Care, 13, 1677–1684.

Gerend, M. A., & Shepherd, J. E. (2012). Predicting human papillomavirus vaccine

uptake in young adult women: Comparing the health belief model and theory of

planned behavior. Annals of Behavioral Medicine, 44, 171–180.

107

Glanz, K., & Bishop, D. B. (2010). The role of behavioral science theory in development

and implementation of public health interventions. Annual Review of Public

Health, 31, 399–418. doi:10.1146/annurev.publhealth.012809.103604

Greco, P. J., Schulman, K. A., Lavizzo-Mourey, R., & Hansen-Flaschen, J. (1991). The

Patient Self-Determination Act and the future of advance directives. Annals of

Internal Medicine, 115, 639–643.

Green, J. (2000). The role of theory in evidence-based health promotion practice. Health

Education Research, 15, 125–129. doi:10.1093/her/15.2.125

Guo, M., Li, S., Liu, J., & Sun, F. (2015). Family relations, social connections, and

mental health among Latino and Asian older adults. Research on Aging, 37, 123–

147.

Gupta, R., & Yick, A.G. (2001). Preliminary validation of the acculturation scale on

Chinese Americans. Journal of Social Work Research and Evaluation, 2, 43–56.

Hamel, C. F., Guse, L. W., Hawranik, P. G., & Bond, J. B., Jr. (2002). Advance

directives and community-dwelling older adults. Western Journal of Nursing

Research, 24, 143–158.

Hanel, P. H. P., & Vione, K. C. (2016). Do student samples provide an accurate estimate

of the general public? PLOS ONE, 11, e0168354.

doi:10.1371/journal.pone.0168354

Harrison, J. A., Mullen, P. D., & Green, L. W. (1992). A meta-analysis of studies of the

health belief model with adults. Health Education Research, 7, 107–116.

Heyland, D. K., Barwich, D., Pichora, D., Dodek, P., Lamontagne, F., You, J.

J.,…Simon, J. (2013). Failure to engage hospitalized elderly patients and their

families in advance care planning. JAMA Internal Medicine, 173, 778–787.

doi:10.1001/jamainternmed.2013.180

Ho, A. H. Y., Dai, A. A. N., Shu-hang, L., Wong, S. W. R., Tsui, A. L. M., Tang, J. C. S.,

& Lou, V. W. Q. (2016). Development and pilot evaluation of a novel dignity-

conserving End-of-Life (EoL) care model for nursing homes in Chinese societies.

Gerontologist, 56, 578–589. doi:10.1093/geront/gnv037

Hong, M., Casado, B. L., & Lee, S. E. (2018). The intention to discuss advance care

planning in the context of Alzheimer’s Disease among Korean Americans. The

Gerontologist. Advance online publication. doi:10.1093/geront/gnx211

108

Houben, C. H., Spruit, M. A., Groenen, M. T., Wouters, E. F., & Janssen, D. J. (2014).

Efficacy of advance care planning: A systematic review and meta-

analysis. Journal of the American Medical Directors Association, 15, 477–489.

Hsiung, Y. (2015). Theoretical issues associated with cultural readiness for end-of-life

care planning in Taiwan. Social Behavior & Personality, 43, 1385–1393.

doi:10.2224/sbp.2015.43.8.1385

Hsiung, Y. F. (2011). Chinese American’s readiness for advance care planning.

(Doctoral dissertation). Available from ProQuest Dissertations and Theses

database. (UMI No. 3551255)

Hsiung, Y. Y., & Ferrans, C. E. (2007). Recognizing Chinese Americans’ cultural needs

in making end-of-life treatment decisions. Journal of Hospice & Palliative

Nursing, 9, 132–140. doi:10.1097/01.NJH.0000269993.13625.49

Huang, I. A., Neuhaus, J. M., & Chiong, W. (2016). Racial and ethnic differences in

advance directive possession: Role of demographic factors, religious affiliation,

and personal health values in a national survey of older adults. Journal of

Palliative Medicine, 19, 149–156. doi:10.1089/jpm.2015.0326

Janz, N. K., & Becker, M. H. (1984). The health belief model: A decade later. Health

Education Quarterly, 11, 1–47.

Kang, S. C., Lin, M. H., Hwang, I. H., Lin, M. H., Chang, H. T., & Hwang, S. J. (2012).

Impact of hospice care on end-of-life hospitalization of elderly patients with lung

cancer in Taiwan. Journal of the Chinese Medical Association, 75, 221–226.

Kataoka-Yahiro, M. R., Yancura, L. A., Page, V., & Inouye, J. (2011). Advance care

planning decision making among Asian Pacific Islander family caregivers of stage

4 to 5 chronic kidney disease patients on hemodialysis: A focus group study.

Journal of Hospice & Palliative Nursing, 13, 426–435.

doi:10.1097/NJH.0b013e31822da9c2

Kent, D. R. (1996). Application of health behavior theory to increase use of advance

directives among individuals with HIV disease. (Doctoral dissertation). Available

from ProQuest Dissertations and Theses database. (UMI No. 9703840)

Ko, E. (2008). Advance care planning with Korean American and non-Hispanic White

older adults. (Doctoral dissertation). Available from ProQuest Dissertations and

Theses database. (UMI No. 3297917)

Ko, E., & Lee, J. (2009). End-of-life communication: Ethnic differences between Korean

American and non-Hispanic White older adults. Journal of Aging and Health, 21,

967–984. doi:10.1177/0898264309344179

109

Kononovas, K., & McGee, A. (2017). The benefits and barriers of ensuring patients have

advance care planning. Nursing Times, 113, 41–44.

Kramer, B. J., Boelk, A. Z., & Auer, C. (2006). Family conflict at the end of life: Lessons

learned in a model program for vulnerable older adults. Journal of Palliative

Medicine, 9, 791–801.

Lee, H. T. S., Cheng, S. C., Dai, Y. T., Chang, M., & Hu, W. Y. (2016). Cultural

perspectives of older nursing home residents regarding signing their own DNR

directives in Eastern Taiwan: A qualitative pilot study. BMC Palliative Care, 15,

1–10. doi:10.1186/s12904-016-0117-4

Lee, M. C., Byon, H., Hinderer, K., & Alexander, C. (2017). Beliefs in advance care

planning among Chinese Americans: Similarities and differences between the

younger and older generations. Asian/Pacific Island Nursing Journal, 2, 83–90.

doi:10.9741/23736658.1055

Lee, M. C., Hinderer, K. A., & Friedmann, E. (2015). Engaging Chinese American adults

in advance care planning: A community-based, culturally sensitive

seminar. Journal of Gerontological Nursing, 41, 17–21. doi:10.3928/00989134-

20150406-01

Lee, M. C., Hinderer, K. A., & Kehl, K. A. (2014). A systematic review of advance

directives and advance care planning in Chinese people from Eastern and Western

cultures. Journal of Hospice & Palliative Nursing, 16, 75–85.

Lindlof, T. R., & Taylor, B. C. (2011). Qualitative communication research methods (3rd

ed.). Thousand Oaks, CA: SAGE Publications Inc.

Low, J. A., Ng, W. C., Yap, K. B., & Chan, K. M. (2000). End-of-life issues: Preferences

and choices of a group of elderly Chinese subjects attending a day care center in

Singapore. Annals of the Academy of Medicine, Singapore, 29, 50–56.

Matsumura, S., Bito, S., Liu, H., Kahn, K., Fukuhara, S., Kagawa-Singer, M., &

Wenger, N. (2002). Acculturation of attitudes toward end-of-life care. Journal of

General Internal Medicine, 17, 531–539. doi:10.1046/j.1525-1497.2002.10734.x

McClenahan, C., Shevlin, M., Adamson, G., Bennett, C., & O’Neill, B. (2007). Testicular

self-examination: A test of the health belief model and the theory of planned

behaviour. Health Education Research, 22, 272–284. doi:10.1093/her/cyl076

McEachan, R. R. C., Conner, M., Taylor, N. J., & Lawton, R. J. (2011). Prospective

prediction of health-related behaviours with the theory of planned behaviour: A

meta-analysis. Health Psychology Review, 5, 97–144.

110

Michie, S., & Johnston, M. (2012). Theories and techniques of behaviour change:

Developing a cumulative science of behaviour change. Health Psychology

Review, 6, 1–6. doi:10.1080/17437199.2012.654964

Miller, B. (2017). Nurses in the know: The history and future of advance directives. The

Online Journal of Issues in Nursing, 22(3). doi:10.3912/OJIN.Vol22No03PPT57

Montanaro, E. A., & Bryan, A. D. (2014). Comparing theory-based condom

interventions: Health belief model versus theory of planned behavior. Health

Psychology, 33, 1251–1260. doi:10.1037/a0033969

Morrison, R. S., & Meier, D. E. (2004). High rates of advance care planning in New York

City’s elderly population. Archives of Internal Medicine, 164, 2421–2426.

doi: 10.1001/archinte.164.22.2421

Nahapetyan, L., Orpinas, P., Glass, A., & Song, X. (2017). Planning ahead: Using the

theory of planned behavior to predict older adults’ intentions to use hospice if

faced with terminal illness. Journal of Applied Gerontology. Advance online

publication. doi:10.1177/0733464817690678

National Hospice and Palliative Care Organization. (2016). Advance care planning.

Retrieved from http://www.nhpco.org/advance-care-planning

National Institute of Aging. (2016). Advance care planning. Retrieved from

https://www.nia.nih.gov/health/publication/advance-care-planning

Ni, P., Zhou, J., Wang, Z. X., Nie, R., Phillips, J., & Mao, J. (2014). Advance directive

and end-of-life care preferences among nursing home residents in Wuhan, China:

A cross-sectional study. Journal of the American Medical Directors Association,

15, 751–756. doi:10.1016/j.jamda.2014.06.006

Noar, S. M., Benac, C. N., & Harris, M. S. (2007). Does tailoring matter? Meta-analytic

review of tailored print health behavior change interventions. Psychological

Bulletin, 133, 673–693. doi: 10.1037/0033-2909.133.4.673

Notani, A. S. (1998). Moderators of perceived behavioral control’s predictiveness in the

theory of planned behavior: A meta‐analysis. Journal of Consumer Psychology, 7,

247–271.

Nussbaum, J. F., & Fisher, C. L. (2009). A communication model for the competent

delivery of geriatric medicine. Journal of Language and Social Psychology, 28,

190–208.

https://www.nia.nih.gov/health/publication/advance-care-planning

111

O’Connor, B. P. (2000). SPSS and SAS programs for determining the number of

components using parallel analysis and Velicer’s MAP test. Behavior Research

Methods, Instrumentation, and Computers, 32, 396–402.

Ortman, J. M., Velkoff, V. A., & Hogan, H. (2014). An aging nation: The older

populationin the United States. Retrieved from

https://www.census.gov/prod/2014pubs/p25-1140.pdf

Parkin, D.M., Boyd, L., & Walker, L.C. (2011). The fraction of cancer attributable to

lifestyle and environmental factors in the UK in 2010. British Journal of Cancer,

105, S77–S81. doi:10.1038/bjc.2011.474

PerryUndem Research/Communication. (2016). Physicians’ views toward advance care

planning and end-of-life care conversations. Retrieved from

https://www.johnahartford.org/images/uploads/resources/ConversationStopper_P

oll_Memo.pdf

Peterson, R. A., & Merunka, D. R. (2014). Convenience samples of college students and

research reproducibility. Journal of Business Research, 67, 1035–1041.

Pett, M. A., Lackey, N. R., & Sullivan, J. J. (2003). Making sense of factor analysis.

Thousand Oaks, CA: Sage Publications Inc.

Pew Research Center. (2017). Chinese in the U.S. Fact Sheet.

Retrieved from http://www.pewsocialtrends.org/fact-sheet/asian-americans-

chinese-in-the-u-s/

Pope, T. M. (2013). Legal briefing: The new Patient Self-Determination Act. Journal of

Clinical Ethics, 24, 156–167.

Rao, J. K., Anderson, L. A., Lin, F.-C., & Laux, J. P. (2014). Completion of advance

directives among U.S. consumers. American Journal of Preventive Medicine, 46,

65–70.

Riebl, S. K., Estabrooks, P. A., Dunsmore, J. C., Savla, J., Frisard, M. I., Dietrich, A. M.,

…Davy, B. M. (2015). A systematic literature review and meta-analysis: The

theory of planned behavior’s application to understand and predict nutrition-

related behaviors in youth. Eating Behaviors, 18, 160–178.

Rivera, F. I., Guarnaccia, P. J., Mulvaney-Day, N., Lin, J. Y., Torres, M., & Alegria, M.

(2008). Family cohesion and its relationship to psychological distress among

Latino groups. Hispanic Journal of Behavioral Sciences, 30, 357–378.

doi:10.1177/0739986308318713

https://www.johnahartford.org/images/uploads/resources/ConversationStopper_

http://www.pewsocialtrends.org/fact-sheet/asian-americans-chinese-in-the-u-s/

112

Rosenstock, I. M. (1974). Historical origins of the health belief model. Health

Education Monographs, 2, 328–335.

Rosenstock, I. M., Strecher, V. J., & Becker, M. H. (1988). Social learning theory and the

health belief model. Health Education & Behavior, 15, 175–183.

Sabatino, C. P. (2010). The evolution of health care advance planning law and policy.

The Milbank Quarterly, 88, 211–239. doi:10.1111/j.1468-0009.2010.00596.x

Sales, A., Smith, J., Curran, G., & Kochevar, L. (2006). Models, strategies, and tools:

Theory in implementing evidence-based findings into health care practice.

Journal of General Internal Medicine, 21(Suppl 2), S43–S49. doi:10.1111/j.1525-

1497.2006.00362.x

Schwartz, S. J., Unger, J. B., Zamboanga, B. L., & Szapocznik, J. (2010). Rethinking the

concept of acculturation: Implications for theory and research. The American

Psychologist, 65, 237–251. doi:10.1037/a0019330

Sharp, T., Moran, E., Kuhn, I., & Barclay, S. (2013). Do the elderly have a voice?

Advance care planning discussions with frail and older individuals: A systematic

literature review and narrative synthesis. British Journal of General Practice, 63,

e657–e668. doi:10.3399/bjgp13X673667

Shrestha, L. B., & Heisler, E. J. (2011). The changing demographic profile of the United

States. Retrieved from https://www.fas.org/sgp/crs/misc/RL32701.pdf.

Şimşekoğlu, Ö., & Lajunen, T. (2008). Social psychology of seat belt use: A comparison

of theory of planned behavior and health belief model. Transportation Research

Part F: Traffic Psychology and Behaviour, 11, 181–191.

Stiff, J. B., & Mongeau, P. A. (2016). Persuasive communication (3rd ed.). New York,

NY: Guilford Press.

Sudore, R. L., & Fried, T. R. (2010). Redefining the “planning” in advance care planning:

Preparing for end-of-life decision making. Annals of Internal Medicine, 153, 256–

261.

Sun, F., Gao, X., & Coon, D. W. (2015). Perceived threat of Alzheimer’s Disease among

Chinese American older adults: The role of Alzheimer’s Disease literacy. The

Journals of Gerontology: Series B, 70, 245–255. doi:10.1093/geronb/gbt095

Szalai, L. C. (2015). Predicting young adults’ engagement in advance care planning.

(Doctoral dissertation). Available from ProQuest Dissertations and Theses

database. (UMI No. 10024069)

113

Ta, V. M., Holck, P., & Gee, G. C. (2010). Generational status and family cohesion

effects on the receipt of mental health services among Asian Americans: Findings

from the national Latino and Asian American study. American Journal of Public

Health, 100, 115–121. doi:10.2105/AJPH.2009.160762

Tang, C., Lam, L., & Chiu, H. (2007). Attitudes to end-of-life decisions: A survey of

elderly Chinese with dementia and their carers. Asian Journal of Gerontology &

Geriatrics, 2, 119–125.

Ting, F. H., & Mok, E. (2011). Advance directives and life-sustaining treatment:

Attitudes of Hong Kong Chinese elders with chronic disease. Hong Kong Medical

Journal, 17, 105–111.

Topa, G., & Moriano, J. A. (2010). Theory of planned behavior and smoking: Meta-

analysis and SEM model. Substance Abuse and Rehabilitation, 1, 23–33.

Thornton, S., & Calam, R. (2011). Predicting intention to attend and actual attendance at

a universal parent-training programme: A comparison of social cognition models.

Clinical Child Psychology and Psychiatry, 16, 365–383.

doi:10.1177/1359104510366278

U.S. Census Bureau. (2012, March 12). 2010 Census shows Asians are fastest-growing

race group [Press release]. Retrieved from

https://www.census.gov/newsroom/releases/archives/2010_census/cb12-

cn22.html

U.S. Census Bureau. (2016). Asian alone or in any combination by selected groups.

Retrieved from

https://factfinder.census.gov/faces/tableservices/jsf/pages/productview.xhtml?src=

bkmk

U.S. Census Bureau. (2017a). Asian-American and Pacific Islander Heritage

Month: May 2017. Retrieved from

https://www.census.gov/content/dam/Census/newsroom/facts-for-

features/2017/cb17-ff07.pdf

U.S. Census Bureau. (2017b). Older Americans Month: May 2017. Retrieved from

https://www.census.gov/newsroom/facts-for-features/2017/cb17-ff08.html

U.S. Census Bureau. (2017c). The nation’s older population is still growing, Census

Bureau reports [Press release]. Retrieved from

https://www.census.gov/newsroom/press-releases/2017/cb17-100.html

114

VandeCreek, L., & Frankowski, D. (1996). Barriers that predict resistance to

completing a living will. Death Studies, 20, 73–82.

doi:10.1080/07481189608253412

Volandes, A. E., & Volandes, A. (2015). The conversation: A revolutionary plan for end-

of-life care. New York: Bloomsbury Publishing USA.

Wilkinson, A., Wenger, N., & Shugarman, L. R. (2007). Literature review on advance

directives. United States: Department of Health and Human Services and RAND

Corporation.

Witte, K. (2000). EPPM: Examples of items. Retrieved from http://www.

msu.edu/~wittek/scale.htm

Wittenberg-Lyles, E., Villagran, M. M., & Hajek, C. (2008). The impact of

communication, attitudes, and acculturation on advance directives decision-

making. Journal of Ethnic & Cultural Diversity in Social Work, 17, 349–364.

doi:10.1080/15313200802467874

World Health Organization. (2000). Information needs for research, policy and action on

ageing and older adults. Retrieved from

http://www.who.int/healthinfo/survey/ageing_mds_report_en_harare.pdf

Wright, A. A., Zhang, B., Ray, A., Mack, J. W., Trice, E., Balboni, T.,...& Prigerson, H.

G. (2008). Associations between end-of-life discussions, patient mental health,

medical care near death, and caregiver bereavement adjustment. Jama, 300,

1665–1673. doi:10.1001/jama.300.14.1665

Yang, Z. J. (2015). Predicting young adults’ intentions to get the H1N1 vaccine: an

integrated model. Journal of Health Communication, 20, 69–79.

doi:10.1080/10810730.2014.904023

Yap, S. S., Chen, K., Detering, K. M., & Fraser, S. A. (2017). Exploring the knowledge,

attitudes and needs of advance care planning in older Chinese-Australians.

Journal of Clinical Nursing. Advance online publication. doi:10.1111/jocn.13886

Yzer, M. (2013). Reasoned action theory: Persuasion as belief-based behavior change. In

J. P. Dillard, & L. J. Shen (Eds.), The SAGE handbook of persuasion:

Developments in theory and practice (pp. 120-136). Thousand Oaks, CA: Sage

Publications Inc.

Zhang, N., Ning, X. H., Zhu, M. L., Liu, X. H., Li, J. B., & Liu, Q. (2015). Attitudes

towards advance care planning and healthcare autonomy among community-

dwelling older adults in Beijing, China. BioMed Research International, 2015, 1-

10. doi:10.1155/2015/453932

115

Zhang, B., Wright, A. A., Huskamp, H. A., Nilsson, M. E., Maciejewski, M. L., Earle, C.

C.,…& Prigerson, H. G. (2009). Health care costs in the last week of life:

associations with end-of-life conversations. Archives of Internal Medicine, 169,

480-488. doi:10.1001/archinternmed.2008.587

Zimmerman, R. S., & Vernberg, D. (1994). Models of preventative health behavior:

Comparison, critique, and meta-analysis. Advances in Medical Sociology, 4, 45–

67.

116

APPENDIX I

IRB APPROVAL

117

EXEMPTION GRANTED

Anthony Roberto

Human Communication, Hugh Downs School of

480/965-4111 [email protected]

Dear Anthony Roberto:

On 3/31/2017 the ASU IRB reviewed the following protocol:

Type of Review: Initial Study

Title: Understanding Chinese American Older Adults'

Intention to Use Advance Care Planning

Investigator: Anthony Roberto

IRB ID: STUDY00006081

Funding: None

Grant Title: None

Grant ID: None

Documents Reviewed: • Survey Questions Chinese, Category: Translations;

• Translation Certificate Form, Category:

Translations;

• PROTOCOL Social Behavioral 0329.docx,

Category: IRB Protocol;

• survey recruitment revised.pdf, Category:

Recruitment Materials;

• CONSENT SOCIAL BEHAVIORAL revised.pdf,

Category: Consent Form;

• Survey Questions English, Category: Measures

(Survey questions/Interview questions /interview

guides/focus group questions);

The IRB determined that the protocol is considered exempt pursuant to Federal

Regulations 45CFR46 (2) Tests, surveys, interviews, or observation on 3/31/2017.

In conducting this protocol you are required to follow the requirements listed in the

INVESTIGATOR MANUAL (HRP-103).

Sincerely,

mailto:[email protected]

https://era.oked.asu.edu/IRB/Personalization/MyProfile?Person=com.webridge.account.Person%5BOID%5B99C2373498FCF440895ED285155398E6%5D%5D

118

IRB Administrator

cc: Yanqin Liu

119

APPENDIX II

THEORY OF PLANNED BEHAVIOR & HEALTH BELIEF MODEL QUESTIONS

ENGLISH VERSION 1

120

Advance care planning helps us before we encounter a medical crisis and are not able

to speak for ourselves. Advance care planning includes:

1. Discussing your end of life care preferences with family members, and

2. Documenting your end of life care preferences.

It is recommended that you do these two things while you are still physically and

mentally able to do so.

This part of the survey focuses on discussing your end of life care plans with family

members. Please think about just this part of advance care planning when answering

the questions in this section.

Please circle the word that best describes your opinion. It is measured on a scale of 1-5.

Attitude

Discussing my end of life care plans with family members is:

1. Very bad-bad-neutral-good-very good

2. Very Harmful-harmful-neutral-beneficial-very beneficial

3. Very useless-useless-neutral-useful-very useful

4. Very worthless-worthless-neutral-valuable-very valuable

Please rate to what degree the following statements are true to you. 1=completely

disagree, 2 =mostly disagree, 3=neither agree nor disagree, 4=mostly agree,

5=completely agree.

Subjective norm

1. Most people who are important to me think I should discuss my end of life care plans

with family members.

2. Most people whose opinions I value would approve of my discussing end of life care

plans with family members.

3. Most people I respect and admire will support my discussing end of life care plans

with family members.

Self-efficacy

1. I am confident that I can discuss my end of life care plans with family members.

2. I have the ability to discuss my end of life care plans with family members.

3. It would be easy for me to discuss my end of life care plans with family members.

4. I have enough knowledge to be able to discuss my end of life care plans with family

members.

121

Control belief

1. Whether I discuss my end of life care plans with family members is entirely up to me.

2. Whether I discuss my end of life care plans with family members is under my personal

control.

3. I feel in complete control over whether I discuss my end of life care plans with family

members.

Behavioral intention

1. I intend to discuss my end of life care plans with family members.

2. I am willing to discuss my end of life care plans with family members.

3. I will discuss my end of life care plans with family members.

Perceived susceptibility

1. If I do NOT discuss my end of life care plans with family members I will receive

unwanted medical treatments.

2. If I do NOT discuss my end of life care plans with family members I will be at risk of

receiving unwanted medical treatments.

3. If I do NOT discuss my end of life care plans with family members it is likely that I

will receive unwanted medical treatments.

4. If do NOT discuss my end of life care plans with family members I will be susceptible

to unwanted medical treatments.

Perceived severity

1. Receiving unwanted medical treatments would be harmful to me.

2. Receiving unwanted medical treatments is a big concern to me.

3. Receiving unwanted medical treatments is a serious problem to me.

4. Receiving unwanted medical treatments would have severe negative consequences.

Perceived benefit

1. Discussing my end of life care plan with family members will help my family know

about my medical treatment preferences in advance.

2. Discussing my end of life care plan with family members will help my doctor know

about my medical treatment references in advance.

3. Discussing my end of life care plan with family members will help me get the wanted

medical treatments in the future.

4. Discussing my end of life care plan with family members will help me relieve family

burdens.

5. Discussing my end of life care plan with family members will help me reduce family

conflicts.

6. Discussing my end of life care plan with family members will increase the quality of

my life in my last days.

122

Perceived barrier

1. It makes me sad to discuss my end of life care plans with family members.

2. Discussing my end of life care plans with family members will increase my family

conflicts.

3. It is difficult to discuss my end of life care plans because I do not know what my

medical treatment preferences will be in the future.

4. I feel uncomfortable to discuss my end of life care plans with family members.

5. Discussing my end of life care plans with family members will cause my death

anxiety.

6. I am not used to considering my end of life care in advance.

7. I am not sick enough to discuss end of life care plans with family members.

8. Discussing my end of life care plans with my family members is bad luck.

9. It will make my family members sad if I discuss my end of life care plans with them.

123

APPENDIX II

THEORY OF PLANNED BEHAVIOR & HEALTH BELIEF MODEL QUESTIONS

MANDARIN VERSION 1

计划行为理论和健康信念模型问题版本一

124

基本提示：

在美国的医疗系统中，预定护理计划(Advance Care Planning)可以帮助我们在身体

健康、精神状态良好、意识清楚的时候对未来的临终照料做出安排。这样做的目的

是尽可能地保证自己的选择得到家人和医生的尊重，更好地处理医疗危机。这一类

的计划主要包括：

1）与家人讨论临终照料计划

2）填写临终照料计划

这部分的问题关于与家人讨论临终照料计划的行为。如果您已经有过讨论，请设想

自己是否愿意继续和家人就这一问题进行讨论。因为我们的生活处于变化之中，对

于临终照料的计划有可能需要调整。

在这一部分，您会发现很多问题看似重复。这是为了科学研究所做出的刻意安排，

因为一个问题有时并不能全面地呈现出您的看法。请仔细阅读并根据您的真实想法

思考每一个问题，在相应选项处划勾。调查对参与者信息严格保密，请放心回答。

答案无对错之分。

请根据真实想法回答，在相应选项处划勾。五个选项分别代表1-5。

态度

与家人讨论我的临终照料计划：

1）很不好不好中立好很好

2）很有害有害的中立有利很有利

3）很没有帮助没有帮助中立有帮助很有帮助

4）很没有价值没有价值中立有价值很有价值

请根据自己的真实想法选出对下列描述的同意程度。1=完全不同意，2=大部分不

同意，3=一半一半，4=大部分同意，5=完全同意。

主观规范

1. 对我重要的大多数人认为我应该与家人讨论临终照料计划。

2. 我重视意见的大多数人会赞同我与家人讨论临终照料计划。

3. 我尊敬的大多数人会支持我与家人讨论临终照料计划。

自我效能

1. 我有信心和家人讨论我的临终照料计划。

2. 我有能力和家人讨论我的临终照料计划。

3. 对我来说，和家人讨论我的临终照料计划是容易的。

125

4. 我知道自己该怎样和家人讨论我的临终照料计划。

控制信念

1. 是否和家人讨论临终照料计划完全由我自己来决定。

2. 是否和家人讨论临终照料计划在我的控制范围之内。

3. 是否和家人讨论临终照料计划，我一个人说了算。

行为动机

1. 我计划和家人讨论我的临终照料事项。

2. 我愿意和家人讨论我的临终照料计划。

3. 我将来会和家人讨论我的临终照料计划。

风险程度

1. 如果不和家人讨论临终照料计划，我会得到不想要的医疗方案。

2. 如果不和家人讨论临终照料计划，我很有可能会得到不想要的医疗方案。

3. 如果不和家人讨论临终照料计划，我也许会得到不想要的医疗方案。

4. 如果不和家人讨论临终照料计划，不想要的医疗方案容易影响到我。

严重性

1. 得到不想要的医疗方案对我是有害的。

2. 我很担心自己会得到不想要的医疗方案。

3. 对我来说，得到不想要的医疗方案是一个严重的问题。

4. 得到不想要的医疗方案会给我带来严重的后果。

好处

1. 与家人讨论临终照料计划可以帮助家人提前了解我的医疗意愿。

2. 与家人讨论临终照料计划可以帮助医生提前了解我的医疗意愿。

3. 与家人讨论临终照料计划可以帮助我将来得到我想要的医疗方案。

4. 与家人讨论临终照料计划可以帮助减轻家庭负担。

5. 与家人讨论临终照料计划可以帮助缓解家庭矛盾。

6. 与家人讨论临终照料计划可以提高我生命最后阶段的生活质量。

126

困难

1. 与家人讨论我的临终照料计划会让我感到悲伤。

2. 与家人讨论我的临终照料计划会增加家庭矛盾。

3. 与家人讨论我的临终照料计划比较困难，因为我还不知道自己未来的医疗意愿会

是什么。

4. 与家人讨论我的临终照料计划会让我心里不舒服。

5. 与家人讨论我的临终照料计划会引起我对死亡的恐惧。

6. 我不习惯提前考虑和临终照料有关的话题。

7. 我还没有病重到需要与家人讨论临终照料计划的程度。

8. 和家人我的讨论临终照料计划不吉利。

9. 与家人讨论我的临终照料计划会让家人感到悲伤。

127

APPENDIX III

THEORY OF PLANNED BEHAVIOR AND HEALTH BELIEF MODEL QUESTIONS

ENGLISH VERSION 2

128

Advance care planning helps us before we encounter a medical crisis and are not able

to speak for ourselves. Advance care planning includes:

1. Discussing your end of life care preferences with family members, and

2. Documenting your end of life care preferences.

It is recommended that you do these two things while you are still physically and

mentally able to do so.

This part of the survey focuses on documenting your end of life care plans.

Documenting your end of life care plan is done using a legal document called an

advance directive (AD). When you encounter a medical crisis and are not able to speak

for yourself, AD mainly provides guidance about the following things:

1. What life-supporting treatments (e.g., CPR and a breathing machine) you do not want.

2. Who is appointed to make medical decisions on your behalf.

Think about just this part of advanced care planning when answering the questions in

this section. Please keep in mind that we may ask similar questions multiple times,

since no one item will assess your understanding perfectly.

Please circle the word that best describes your opinion. It is measured on a scale of 1-5.

Attitude

For me, completing an advance directive is:

1. Very bad-bad-neutral-good-very good

2. Very Harmful-harmful-neutral-beneficial-very beneficial

3. Very useless-useless-neutral-useful-very useful

4. Very worthless-worthless-neutral-valuable-very valuable

Please rate to what degree the following statements are true to you. 1=completely

disagree, 2=mostly disagree, 3=neither agree nor disagree, 4=mostly agree, 5=completely

agree.

Subjective norm

1. Most people who are important to me think that I should complete an advance

directive.

2. Most people whose opinions I value would approve of my completing an advance

directive.

3. Most people I respect and admire will support my completing an advance directive.

129

Self-efficacy

1. I am confident that I can complete an advance directive.

2. I have the ability to complete an advance directive.

3. It would be easy for me to complete an advance directive.

4. I have enough knowledge to be able to complete an advance directive.

Control belief

1. Whether I complete an advance directive is entirely up to me.

2. Whether I complete an advance directive is under my personal control.

3. I feel in complete control over whether I complete an advance directive.

Behavioral intention

1. I intend to complete an advance directive.

2. I am willing to complete an advance directive.

3. I will complete an advance directive.

Perceived susceptibility

1. If I do NOT complete an advance directive I will receive unwanted medical treatments.

2. If I do NOT complete an advance directive I am at risk of receiving unwanted medical

treatments.

3. If I do NOT complete an advance directive it is likely that I will receive unwanted

medical treatments.

4. If do NOT complete an advance directive I will be susceptible to unwanted medical

treatments.

Perceived benefit

1. Completing an advance directive will help my family know about my medical

treatment preferences in advance.

2. Completing an advance directive will help my doctor know about my medical

treatment references in advance.

3. Completing an advance directive will help me get the wanted medical treatments in the

future.

4. Completing an advance directive will help me relieve family burdens.

5. Completing an advance directive will help me reduce family conflicts.

6. Completing an advance directive will increase the quality of my life in my last days.

130

Perceived barrier

1. It makes me sad to complete an advance directive.

2. Completing an advance directive will increase my family conflicts.

3. It is difficult to complete an advance directive because I do not know what my medical

treatment preferences will be in the future.

4. I feel uncomfortable to complete an advance directive.

5. Completing an advance directive will cause my death anxiety.

6. I am not used to considering my end of life care in advance.

7. I am not sick enough to complete an advance directive.

8. Completing an advance directive is bad luck.

9. It will make my family members sad if I complete an advance directive.

131

APPENDIX III

THEORY OF PLANNED BEHAVIOR AND HEALTH BELIEF MODEL QUESTIONS

MANDARIN VERSION 2

计划行为理论和健康信念模型问题版本二

132

基本提示：

在美国的医疗系统中，预定护理计划(Advance Care Planning)可以帮助我们在身体

健康、精神状态良好、意识清楚的时候对未来的临终照料做出安排。这样做的目的

是尽可能地保证自己的选择得到家人和医生的尊重，更好地处理医疗危机。这一类

的计划主要包括：

3）与家人讨论临终照料计划

4）填写临终照料计划

在美国的医疗系统中，书面填写临终照料计划是通过签署医疗照料委托书 (advance

directive) 实现的。医疗照料委托书是一份具有法律效力的文件，帮助我们在遇到

医疗危机、没有能力做决定时表达重要的医疗意愿。它主要包括：

1）指示是否需要维持生命治疗手段（比如使用心肺复苏、呼吸机等治疗方

式）。

2）指定当我不能做决定时能够代替我做医疗决定的代理人。

虽然临终照料计划的文件主要包括两方面的内容。但是在这一部分的问卷调查中，

请您把“书面填写临终照料计划”作为一个整体来考虑。

这部分的问题关于书面填写临终照料计划的行为。如果您已经填写，请设想自己是

否愿意以后书面更新自己的临终照料事项。因为我们的生活处于变化之中，对于临

终照料的计划有可能需要调整。

在这一部分，您会发现很多问题看似重复。这是为了科学研究所做出的刻意安排，

因为一个问题有时并不能全面地呈现出您的看法。请仔细阅读并根据您的真实想法

思考每一个问题，在相应选项处划勾。调查对参与者信息严格保密，请放心回答。

答案无对错之分。

请根据真实想法回答，在相应选项处划勾。五个选项分别代表1-5。

态度

对我来说，填写临终照料计划

1. 很不好不好中立好很好

2. 很有害有害的中立有利很有利

3. 很没有帮助没有帮助中立有帮助很有帮助

4. 很没有价值没有价值中立有价值很有价值

133

请根据自己的真实想法选出对下列描述的同意程度。1=完全不同意，2=大部分不

同意，3=一半一半，4=大部分同意，5=完全同意。

主观规范

1. 对我重要的大多数人认为我应该填写临终照料计划。

2. 我重视意见的大多数人会赞同我填写临终照料计划。

3. 我尊敬的大多数人会支持我填写临终照料计划。

自我效能

1. 我对填写自己的临终照料计划有信心。

2. 我有能力填写临终照料计划。

3. 对我来说，填写临终照料计划是容易的。

4. 我知道自己该怎样填写临终照料计划。

控制信念

1. 是否填写临终照料计划完全由我自己来决定。

2. 是否填写临终照料计划在我的控制范围之内。

3. 是否填写临终照料计划，我一个人说了算。

行为动机

1. 我计划填写自己的临终照料事项。

2. 我愿意填写临终照料计划。

3. 我将来会填写临终照料计划。

风险程度

1. 如果不填写临终照料计划，我会得到不想要的医疗方案。

2. 如果不填写临终照料计划，我很有可能会得到不想要的医疗方案。

3. 如果不填写临终照料计划，我也许会得到不想要的医疗方案。

4. 如果不填写临终照料计划，不想要的医疗方案容易影响到我。

严重性

1. 得到不想要的医疗方案对我是有害的。

2. 我很担心自己会得到不想要的医疗方案。

3. 对我来说，得到不想要的医疗方案是一个严重的问题。

4. 得到不想要的医疗方案会给我带来严重的后果。

134

好处

1. 填写临终照料计划可以帮助家人提前了解我的医疗意愿。

2. 填写临终照料计划可以帮助医生提前了解我的医疗意愿。

3. 填写临终照料计划可以帮助我将来得到我想要的医疗方案。

4. 填写临终照料计划可以帮助减轻家庭负担。

5. 填写临终照料计划可以帮助缓解家庭矛盾。

6. 填写临终照料计划可以提高我生命最后阶段的生活质量。

困难

1. 填写临终照料计划会让我感到悲伤。

2. 填写临终照料计划会增加家庭矛盾。

3. 填写临终照料计划比较困难，因为我还不知道自己未来的医疗意愿会是什么。

4. 填写临终照料计划会让我心里不舒服。

5. 填写临终照料计划会引起我对死亡的恐惧。

6. 我不习惯提前考虑和临终照料有关的话题。

7. 我还没有病重到需要填写临终照料计划的程度。

8. 填写临终照料计划不吉利。

9. 填写临终照料计划会让家人感到悲伤。

135

APPENDIX IV

SURVEY QUESTIONS ABOUT DEMOGRAPHICS, ACCULTURATION, AND

FAMILY COHESION

ENGLISH VERSION

136

Are you 1. Male 2. Female

What is your date of birth? __ __ (month) __ __ __ __ (year)

What is the highest grade or year of regular school that you have completed?

1 6th grade or lower

2 9th grade

3 12th grade

4 Vocational or trade school

5 College graduate – 4 yr.

6 Postgraduate or higher

In which territory were you born?

1 USA (If you have chosen this one, skip question 2, and go to question 3)

2 Mainland China

3 Hong Kong

4 Taiwan

5 Someplace else (Specify: __________________ )

(If you were born in the U.S., skip this question). In which year you came to the United

States to stay for the long term? __ __ __ ___ Year

Do you have any religious beliefs?

Yes

No

If yes, please specify 1 Buddhism 2. Christianity 3 Others __________

Have you discussed your end of life care plan with family members?

Yes

No

Have you documented your end of life care plan through advance directive?

Yes

No

137

Please rate to what degree the following statements are true to you. 1=completely disagree,

2=mostly disagree, 3=neither agree nor disagree, 4=mostly agree, 5=completely agree.

Acculturation

1. I speak Chinese more often than English.

2. My English is much more fluent than Chinese.

3. I celebrate Chinese holidays (e.g., Chinese Spring Festival, Mid Autumn Day) more

frequently than American holidays (e.g., Christmas, Thanksgiving).

4. I watch Chinese TV programs and movies more frequently than English ones.

5. I read Chinese books or newspapers than English ones.

6. I write in English more often than in Chinese.

7. What I eat daily is mostly Chinese food.

8. Most activities I attend are associated with Chinese communities.

9. I mainly go shopping at Asian markets or grocery stores.

10. I feel at home living in the U.S.

Please rate to what degree the following statements are true to you. 1=strongly agree,

2=somewhat agree, 3=somewhat disagree, 4=strongly disagree.

Family cohesion

1. Family members respect one another.

2. We share similar values and beliefs as a family.

3. Family members feel loyal to the family.

4. Family members share their feelings with each other.

5. Family members like to be together when then have free time.

138

APPENDIX IV

SURVEY QUESTIONS ABOUT DEMOGRAPHICS, ACCULTURATION, AND

FAMILY COHESION

MANDARIN VERSION

关于研究对象背景、文化适应程度和家庭亲密度的问题

139

您的性别 1 男 2 女

您的出生年月： ______ 年_______ 月

您的教育程度? ___________________

1 小学或者以下

2 初中

3 高中

4 大专或者大学肄业

5 四年本科毕业

6 研究生及以上

您在哪里出生?

1 美国

2 中国大陆

3 香港

4 台湾

5 其他 (请注明: __________________ )

(如果选择出生地在美国，请跳过这个问题) 请问您哪一年__________来美国定居（长期居住）。

请问您有宗教信仰吗？

有

没有

如果有，请问是 1 佛教 2 基督教 3 其它_______

您和家人讨论过您的临终照料计划吗？

有

没有

您填写过您的临终照料计划（医疗照料委托书）吗？

有

没有

140

请根据自己的真实想法选出对下列描述的同意程度。1=完全不同意，2=大部分不

同意，3=一半一半，4=大部分同意，5=完全同意。

文化适应程度

1. 我说中文 (普通话、粤语、或其他方言) 的时间超过英文。

2. 我的英文水平比中文水平高。

3. 我庆祝华人的节日（春节，中秋节）的时候多过美国的节日（圣诞节，感恩

节）。

4. 我看中文电视 (普通话、粤语、等其他方言) 和电影超过看英文的节目。

5. 我看中文书籍或报纸超过看英文的。

6. 我用英语写信/email 超过用中文写的。

7. 我大部分时间吃的都是中餐。

8. 我参加的活动大都是华人举办的。

9. 我主要去亚洲（或中国）超市买东西。

10. 我对美国的生活很有认同感。

请根据自己的真实想法选出对下列描述的同意程度。1=非常赞同，2=比较赞同，

3=比较反对，4=非常反对。

家庭亲密度

1. 你的家庭成员互相尊重对方。

2. 你们有相似的价值观和信仰。

3. 你的家庭成员之间彼此忠诚。

4. 你们可以向彼此讲心事。

5. 你的家庭成员喜欢有空就聚在一起。