Message Production

This article was downloaded by: [Western Illinois University] On: 19 November 2013, At: 06:53 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Western Journal of Communication Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/rwjc20

Do Message Features Influence Responses to Depression Disclosure? A Message Design Logics Perspective Allison M. Scott a , John P. Caughlin b , Erin Donovan-Kicken c & Sylvia L. Mikucki-Enyart d a Department of Communication , University of Kentucky b Department of Communication , University of Illinois , Urbana- Champaign c Department of Communication Studies , University of Texas , Austin d Division of Communication , University of Wisconsin-Stevens Point Published online: 07 Feb 2013.

To cite this article: Allison M. Scott , John P. Caughlin , Erin Donovan-Kicken & Sylvia L. Mikucki-Enyart (2013) Do Message Features Influence Responses to Depression Disclosure? A Message Design Logics Perspective, Western Journal of Communication, 77:2, 139-163, DOI: 10.1080/10570314.2012.694007

To link to this article: http://dx.doi.org/10.1080/10570314.2012.694007

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &

Conditions of access and use can be found at http://www.tandfonline.com/page/terms- and-conditions

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

Do Message Features Influence Responses to Depression Disclosure? A Message Design Logics Perspective Allison M. Scott, John P. Caughlin, Erin Donovan-Kicken, & Sylvia L. Mikucki-Enyart

This project uses a multiple goals theoretical perspective to examine the sophistication of

responses to depression disclosure. In the first study, 504 college students imagined that a

friend revealed a depression diagnosis using one of several messages exhibiting different

identity management goals. Their responses were coded for expressive, conventional, or

rhetorical message design logic. Variation in how depression was disclosed influenced

the sophistication of responses. In the second study, 352 individuals with depression rated

messages taken verbatim from the first study. Expressive messages were evaluated less

favorably than conventional messages, which in turn were rated less favorably than rhe-

torical messages. The discussion focuses on the connection between attention to multiple

goals and communication sophistication in responding to important disclosures.

Keywords: Depression; Disclosure; Message Design Logic; Multiple Goals

Depression is a growing public health concern with widespread personal, social, and

economic consequences. A recent survey conducted by the Centers for Disease

Allison M. Scott (PhD, University of Illinois at Urbana–Champaign) is an Associate Professor in the Department

of Communication at the University of Kentucky. John P. Caughlin (PhD, University of Texas at Austin) is

Associate Head and Conrad Humanities Scholar in the Department of Communication at the University of Illi-

nois at Urbana–Champaign. Erin Donovan-Kicken (PhD, University of Illinois at Urbana–Champaign) is an

Assistant Professor in the Department of Communication Studies at the University of Texas at Austin. Sylvia

L. Mikucki-Enyart (PhD, University of Illinois at Urbana–Champaign) is an Assistant Professor in the Division

of Communication at the University of Wisconsin–Stevens Point. The authors would like to thank Dale

Brashers, Kristin Satterlee, and Marcia Scott for their assistance in conducting this research. An earlier version

of this paper was presented at the 2010 annual meeting of the National Communication Association. Correspon-

dence to: Allison M. Scott, 236 Grehan Building, Department of Communication, University of Kentucky,

Lexington, KY 40506-0042, USA. E-mail: [email protected]

Western Journal of Communication

Vol. 77, No. 2, March–April 2013, pp. 139–163

ISSN 1057-0314 (print)/ISSN 1745-1027 (online) # 2013 Western States Communication Association

DOI: 10.1080/10570314.2012.694007

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

Control and Prevention in the United States indicated that approximately 1 in 20

Americans suffers from depression (Pratt & Brody, 2008), and the World Health

Organization estimates that by the year 2020, depression will be a leading cause of

disability among citizens of the world, second only to cardiac disease (World Health

Organization, 2001). The individual and societal costs of mental illnesses like

depression can be devastating (U.S. Department of Health and Human Services,

1999). Included among the challenges faced by people coping with depression are

concerns and questions about how to reveal their diagnosis to others. Disclosing ill-

ness is important for accessing social support, seeking professional help, and adhering

to treatment regimens, all of which can ultimately lead to recovery from the illness

(e.g., Caughlin et al., 2009; Müller, Bode, Myer, Stahl, & von Steinbüchel, 2011).

Research on depression and self-disclosure has tended to focus on questions such

as what motivates a person to disclose (or not disclose) their diagnosis (e.g., Dennis

& Chung-Lee, 2006; Garcia & Crocker, 2008). This work primarily addresses the dis-

closer’s decisions, while leaving questions about responses to disclosure messages lar-

gely unexamined. There are, of course, some exceptions, such as research examining

what linguistic cues people believe are indicative of depression (Rodriguez, Holleran,

& Mehl, 2010). Still, little is known about what constitutes better and worse reactions

to disclosures of a depression diagnosis.

Understanding variations in how people respond to depression disclosure is

important because some of the potential benefits of disclosure, such as receiving

social support, depend on the response to the disclosure. Also, poor responses to

depression can intensify the discloser’s feelings of dysphoria and isolation and

thereby perpetuate depressive symptoms (Coyne, 1976). Thus, it is important to

understand factors affecting the quality of responses to depression disclosures. The

overall purpose of the present investigation was to provide a theoretical means of

explaining message sophistication in responses to depression disclosure. The objec-

tive of the first study was to examine how responses to depression disclosure might

be meaningfully influenced by different ways of disclosing the illness, and the objec-

tive of the second study was to test whether the different ways of responding to

depression disclosure affect evaluations of the responses. To accomplish these objec-

tives, we utilized a multiple goals theoretical perspective to conceptualize diversity in

disclosure and response messages.

Multiple Goals Theoretical Perspective

Many interpersonal communication scholars have noted that communication is often

aimed at addressing multiple goals (e.g., Clark & Delia, 1979; Dillard, 1990;

Goldsmith, 2004; Wilson, 2002). There are numerous multiple goals perspectives,

but one common assumption is that when individuals interact, they attempt to

accomplish multiple purposes. Although there are many possible communication

goals, certain types of goals are relevant across social situations, including instrumen-

tal goals (which describe the primary objective of the interaction, such as persuasion

or support), identity goals (which address how the interactants create and manage

140 A. M. Scott et al.

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

impressions), and relational goals (which involve the maintenance of desired rela-

tional qualities; Clark & Delia, 1979). These multiple goals sometimes present com-

munication dilemmas when accomplishing one purpose may come at the expense of

accomplishing other relevant goals. For instance, an individual may want to give

advice to another person who has just revealed a depression diagnosis, but the advice

might make the person with depression seem incompetent or the adviser seem intrus-

ive (e.g., Goldsmith & Fitch, 1997).

A second common assumption of multiple goals theories is that some ways of talk-

ing are more effective than others at accomplishing the multiple (and potentially con-

flicting) purposes relevant in certain situations. According to this assumption, the

quality of communication depends on the extent to which it achieves the primary

interaction task (e.g., disclosing or responding to depression) while also successfully

managing what the conversation means for identities and relationships. An impor-

tant implication of this assumption is that goals are not only a means of accounting

for message construction, but goals also provide an account for how people interpret

messages (Goldsmith, 2004). For instance, variations in whether and how certain

purposes are pursued in conversation influence how people assess communication

(Caughlin, 2010). Furthermore, a person’s attention to more goals and certain com-

binations of goals can influence how conversational partners perceive and respond to

talk (Caughlin et al., 2009).

Message Design Logics

The theory of message design logics is a particular multiple goals perspective that

focuses on how different kinds of messages address different communicative goals

(B. O’Keefe, 1988, 1990). The design logics themselves do not refer to whether a per-

son is pursuing minimal purposes or multiple ones; instead, the design logics refer to

the reasoning that implicitly connects a given set of goals and the messages produced.

Inferences about design logics presume that certain goals are normatively accepted as

salient in certain social contexts (B. O’Keefe, 1988). For example, identity concerns

are inherent in advice-giving situations: Giving an individual advice inevitably threa-

tens the advisee’s negative face by imposing on the person’s autonomy; moreover,

advice that is not followed poses a threat to the advisor’s positive face by undermin-

ing a sense of cooperativeness between the advice giver and recipient (Goldsmith &

Fitch, 1997). Notice that not attending to such potential threats is conventionally

problematic regardless of whether a message producer is aware of the conventionally

relevant goal.

According to message design logic theory (B. O’Keefe, 1988), there are three ways

of reasoning by which individuals can pursue conventionally relevant goals in com-

munication. Each of these message design logics entails a distinct view of the nature

and purpose of communication. The expressive message design logic treats commu-

nication as a means of expressing thoughts or feelings. People utilizing an expressive

design logic pay little attention to contextual information, rely little on interpretation

in assigning meaning in interaction, and thus may fail to attend to goals that are

Western Journal of Communication 141

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

conventionally relevant in certain situations. The conventional design logic treats

interaction as a cooperative effort in which people follow the rules governing what

is normatively expected in a situation and avoid socially inappropriate action. In

the conventional design logic, social structures (e.g., roles, obligations, formalized

relationships) are treated as fixed instead of flexible. Rather than seeking to simply

express thoughts and feelings, individuals using conventional design logic try to

accomplish conventionally relevant goals through socially appropriate means. Finally,

the rhetorical message design logic treats communication as a process of coordi-

nation in which identities and contexts are established through what is said and

how it is said. In rhetorical messages, social structures are flexible rather than fixed.

Instead of seeking to give conventionally appropriate responses as defined by the con-

text, people who produce rhetorical messages seek to redefine the social situation

through their communication.

These three message design logics are developmentally ordered in terms of sophis-

tication and functional utility: Expressive design logic is the least sophisticated and

least capable of producing effective messages, and rhetorical design logic is the most

sophisticated and most capable of producing effective messages (B. O’Keefe, 1988).

This hierarchy is theoretically grounded: A person must be able to express thoughts

and feelings before being able to do so in a conventionally appropriate manner, and a

person must understand conventional goals and contexts before being able to rede-

fine such goals and contexts. There is also empirical evidence for this hierarchy:

Sophistication of message design logic has been associated with higher cognitive

complexity (B. O’Keefe & Shepherd, 1987, 1989), higher communication competence

(B. O’Keefe & McCornack, 1987), more effective social support provision (Caughlin

et al., 2008; Peterson & Albrecht, 1996), and more effective persuasion (Lambert &

Gillespie, 1994).

In short, a multiple goals perspective explains how individuals manage various

goals in producing messages and how variations in goal attention influence how

people evaluate the messages. More specifically, the theory of message design logics

proposes that certain goals are conventionally understood to be salient in certain

situations, and communication that does not address such goals is normatively eval-

uated as ineffective or inappropriate. A multiple goals theoretical framework can be

applied to any communication situation, but the particular goals that are conven-

tionally relevant vary by social context. It is thus important to consider goals that

are commonly salient in conversations when a person discloses and responds to

depression.

Relevant Goals in Disclosing and Responding to Depression

In the case of disclosing depression to another person, the primary goal of the inter-

action presumably is to reveal the diagnosis. However, some research suggests that

another common goal of people who disclose depression is to gain social support.

In fact, there is evidence that many individuals with depression rank receiving sup-

port as more important to their quality of life than is their physical and mental health

142 A. M. Scott et al.

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

(Michalak, Yatham, Kolesar, & Lam, 2006). Sometimes this social support involves

emotional support, such as expressions of caring, concern, empathy, or reassurance;

in other cases, individuals with depression seek instrumental support, such as offers

of assistance or advice (Dennis & Chung-Lee, 2006).

The task goals of disclosing depression and thereby accessing social support are

constrained by relevant relational and identity goals. For instance, one goal people

pursue in disclosing depression is to maintain relationships. Many people with

depression report that others often treat them differently after learning about the

diagnosis or that their relationships are lost altogether, leading them to want to dis-

close their illness in ways that minimize relational damage (Michalak et al., 2006). A

second goal individuals have in disclosing depression is avoiding negative evaluation

from others. Because depression is a stigmatized illness, people considering disclosing

their depression commonly are worried about maintaining a positive identity and

often work to minimize any damage to their identity as a result of unfavorable con-

notations (Barney, Griffiths, Jorm, & Christensen, 2006).

The way individuals manage identity in disclosing illness can affect others’

responses to such disclosure. Recent research by Caughlin et al. (2009) has shown

that HIV disclosure messages designed to avoid negative evaluation of the discloser

elicited more relationally affirming responses but less instrumentally supportive

responses than messages that attended to other goals. It is possible that pursuing

identity goals in disclosing depression may likewise prompt differences in the sophis-

tication of support from others. Caughlin et al. (2009) manipulated attention to

identity goals in a direct manner; that is, certain disclosure messages explicitly asked

recipients not to evaluate the discloser negatively (e.g., ‘‘Please don’t judge me’’).

Although direct efforts to attend to identity issues appear useful, there is evidence

that identity management during illness disclosures is often less direct. Cohen Silver,

Wortman, and Crofton (1990) proposed support-seeking through self-presentation

of coping style as one means of identity work in illness disclosure. In their study

on people disclosing cancer, they examined three self-presentations of coping style:

‘‘Poor coping’’ involved displaying intense distress about the illness diagnosis and

appearing to have great difficulty coping with circumstances; ‘‘good coping’’ entailed

expressing an optimistic view of the illness and appearing to cope very well with the

situation; finally, ‘‘balanced coping’’ involved conveying some distress about the ill-

ness diagnosis, but also acknowledging some success in coping with the situation.

Cohen Silver et al. (1990) found that self-presenting a poor coping style elicited less

favorable responses than self-presenting a good coping style, but the good coping

style still did not elicit support; rather, it was the balanced coping style that elicited

the most supportive responses. The balanced coping style represented a particularly

effective strategy because it indirectly acknowledged the need for support while also

preserving the discloser’s positive face (because it acknowledged some success in

coping). It may be that indirect identity work through self-presentation of coping

also influences responses to depression disclosure.

Previous work on depression disclosure has focused primarily on the perspective

and goals of the discloser; however, a deeper understanding of the disclosure process

Western Journal of Communication 143

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

requires also examining the perspective and goals of the respondent. Although reci-

pients of a depression disclosure would undoubtedly have some idiosyncratic conver-

sational goals, it is also possible to discern certain goals that are conventionally

understood to be relevant in this situation. For example, being told that someone

has been diagnosed with depression would generally be considered ‘‘bad news,’’

and the usual expectation in such situations is to provide some form of support or

comforting. Thus, the goal of comforting the person with depression would be rel-

evant, regardless of whether any particular individual recognized the relevance of this

goal or provided adequate comforting (see Burleson, 1994). Furthermore, the goal of

providing adequate comfort makes identity and relational goals relevant given that

the quality of comforting messages often depends on the extent to which the message

is person-centered (Burleson, 2003) and contains relational affirmation (Samter,

Burleson, & Basden-Murphy, 1989).

In sum, an examination of the depression disclosure literature indicates that

instrumental, identity, and relational goals are relevant in disclosing a depression

diagnosis as well as in responding to such a disclosure. This broad understanding

of conventionally relevant goals in depression disclosure interactions provides a basis

for assessing the quality of responses to depression disclosure.

Study 1

The primary objective of the first study was to assess how different ways of disclosing

depression might meaningfully influence the way others respond. The first step in

achieving this objective was to examine how responses to depression disclosure vary

and how such variation could be reliably classified. Given that the process of dep-

ression disclosure is distinct from the contexts of other research on message design

logics, we first asked two basic research questions:

RQ1: Do responses to depression disclosure vary with respect to message design logic?

RQ2: What message features characterize expressive, conventional, and rhetorical responses to depression disclosure?

The second step was to explore how different ways of disclosing depression affect

responses to such disclosure. Clearly, an infinite number of potential disclosure mes-

sages are conceivable, so an important issue is how to conceptualize message varia-

tions. One previous study of HIV disclosures (Caughlin et al., 2009) suggested that

blunt statements about identity goals can influence disclosure responses, but research

by Cohen Silver et al. (1990) demonstrates that attention to identity during illness

disclosures is often more subtle. Thus, self-presentation styles (i.e., poor coping, good

coping, balanced coping) examined by Cohen Silver et al. were used in the current

study as a primary way of conceptualizing variations in disclosure message. These

self-presentational styles represent indirect ways of managing identity while discuss-

ing an illness disclosure, and we sought to compare messages demonstrating each of

these identity management strategies with two other types of messages: one that

focused solely on disclosing the information (with no explicit attention to any other

144 A. M. Scott et al.

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

goal) and one that disclosed depression and explicitly sought support. We were inter-

ested in whether these different types of disclosures would elicit better or worse

responses as conceptualized by the theory of message design logics, which prompted

the third research question:

RQ3: Do different ways of disclosing depression elicit different proportions of expressive, conventional, and rhetorical responses?

Method

Participants

Participants included 504 students from communication courses at a large public

midwestern university. Given the rates of depression among college students (e.g.,

Furr, Westefeld, McConnell, & Jenkins, 2001), the topic of depression was undoubt-

edly familiar to these participants. The sample included 293 (58.13%) females, 191

(37.90%) males, and 20 (3.97%) respondents who provided no information about

their sex. Participants averaged 20.14 years of age (SD¼ .90, range¼ 18–28). The

sample included White (n¼ 311, 61.71%), African American (n¼ 82, 16.27%), Asian

American (n¼ 38, 7.53%), and Latino (n¼ 33, 6.55%) participants, with 22 (4.37%)

respondents reporting other ethnic backgrounds and 18 (3.57%) declining to report

ethnicity.

Procedure

Participants completed an online questionnaire that contained an open-ended item

that directed students to ‘‘Imagine that you are talking alone with a good friend.

Please think of a particular friend. Now, imagine that after a few minutes of talking,

your friend says, ‘Well, I have something to tell you . . . ’’’ This statement was fol-

lowed by one of 20 disclosure messages indicating that the friend had been recently

diagnosed with depression. Respondents were asked, ‘‘If this were to happen, what

would you do or say in response? Please write the exact words you would say, as

though you were writing a script.’’ Participants typed their responses and then com-

pleted several closed-ended questions. The 20 specific disclosure messages were

grouped into five conditions, which reflected different ways of attending to identity

goals. The kernel condition (n¼ 100) contained four messages that only disclosed the

diagnosis and contained no information about the discloser’s coping style. The mes-

sages in each of the remaining four conditions included a kernel message and refer-

ence to poor coping (n¼ 87), good coping (n¼ 96), balanced coping (n¼ 122), or

explicit support-seeking (n¼ 98). Overall, participants rated the disclosure messages

as fairly realistic (on a 5-point scale, M¼ 3.71, SD¼ .70). The messages for each of

these conditions are presented in Appendix A.

This multiple messages design allowed us to manipulate features of messages that

theoretically should matter (i.e., identity management strategies). The strategy

required us to expose a large sample of individuals to the messages, making a

Western Journal of Communication 145

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

confederate study impractical. And although it might be possible to study some nat-

urally occurring disclosures of depression, it is unlikely that we could obtain a large

enough sample of messages to examine systematic differences in the message features

of interest. Thus, the current method was considered the best option given the

reviewed theory and literature as well as what is feasible and ethical.

Coding

In coding the open-ended responses for message design logic (using each person’s

entire response as the unit of analysis), we based our coding scheme on B. O’Keefe’s

(1988) description of message design logics. We also drew from Caughlin et al. (2008)

as a recent example of how the theory of message design logics can be utilized in a

coding scheme. Prior to coding, all authors collaborated on adapting the design logic

categories to the current research problem. Then three of the authors coded each

message into one of three categories: expressive (i.e., messages that communicated

thoughts and feelings, often without attending to conventionally relevant situational

goals), conventional (i.e., messages that contained socially normative content and

structure), and rhetorical (i.e., messages that attended to multiple goals, often

attempting to redefine the situation, identities, or relationships). Reliability among

the coders, which was assessed with an intraclass correlation (Fleiss, 1981), was good

(intraclass correlation coefficient ¼ .80). Discrepancies in coding were resolved in

meetings with all coders.

Results

In general, the results demonstrated that responses to depression disclosure varied in

sophistication, and this variation was associated with the way in which depression

was disclosed.

Sophistication of messages

The first research question asked if the responses to disclosure of a depression diag-

nosis would vary in the sophistication of the message design logics used. We found

that responses did vary in level of sophistication, with conventional messages occur-

ring most frequently (n¼ 370, 73.41%), followed by expressive (n¼ 102, 20.24%)

and rhetorical (n¼ 31, 6.15%) messages. A test of two proportions (distributed

as z; see Brunig & Kintz, 1997) indicated that the differences in the proportions

between design logics were significant: There were significantly more conventional

messages than expressive (z¼ 16.87, p< .001) and rhetorical (z¼ 6.52, p< .001) mes-

sages, and there were more expressive messages than rhetorical messages (z¼ 21.77,

p< .001).

Characteristics of message design logics

In the second research question, we asked about the defining characteristics of

expressive, conventional, and rhetorical response messages in the context of

146 A. M. Scott et al.

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

depression disclosure. We found that responses to depression disclosure could be

reliably classified as described by the theory of message design logics. Many expressive

messages in this study revealed the recipient’s cognitive or emotional state by express-

ing surprise or sadness, or the desire to satisfy the respondent’s curiosity about the

situation, but these messages typically lacked explicit comforting features, as shown

in Example 1.

Example 1: ‘‘Really? What made you think you have it in the first place? And is it something really serious that you have to get medication for, or can you just do something in your everyday life to help you overcome it?’’

Messages containing explicitly negative features, such as minimizing or denying

the diagnosis (Example 2), implicitly blaming or criticizing the discloser (Example

3), or implying that the discloser would commit self-harm (Example 4) were also

categorized as expressive.

Example 2: ‘‘Dude, you don’t have depression. You’re just a little down. You and me are too strong to have depression. You’re talking crazy.’’

Example 3: ‘‘Please tell me this is a joke. All you do is sit around all day and enable your ‘depression.’ This is insane . . . ’’

Example 4: ‘‘That’s fucked up, man. Don’t kill yourself.’’

Conventional responses to depression disclosure tended to follow socially norma-

tive expectations by, at a minimum, containing an attempt to comfort or offer assist-

ance to the discloser. Some conventional messages contained explicit emotional

support (Examples 5 and 6), whereas others included more indirect offers of support

(Examples 7 and 8).

Example 5: ‘‘I’m so sorry. Is there anything that I can do to help? Is there anything that you would like to talk about? I want you to know that I am always here for you.’’

Example 6: ‘‘Do you want to talk about it? If you do not, I just want to let you know that you are my best friend and I wouldn’t want anything bad to happen to you.’’

Example 7: ‘‘I’m here for you.’’ Example 8: ‘‘I’m so sorry. How do you feel about this?’’

Rhetorical messages reframed the meaning of the depression diagnosis in various

ways. In some rhetorical responses, the participant redefined the relational context,

reframing the diagnosis as a shared problem, as in Example 9.

Example 9: ‘‘You are my best friend in the world and I want you to know that I am here for you through this, and even on your worst days you should come to me and I will be there for you. We will get through this. Don’t even think of it as something you are going through alone. I know it is hardest for you because you have to live through it every day, but it is still something that I will go through with you like it is my own issue. I love you!’’

Western Journal of Communication 147

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

In other cases, rhetorical messages took the conventionally understood definition

of depression as a weakness and redefined disclosure of depression as an indication of

strength (Example 10).

Example 10: ‘‘Oh my gosh. I’m sorry to hear that. I am here for you every step of the way. It is not your fault that you are suffering from this. There are a lot of people who get this, so don’t feel like you are the only one. It was strong of you to go to the doctor and get help. This will not be easy, but I promise you that I will be here for anything that you need.’’

Relationship between disclosure conditions and response message design logics

In response to the third research question, which asked if the various ways of disclos-

ing a depression diagnosis affected the sophistication of responses, we conducted a

5� 3 (disclosure condition by response message design logic) chi-square difference

test. Overall, the design logics of the responses differed significantly by condition:

v2(8)¼ 59.16, p< .001, u¼ .24. We further examined the differences between the

conditions by comparing the proportions of the three message design logics for each

condition. These differences are presented in Table 1.

In general, the good coping disclosure condition received the fewest rhetorical

responses and elicited significantly fewer rhetorical messages than did the kernel

(z¼ 2.61, p< .01) and poor coping (z¼ 2.44, p< .05) conditions. Good coping dis-

closure messages also prompted a higher proportion of expressive messages than did

any other type of message, and the proportion of expressive messages was significantly

higher for the good coping condition than for the poor coping (z¼ 2.15, p< .05),

balanced coping (z¼ 4.73, p< .001), and support-seeking (z¼ 5.35, p< .001) con-

ditions. Good coping messages elicited significantly fewer conventional messages than

the balanced coping (z¼ 4.02, p< .001) and support-seeking (z¼ 4.44, p< .01)

conditions.

There were additional significant differences in the proportions of message design

logics. The balanced coping disclosure condition received fewer expressive messages

Table 1 Proportions and Frequencies of Response Message Design Logic for Each

Disclosure Condition

Message Design Logic of Response

Disclosure condition Expressive Conventional Rhetorical

Kernel .30 (30)ab .59 (59)ab .11 (11)a

Poor coping .22 (19)cde .68 (59)cd .10 (9)b

Good coping .38 (36)cfg .61 (59)ef .01 (1)ab

Balanced coping .10 (12)adf .86 (105)ace .04 (5)

Support-seeking .05 (5)beg .90 (88)bdf .05 (5)

Note. Proportions appear outside and frequencies appear inside the parentheses. Proportions in the same col-

umn that share subscripts differ significantly at p< .05.

148 A. M. Scott et al.

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

than did the kernel (z¼ 3.64, p< .001) and poor coping (z¼ 2.21, p< .05) conditions

and more conventional messages than the kernel (z¼ 4.41, p< .001) and poor coping

(z¼ 2.99, p< .01) conditions. Similarly, the explicit support-seeking message elicited

fewer expressive messages than did the kernel (z¼ 4.41, p< .001) and poor coping

(z¼ 3.16, p< .01) conditions and more conventional messages than the kernel

(z¼ 4.79, p< .001) and poor coping (z¼ 3.51, p< .001) conditions.

Discussion

The findings from this first study provide an important first step toward understand-

ing what counts as a sophisticated response to depression disclosure and how features

of depression disclosure messages influence the impact of those messages.

Sophistication of responses to depression disclosure

The present results demonstrate that it is possible to reliably classify responses to

depression disclosure using B. O’Keefe’s (1988, 1990) message design logic theory,

and the hierarchical ordering implied in the theory provides a coherent way to

describe and classify the sophistication of responses to disclosure of depression. These

findings are consistent with previous work that has confirmed the utility of the theory

for conceptualizing message sophistication in a variety of contexts, including regulat-

ive (B. O’Keefe, 1988), compliance-gaining (Lambert & Gillespie, 1994), and illness

disclosure (Caughlin et al., 2008) situations. However, the proportion of message

design logics produced in the current study differed from the proportions yielded

in previous studies. For example, the distribution of design logics in the study con-

ducted by Caughlin et al. (2008), which examined responses to HIV disclosure, was

32% expressive, 55% conventional, and 12% rhetorical (compared to the current dis-

tribution of 20% expressive, 74% conventional, and 6% rhetorical). This difference is

statistically significant, v2 (2)¼ 37.65, p< .001, u¼ .19, and it suggests that the spe-

cific tasks examined in each study are different in potentially important ways. The

differences between the current findings and previous ones are consistent with

message design logic theory, which suggests that different types of tasks make relevant

different kinds of goals and thus require more or less sophisticated messages

(B. O’Keefe, 1988). In this way, the communication situation influences the talk that

is produced (e.g., routine contexts prompt routine messages).

Although the situation appears to influence the responses, the first study does not

examine exactly what it is about the situation that elicits particular responses. It is not

clear, for example, why so few rhetorical responses were produced. Perhaps the task

of responding to depression disclosure does not seem exceptional enough to warrant

producing rhetorical responses over conventional ones. It is also possible that design-

ing a rhetorical message in response to depression disclosure is particularly difficult

because people with depression often exhibit aversive depressive symptoms that

people want to counteract (Coyne, 1976). This can lead recipients of depression dis-

closure to try to comfort in supposedly responsive but potentially less genuine (i.e.,

Western Journal of Communication 149

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

less sophisticated) ways because the support is offered in the interest of alleviating the

respondent’s own discomfort. Another possible explanation for the relative rarity of

rhetorical responses is that individuals may have a tacit understanding that the

expected response to the mention of an illness like depression is a conventional sup-

port response, and only if support is more directly requested would people feel that it

is appropriate to produce rhetorical responses. Thus, different messages may have

elicited more rhetorical responses in this context; for instance, messages that overtly

signaled a need for help with sense-making may have prompted a higher proportion

of rhetorical messages.

Clearly, future research should examine the particular reasons why situations elicit

different types of response logics, but the current study is important in that it pro-

vides evidence that message design logics contain a situational component. Most pre-

vious work on message design logics has highlighted the individual differences in

message design that are evident in communicatively complex situations. The results

of the present study, however, demonstrate that different situations evoke different

patterns of design logics.

Variation in depression disclosures and responses

Not only do different contexts elicit responses that differ in message design logics, but

this first study indicates that different ways of disclosing depression (i.e., self-

presentation of coping style) also elicit responses that vary in design logic. Consider-

able research on illness disclosure has focused on situational factors affecting

disclosure, such as barriers people face in revealing their illness and the timing of dis-

closure (e.g., Charmaz, 1991; Greene, Derlega, Yep, & Petronio, 2003). The current

study suggests that, in addition to such situational factors, self-presentation features

of the disclosure messages also influence responses to such disclosures.

We found that variation in identity work in depression disclosure messages influ-

enced variation in the message design logics of responses to disclosure. Some disclos-

ure conditions tended to elicit less sophisticated responses. For instance, the good

coping condition received the lowest proportion of rhetorical responses and the high-

est proportion of expressive responses of all the disclosure conditions. One possible

reason for this finding is that self-presenting a good coping style may imply that the

discloser is so successful in dealing with the depression that ‘‘good’’ comforting is not

necessary and thus respondents do not exert the effort needed to construct a

sophisticated message. By contrast, the poor coping disclosure condition prompted

a significantly higher proportion of rhetorical responses and a significantly lower pro-

portion of expressive responses than the good coping condition. This finding suggests

that when a person self-presents as having great difficulty in dealing with the

depression, individuals assess the situation as calling for a better response.

The present results concerning poor coping and good coping differ from the find-

ings reported by Cohen Silver et al. (1990), who found that people responded more

favorably to individuals who presented a good coping style and less favorably to indi-

viduals who presented a poor coping style. One possible reason for these divergent

150 A. M. Scott et al.

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

findings could be the difference in outcome measures: Cohen Silver and her collea-

gues examined the distress felt by individuals after hearing the illness disclosure, and

we examined the competence of people’s responses to the disclosure, making direct

comparisons between the studies difficult. It is possible, however, that respondents to

illness disclosure produce sophisticated messages because the distress of the situation

prompts them to put effort into designing their response.

We also found that some disclosure conditions functioned similarly in eliciting

more or less sophisticated responses. For example, the balanced coping and support-

seeking conditions each received fewer expressive messages and more conventional

messages than the kernel and poor coping conditions. Given that self-presenting as

having some success at coping prompted supportive responses approximately as well

as explicitly asking for support, it could be that presenting a balanced coping style is

recognized by others as a request for support (e.g., Goldsmith, 2004). Although the

two conditions fared similarly, there may be an advantage to presenting a balanced

coping style because it appears to be as effective at eliciting support as a direct request

but is less likely to carry the potential for identity loss (e.g., being seen as incompetent

or not self-reliant) that can accompany on-record requests for support (e.g.,

Goldsmith, Lindholm, & Bute, 2006).

Study 2

The first study demonstrated that responses to depression disclosure meaningfully

vary in sophistication. However, the first study could not provide direct evidence that

messages of greater sophistication are more effective or appropriate in responding to

depression disclosure. This was a significant limitation given that sophistication of

message design logic is not always related to the quality of supportive messages in

a straightforward manner (e.g., B. O’Keefe, Lambert, & Lambert, 1997; Peterson &

Albrecht, 1996). To address this limitation, the objective of the second study was

to explore how responses to depression disclosure with different message design

logics are evaluated in terms of perceived quality or competence.

Communication competence is often conceptualized as the degree to which com-

munication is appropriate and effective (Canary & Spitzberg, 1987). Given the con-

ventional relevance of support to a situation in which a person discloses a depression

diagnosis, perceived effectiveness in the current context is based on the extent to

which the response is viewed as supportive (Goldsmith, McDermott, & Alexander,

2000). Using the assumption that supportiveness is the key indicator of effectiveness,

we conceptualized competence here as involving supportiveness and appropriateness.

To determine the sophistication of responses to depression disclosure, it is neces-

sary to obtain a cultural informant’s perspective on what counts as supportive and

appropriate in the particular context of depression disclosure. The theory of message

design logics does not require a researcher to make inferences about the goals that a

person actually has when producing a message to make a determination of quality.

Instead, what is needed is cultural insight into what kinds of communicative

purposes should be attended to in a particular situation. Thus, assessing the

Western Journal of Communication 151

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

sophistication of responses to depression disclosure is a very different task from pro-

ducing a response to a depression disclosure. Anyone could potentially be a recipient

of a depression disclosure, and the first time that happened, that recipient would have

to draw upon whatever limited experiences the recipient had to respond. Thus, no

particular experience is necessary to produce some response (as in Study 1), but eval-

uating what counts as an appropriate response requires the perspective of individuals

who understand the particular challenges associated with having and disclosing

depression. That is, people who have experienced depression are ideal informants

for making assessments about what counts as a sophisticated response to depression

disclosure because they are likely to understand the particular context of disclosing

depression and its normatively relevant goals. In Study 2, it was therefore crucial

to recruit participants with a history of experiencing depression.

The hypotheses for the second study were guided by the hierarchical ordering of

message design logics implied in B. O’Keefe’s (1988) theory:

H1: Responses demonstrating different message design logics will be rated differently in terms of supportiveness. H1a: Expressive messages will be rated as less supportive than conventional

messages. H1b: Conventional messages will be rated as less supportive than rhetorical

messages. H2: Responses demonstrating different message design logics will be rated

differently in terms of appropriateness. H2a: Expressive messages will be rated as less appropriate than conventional

messages. H2b: Conventional messages will be rated as less appropriate than rhetorical

messages.

Although the hypotheses involved straightforward predictions about which types

of messages would be rated most favorably overall, there was one construct that

was particularly important to account for because it could potentially confound

the results. As noted above, more sophisticated message design logics are not always

viewed as more competent; for example, individual differences among message reci-

pients may influence their perceptions of competence. There is evidence that the cog-

nitive ability of message recipients can influence the extent to which a message is

perceived to be well suited for a particular interaction. For instance, B. O’Keefe

et al. (1997) found that individuals prone to using expressive design logic often

evaluated rhetorical strategies as overly indirect or even disingenuous. Samter et al.

(1989) found that individuals with greater cognitive ability appreciated message com-

plexity more so than did less cognitively skilled individuals. Such findings suggest

that the relatively sophisticated identity and relational work of rhetorical messages

may be lost on less sophisticated message recipients. Although rhetorical messages

may be most effective and appropriate on average, the advantages of relatively soph-

isticated messages may not matter to recipients who view communicative situations

in comparatively simplistic terms, and thus perceived message quality depends, in

part, on the sophistication with which a recipient perceives the situation. Given this

152 A. M. Scott et al.

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

possibility, it was important to be able to control for variations in recipients’ com-

municative skill levels. One particularly useful construct for assessing individual dif-

ferences in people’s ability to discern nuances in situations is cognitive complexity,

which refers to the degree of differentiation in a person’s system of interpersonal

constructs (Burleson & Caplan, 1998). Because cognitive complexity may influence

perceptions of message quality, we asked:

RQ4: Does statistically controlling for individuals’ cognitive complexity influence any associations between the message design logics and participants’ percep- tions of depression disclosure response quality?

Method

Participants

Given that they would be uniquely able to evaluate the quality of responses to

depression disclosures, we recruited 352 participants who had been diagnosed with

depression. Participants resided in many regions of the United States, including

the North, South, Midwest, Southwest, and Northeast. To be eligible, participants

had to report having been professionally diagnosed with depression. Most partici-

pants reported being diagnosed with major depressive disorder (n¼ 159, 45.17%),

and others reported receiving a diagnosis of depressive disorder NOS (n¼ 73,

20.74%),major depressive episode (n¼ 66, 18.75%), bipolar disorder (n¼ 42, 11.93%),

dysthymic disorder (n¼ 11, 3.13%), or cyclothymic disorder (n¼ 1, .28%). The

majority of participants had been diagnosed with depression over 2 years ago

(n¼ 258, 73.30%), and others had been diagnosed 1–2 years ago (n¼ 42, 11.93%),

7–12 months ago (n¼ 19, 5.40%), 4–6 months ago (n¼ 14, 3.98%), or less than 3

months ago (n¼ 18, 5.11%). Demographic questions appeared at the end of the sur-

vey and had much lower response rates than the rest of the study; 115 participants

(32.67%) declined to report their sex or ethnicity. Based on those who did respond,

the sample included 200 (56.82%) females, 37 (10.51%) males, and the average age of

participants was 34.52 years old (SD¼ 12.40, range¼ 18–67). The sample consisted

of White (n¼ 192, 54.55%), African American (n¼ 16, 4.55%), Latino (n¼ 11,

3.13%), and Asian American (n¼ 10, 2.84%) participants, with 8 (2.27%) respon-

dents reporting other ethnic backgrounds.

Procedure

We posted information about the study on numerous online bulletin boards, and

potential participants were directed to a Web page that contained information about

the study, including informed consent information. Those who chose to participate

selected a link to a secure survey Web site. The link randomly directed participants to

one of the 18 different versions of the survey, and at least 13 participants completed

each version. Participants who were willing to provide contact information were

entered in a lottery to win one of five $50 prizes.

Western Journal of Communication 153

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

Measures

Participants were first asked to ‘‘Imagine that you just told one of your good friends

that you have been diagnosed with depression for the first time. Imagine that your

friend reacts in the way described below.’’ This statement was followed by one of

18 reactions demonstrating each of the three message design logics (6 expressive, 6

conventional, 6 rhetorical) that were randomly selected from the 504 coded responses

to depression disclosure from the first study. Each respondent was asked to imagine

one particular reaction. Other than the specific reaction message, the 18 versions of

the questionnaire were identical. Appendix B contains the 18 responses used in this

study.

After being asked to consider the response message, participants rated the suppor-

tiveness and appropriateness of the messages using 12 Likert-type items assessed on a

5-point scale. Message supportiveness was measured using six items (‘‘comforting,’’

‘‘generous,’’ ‘‘helpful,’’ ‘‘reassuring,’’ ‘‘sensitive,’’ ‘‘supportive’’) based on Goldsmith

et al. (2000), andmessage appropriateness was measured with six items (‘‘appropriate,’’

‘‘considerate,’’ ‘‘decent,’’ ‘‘proper,’’ ‘‘respectful,’’ ‘‘understanding’’) based on Caughlin

et al. (2008). Both subscales had good reliability (Cronbach’s a¼ .85 for supportive-

ness, Cronbach’s a¼ .87 for appropriateness).

Participants’ cognitive complexity was measured using the two-peer version of

Crockett’s (1965) Role Category Questionnaire, which has amassed considerable con-

struct, concurrent, and discriminant validity (Burleson & Waltman, 1988). Parti-

cipants wrote a free-response description of two well-known peers, one liked and

one disliked. The descriptions were coded following the coding rules given in

Crockett, Press, Delia, and Kenny (1974). Two coders counted the number of inter-

personal constructs reflected in each description. The number of constructs in each of

the two descriptions were then summed to create the cognitive complexity score,

which represents an index (rather than an exhaustive measure) of the participants’

cognitive complexity. Both coders coded the same 25% of the total sample of descrip-

tions to establish intercoder reliability, which was excellent (intraclass correlation

coefficient¼ .95). Disagreements in coding were resolved by the two coders by dis-

cussing the descriptions and reaching consensus on the coding decisions. Once inter-

coder reliability was established, the two coders evenly divided and independently

coded the remaining descriptions.

Results

As a preliminary analysis, we conducted an ANOVA to determine if the different

types of depression reported by participants were related to the dependent variables.

There was no evidence that the results were dependent on the particular diagnosis

experienced by participants: The ANOVA was not significant for supportiveness,

F (5, 345)¼ 1.41, p¼ .22, or appropriateness, F (5, 345)¼ .39, p¼ .86. We therefore

included participants with all types of depression in the remaining analyses.

To test the hypotheses, we conducted an ANCOVA that treated the specific res-

ponse messages as a random factor within the three message design logics. Treating

154 A. M. Scott et al.

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

the messages as a random factor accounts for variations that may occur because of

differences in messages within a particular category, and it makes it possible to gen-

eralize the findings beyond the specific response messages examined in the present

research to the larger categories of message design logics (Jackson, 1992; Jackson &

Brashers, 1994). Cognitive complexity was added as a covariate, and we included

planned contrasts to test if conventional messages were rated more favorably than

were expressive messages and if rhetorical messages were rated more favorably than

were conventional messages.

In an initial ANOVA without the covariate, there was an overall significant differ-

ence among the three message design logics for supportiveness, F (2, 15)¼ 5.92,

p¼ .01, g2¼ .05. To address the research question, the same analysis was repeated

while controlling for cognitive complexity; the significant result held, and the effect

size was slightly larger, F (2, 15)¼ 7.22, p< .01, g2¼ .06. In short, there was support

for the first hypothesis overall, and it did not appear to be dependent on participants’

cognitive complexity.

To examine the subparts of the first hypothesis, the specific message design logics

were compared, using the analyses that controlled for cognitive complexity. Because a

covariate was included in the model, the planned contrasts were based on estimated

marginal means (EMM). Consistent with H1a, messages with an expressive logic

(EMM¼ 2.12, SE¼ .12) were rated significantly lower in supportiveness than were

messages with a conventional logic (EMM¼ 2.47, SE¼ .12), F (1, 10)¼ 3.42, p< .05,

g2¼ .03. Consistent with H1b, rhetorical messages (EMM¼ 2.79, SE¼ .12) were

evaluated as significantly more supportive than conventional messages, F (1, 10)¼ 3.76, p< .05, g2¼ .02.

By contrast, the findings for appropriateness were not significant, F (2, 15)¼ 0.43,

p¼ .67, g2¼ .001. Controlling for cognitive complexity did not meaningfully change

the findings; there was still no evidence of an association between design logic and

ratings of appropriateness, F (2, 15)¼ 0.25, p¼ .78, g2¼ .001. Thus, the second

hypothesis was not supported.

The research question asked how controlling for cognitive complexity might influ-

ence the findings. We found that adding cognitive complexity as a covariate did not

meaningfully change the results of the analyses, although controlling for cognitive

complexity allowed a slightly larger effect size for supportiveness to emerge.

Discussion

The purpose of the second study was to analyze what counts as a high-quality

response to depression disclosure. Consistent with B. O’Keefe’s (1988, 1990) theory

of message design logics, we found that messages that demonstrated greater sophis-

tication in attending to conventionally relevant goals were rated higher in quality

than less sophisticated messages. More specifically, rhetorical responses to depression

disclosure were evaluated as significantly more supportive than conventional

responses, and conventional responses were evaluated as more supportive than

expressive responses.

Western Journal of Communication 155

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

The finding that rhetorical messages were evaluated as more supportive than

conventional and expressive ones suggests that more sophisticated messages are

evaluated more favorably in general, but it is important to note that more sophisti-

cated messages will not always be evaluated most favorably. Many factors may affect

the overall tendency for sophisticated messages to be assessed as higher in quality. We

investigated cognitive complexity as one potential moderator and found that parti-

cipants’ cognitive complexity did not meaningfully influence assessments of response

messages, which strengthened our confidence in the general applicability of the

ordering of the message designs. Still, there are other factors that may provide caveats

to conclusions about which type of messages are evaluated most favorably in general.

For example, the relative importance of conventionally relevant goals might influence

evaluations of messages. The current finding that messages with different design

logics differed significantly in terms of supportiveness but not appropriateness may

indicate that supportiveness is the overriding imperative in the situation of respond-

ing to depression disclosure.

General Discussion

Taken together, the results from the two studies reported here have a number of the-

oretically interesting implications. First, the present research provides a theoretical

means of conceptualizing message sophistication in responses to depression disclos-

ure. In the first study, we found that individuals’ responses to depression disclosure

varied in the way they attempted to rectify the multiple goals relevant to the situation.

The second study showed that response messages using more sophisticated design

logics for managing the multiple goals were rated as higher in quality than those with

less sophisticated design logics. This finding provides valuable empirical evidence not

only for what counts as a better or worse response to depression disclosure, but also

for the usefulness of message design logics and a multiple goals theoretical perspec-

tive as ways of understanding what makes responses to important disclosures more or

less effective in general. Furthermore, the present results demonstrate the utility of a

multiple goals perspective as a theory of message interpretation. Most research utiliz-

ing a multiple goals framework has focused on how goals influence message construc-

tion (Berger, 2005), but multiple goals approaches also represent a useful theoretical

lens for explaining how people interpret and assess communication (Caughlin, 2010).

By demonstrating that messages that attended well to multiple goals were assessed

more favorably than messages that did not, the current investigation represents an

important step toward tapping into the largely unrealized potential of multiple goals

perspectives to explain interpretive processes in conversations.

In addition to bolstering the connection between managing multiple goals and

communication sophistication, our results also demonstrate that the hierarchical

ordering implied in message design logic theory affords a coherent way to classify

the sophistication of responses to disclosure of depression and thus provide further

evidence of the theory’s utility in explaining the process of illness disclosure more

broadly. Extant work on message design logics has largely focused on regulative

156 A. M. Scott et al.

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

interpersonal communication, or situations in which one person is faced with a need

to correct or control the behavior of another person (e.g., Lambert & Gillespie, 1994;

B. O’Keefe, 1988, 1990), although recent work has begun to investigate the utility of

the theory in explaining HIV disclosure (Caughlin et al., 2008). The current investi-

gation demonstrates that the application of message design logics can be usefully

extended to multiple illness contexts.

The present research also extends our understanding of the features of messages

that can elicit better or worse responses. Cohen Silver et al. (1990) asserted that

how one self-presents coping with illness affects how others respond, and consistent

with this, we found that messages varying in how the discloser was coping influenced

reactions. We believe that the current research is the first to demonstrate this effect

with respect to disclosures about depression.

Such findings have practical significance. The first study demonstrates that the way

individuals manage identity in disclosing illness can affect others’ responses to such

disclosure. This suggests that different ways of attending to identity goals in disclos-

ing depression elicit different kinds of responses, and thus the quality of the response

is based at least in part on how the information is disclosed. Given the challenges

involved in disclosing depression, it is useful to know that certain kinds of messages

(e.g., self-presenting a balanced coping style: ‘‘I’m having a hard time with this, but I

feel like I’m making some progress’’) have a greater likelihood than do others (e.g.,

self-presenting a good coping style: ‘‘Things are going great’’) of allowing disclosure

while minimizing some of the negative reactions to it. Such findings suggest that the

manner in which goals are pursued (not just which goals are addressed) is important.

In providing a general framework for understanding what constitutes a better and

worse response to depression disclosure, these studies highlight several specific mess-

age features that can be incorporated into concrete recommendations about how to

respond when a person discloses a depression diagnosis. For instance, given that

expressive messages were rated as least supportive, it would be prudent for those

wanting to respond supportively to a person’s depression diagnosis to avoid express-

ing their own feelings about the diagnosis, asking questions to satisfy their own curi-

osity, or implicitly or explicitly blaming or criticizing the person. Moreover, given

that rhetorical messages were rated most favorably, recipients of depression disclos-

ure would do well to put the diagnosis in a broader perspective, frame the illness as a

shared problem, or affirm the disclosure as a sign of strength.

Conclusion

The current investigation provides initial evidence for what is helpful and not helpful

in responding to depression disclosure. However, like any study, the present research

is not without limitations. The disclosures used in the first study represented hypo-

thetical rather than actual messages. It is a limitation that our evidence for what

counts as a high-quality response to depression does not come from observing people

in the midst of actually disclosing depression, but given the ethical implications of

giving certain (i.e., low-quality) messages to people living with depression, future

Western Journal of Communication 157

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

research should tread lightly in seeking to maximize realism in this line of inquiry.

Second, although we asked participants in the second study to report their specific

diagnosis in medical terms (which would discourage individuals who had not been

diagnosed with depression from continuing), it is possible that some participants

may not have been honest about having depression. A further limitation is we did

not attempt to investigate the cognitive processes that may mediate the various effects

we found. This choice is consistent with D. O’Keefe’s (2003) recommendations that

message effects research is more fruitful when the focus is on connecting message fea-

tures to outcomes rather than attempting to define messages in terms of psychologi-

cal states. From this perspective, manipulation checks that assess psychological states

are irrelevant to the claims at hand (although, obviously, the study is limited in that it

does not make claims about psychological states, such as whether participants in

Study 1 actually recognized differences in how coping was variously portrayed).

Finally, the disclosure and response messages used in this research represented single

episodes, but it is likely that the interactants would continue discussing a depression

diagnosis over time. Although we would expect the initial reaction to disclosure to

remain important, examining later responses to depression disclosure (e.g., people

who respond poorly at first but become more supportive later) would deepen our

understanding of the process of depression disclosure.

In conclusion, disclosing a diagnosis of depression and responding to such a dis-

closure present unique communication challenges. The current research demon-

strates that the nature of a depression disclosure message can influence the

sophistication of the response and that responses that effectively attend to multiple

goals are higher in quality than messages that do not manage multiple goals. Con-

tinuing to further our understanding of the relationship between variation in illness

disclosure messages and variation in responses to such disclosure will help communi-

cation scholars in developing recommendations for people who would disclose their

depression and for those who would respond to important disclosures.

References

Barney, L. J., Griffiths, K. M., Jorm, A. F., & Christensen, H. (2006). Stigma about depression and

its impact on help-seeking intentions. Australian and New Zealand Journal of Psychiatry, 40,

51–54. doi:10.1111=j.1440–1614.2006.01741.x Berger, C. R. (2005). Interpersonal communication: Theoretical perspectives, future prospects.

Journal of Communication, 55, 415–447. doi:10.1111=j.1460–2466.2005.tb02680.x Brunig, J. L., & Kintz, B. L. (1997). Computational handbook of statistics (4th ed.). New York, NY:

Cambridge University Press.

Burleson, B. R. (1994). Comforting communication: Significance, approaches, and effects. In B. R.

Burleson, T. L. Albrecht, & I. G. Sarason (Eds.), Communication of social support: Messages,

interactions, relationships, and community (pp. 3–28). Thousand Oaks, CA: Sage.

Burleson, B. R. (2003). Emotional support skills. In J. O. Greene & B. R. Burleson (Eds.), Handbook

of communication and social interaction skills (pp. 551–594). Mahwah, NJ: Erlbaum.

Burleson, B. R., & Caplan, S. E. (1998). Cognitive complexity. In J. C. McCroskey, J. A. Daly, & M.

M. Martin (Eds.), Communication and personality: Trait perspectives (pp. 230–286). Cresskill,

NJ: Hampton Press.

158 A. M. Scott et al.

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

Burleson, B. R., & Waltman, M. S. (1988). Cognitive complexity: Using the Role Category

Questionnaire. In C. H. Tardy (Ed.), A handbook for the study of human communication:

Methods and instruments for observing, measuring, and assessing communication processes

(pp. 1–35). Norwood, NJ: Ablex.

Canary, D. J., & Spitzberg, B. H. (1987). Appropriateness and effectiveness of perceptions of conflict stra-

tegies. Human Communication Research, 14, 93–118. doi:10.1111=j.1468–2958.1987.tb00123.x Caughlin, J. P. (2010). A multiple goals theory of personal relationships: Conceptual integration

and program overview. Journal of Social and Personal Relationships, 27, 824–848.

doi:10.1177=0265407510373262 Caughlin, J. P., Brashers, D. E., Ramey, M. E., Kosenko, K. A., Donovan-Kicken, E., & Bute, J. J.

(2008). The message design logics of responses to HIV disclosures. Human Communication

Research, 34, 655–684.

Caughlin, J. P., Bute, J. J., Donovan-Kicken, E., Kosenko, K. A., Ramey, M. E., & Brashers, D. E.

(2009). Do message features influence reactions to HIV disclosures? A multiple-goals

perspective. Health Communication, 24, 270–283. doi:10.1080=10410230902806070 Charmaz, K. (1991). Good days, bad days: The self in chronic illness and time (Chapter 5: Disclosing

illness, pp. 107–133). New Brunswick, NJ: Rutgers University Press.

Clark, R. A., & Delia, J. G. (1979). Topoi and rhetorical competence. Quarterly Journal of Speech, 65,

187–206. doi:10.1080=00335637909383470 Cohen Silver, R., Wortman, C. B., & Crofton, C. (1990). The role of coping in support provision:

The self-presentational dilemma of victims of life crises. In B. R. Sarason, I. G. Sarason, &

G. R. Pierce (Eds.), Social support: An interactional view (pp. 397–426). New York, NY: Wiley.

Coyne, J. C. (1976). Toward an interactional description of depression. Psychiatry, 39, 28–40.

Crockett, W. H. (1965). Cognitive complexity and impression formation. In B. A. Maher (Ed.),

Progress in experimental personality research (Vol. 2, pp. 37–90). New York, NY: Academic Press.

Crockett, W. H., Press, A. N., Delia, J. G., & Kenny, C. T. (1974). Structural analysis of the organi-

zation of written impressions. Unpublished manuscript, University of Kansas.

Dennis, C-L., & Chung-Lee, L. (2006). Postpartum depression help-seeking barriers and maternal

treatment preferences: A qualitative systematic review. Birth, 33, 323–331. doi:10.1111= j.1523–536X.2006.00130.x

Dillard, J. P. (1990). A goal-driven model of interpersonal influence. In J. P. Dillard (Ed.), Seeking

compliance: The production of interpersonal influence messages (pp. 41–56). Scottsdale, AZ:

Gorsuch Scarisbrick.

Fleiss, J. L. (1981). Statistical methods for rates and proportions (2nd ed.). New York, NY: Wiley.

Furr, S. R., Westefeld, J. S., McConnell, G. N., & Jenkins, J. M. (2001). Suicide and depression

among college students: A decade later. Professional Psychology: Research and Practice, 32,

97–100. doi:10.1037==0735–7028.32.1.97 Garcia, J. A., & Crocker, J. (2008). Reasons for disclosing depression matter: The consequences of

having egosystem and ecosystem goals. Social Science and Medicine, 67, 453–462.

doi:10.1016=j.socscimed.2008.03.016

Goldsmith, D. J. (2004). Communicating social support. New York, NY: Cambridge University Press.

Goldsmith, D. J., & Fitch, K. A. (1997). The normative context of advice as social support. Human

Communication Research, 23, 454–476. doi:10.1111=j.1468–2958.1997.tb00406.x Goldsmith, D. J., Lindholm, K. A., & Bute, J. J. (2006). Dilemmas of talking about lifestyle changes

among couples coping with a cardiac event. Social Science and Medicine, 63, 2079–2090.

doi:10.1016=j.socscimed.2006.05.005

Goldsmith, D. J., McDermott, V. M., & Alexander, S. C. (2000). Helpful, supportive, and sensitive:

Measuring the evaluation of enacted social support in personal relationships. Journal of

Social and Personal Relationships, 17, 379–392. doi:10.1177=0265407500173004 Greene, K., Derlega, V. J., Yep, G. A., & Petronio, S. (2003). Privacy and disclosure of HIV in inter-

personal relationships: A sourcebook for researchers and practitioners. Mahwah, NJ: Erlbaum.

Western Journal of Communication 159

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

Jackson, S. (1992).Message effects research: Principles of design and analysis. New York, NY: Guilford

Press.

Jackson, S., & Brashers, D. E. (1994). Random factors in ANOVA. Thousand Oaks, CA: Sage.

Lambert, B. L., & Gillespie, J. L. (1994). Patient perceptions of pharmacy students’ hypertension

compliance-gaining messages: Effects of message design logic and content themes. Health

Communication, 6, 311–325. doi:10.1207=s15327027hc0604_6 Michalak, E. E., Yatham, L. N., Kolesar, S., & Lam, R. W. (2006). Bipolar disorder and quality of

life: A patient-centered perspective. Quality of Life Research, 15, 25–37. doi:10.1007=s11136– 005-0376–7

Müller, A. D., Bode, S., Myer, L., Stahl, J., & von Steinbüchel, N. (2011). Predictors of adherence to

antiretroviral treatment and therapeutic success among children in South Africa. AIDS Care,

23, 129–138. doi:10.1080=09540121003758523 O’Keefe, B. J. (1988). The logic of message design: Individual differences in reasoning about com-

munication. Communication Monographs, 55, 80–103. doi:10.1080=03637758809376159 O’Keefe, B. J. (1990). The logic of regulative communication: Understanding the rationality of

message designs. In J. P. Dillard (Ed.), Seeking compliance: The production of interpersonal

influence messages (pp. 87–104). Scottsdale, AZ: Gorsuch Scarisbrick.

O’Keefe, B. J., Lambert, B. L., & Lambert, C. (1997). Conflict and communication in a research and

development unit. In B. D. Sypher (Ed.), Case studies in organizational communication 2:

Perspectives on contemporary work life (pp. 31–52). New York, NY: Guilford Press.

O’Keefe, B. J., & McCornack, S. A. (1987). Message design logic and message goal structures: Effects

on perceptions of message quality in regulative communication situations. Human

Communication Research, 14, 68–92. doi:10.1111=j.1468–2958.1987.tb00122.x O’Keefe, B. J., & Shepherd, G. J. (1987). The pursuit of multiple objectives in face-to-face persuasive

interactions: Effects of construct differentiation on message organization. Communication

Monographs, 54, 396–419. doi:10.1080=03637758709390241 O’Keefe, B. J., & Shepherd, G. J. (1989). The communication of identity during face-to-face

persuasive interactions: Effects of perceiver’s construct differentiation and target’s message

strategies. Communication Research, 16, 375–404. doi:10.1177=009365089016003004 O’Keefe, D. J. (2003). Message properties, mediating states, and manipulation checks: Claims, evi-

dence, and data analysis in experimental persuasive message effects research. Communication

Theory, 13, 251–274. doi:10.1111=j.1468–2885.2003.tb00292.x Peterson, L. W., & Albrecht, T. L. (1996). Message design logic, social support, and mixed-status

relationships. Western Journal of Communication, 60, 291–309. doi:10.1080= 10570319609374551

Pratt, L. A., & Brody, D. J. (2008, September). Depression in the United States household population,

2005–2006 (Centers for Disease Control and Prevention National Center for Health Statistics

Data Brief No. 7). Washington, DC: U.S. Department of Health and Human Services.

Rodriguez, A. J., Holleran, S. E., & Mehl, M. R. (2010). Reading between the lines: The lay

assessment of subclinical depression from written self-descriptions. Journal of Personality,

78, 575–598. doi:10.1111=j.1467–6494.2010.00627.x Samter, W., Burleson, B. R., & Basden-Murphy, L. (1989). Behavioral complexity is in the eye of the

beholder: Effects of cognitive complexity and message complexity on impressions of the

source of comforting messages. Human Communication Research, 15, 612–629.

doi:10.1111=j.1468–2958.1989.tb00201.x U.S. Department of Health and Human Services. (1999). Mental health: A report of the Surgeon

General. Rockville, MD: U.S. DHHS.

Wilson, S. R. (2002). Seeking and resisting compliance: Why people say what they do when trying to

influence others. Thousand Oaks, CA: Sage.

World Health Organization. (2001). The world health report 2001: Mental health: New understand-

ing, new hope. Geneva, Switzerland: World Health Organization.

160 A. M. Scott et al.

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

Appendix A: Disclosure Messages

‘‘Imagine that you are talking alone with a good friend. Please think of a particular

friend. Now, imagine that after a few minutes of talking, your friend says ‘Well, I have

something to tell you. . .

1. Kernel (no information about coping)

a. . . . I just got diagnosed with depression a few weeks ago.’’’

b. . . .A few weeks ago I found out I have depression.’’’

c. . . .My doctor just diagnosed me with depression a few weeks ago.’’’

d. . . .A few weeks ago my doctor told me I have depression.’’’

2. Self-presenting poor coping

a. . . . I just got diagnosed with depression a few weeks ago. I’m having a really

hard time dealing with all of this.’’’

b. . . .A few weeks ago I found out I have depression. It’s been really hard for me

to relate to people. I feel pretty upset about the whole situation.’’’

c. . . .My doctor just diagnosed me with depression a few weeks ago. I feel

like things are just so unfair and this is making it really hard for me to have

a positive attitude about anything.’’’

d. . . .A few weeks ago my doctor told me I have depression. I just feel like it’s so

hard for me to enjoy life or feel hopeful about anything. I still don’t feel like

it’s even worth it to get up in the morning.’’’

3. Self-presenting good coping

a. . . . I just got diagnosed with depression a few weeks ago. I feel like things

are actually going really well. I’m much more aware of the good things in

life.’’’

b. . . .A few weeks ago I found out I have depression, and I feel like I’m manag-

ing everything really well. This has made me value everything in life a lot more

than I did before.’’’

c. . . .My doctor just diagnosed me with depression a few weeks ago, and I feel

like a cloud has lifted. I’m starting to feel really good again.’’’

d. . . .A few weeks ago my doctor told me I have depression. I feel like in a way

this has improved the quality of my life. I’m trying to live each day to the

fullest now.’’’

4. Self-presenting balanced coping

a. . . . I just got diagnosed with depression a few weeks ago. Some days I feel okay

about it and other days are a little harder. I guess all I can do is keep trying.’’’

b. . . .A few weeks ago I found out I have depression. I sometimes wonder why

this happened to me, but I’m trying to make the best of the situation. Some

days that’s easier to do than others.’’’

c. . . .My doctor just diagnosed me with depression a few weeks ago. It’s been

one of those things where sometimes I feel like I’m spinning out of control,

but then I have days where I feel like I’m managing everything well.’’’

Western Journal of Communication 161

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

d. . . .A few weeks ago my doctor told me I have depression. On the one

hand, this has been a really hard thing to deal with, but on the other

hand, I feel like I’m learning some really valuable lessons from going through

this.’’’

5. Explicit support-seeking

a. . . . I just got diagnosed with depression a few weeks ago, and I think I’m really

going to need your help with all of this.’’’

b. . . .A few weeks ago I found out I have depression, and I’m hoping I can count

on you for support with everything.’’’

c. . . .My doctor just diagnosed me with depression a few weeks ago, and I’m

telling you because I really need your support.’’’

d. . . .A few weeks ago my doctor told me I have depression, and I’m telling you

because I know I can come to you for help.’’’

Appendix B: Response Messages

‘‘Imagine that you just told one of your good friends that you have been diagnosed

with depression for the first time. Imagine that your friend reacts in the way

described below:

1. Expressive responses

a. ‘How so? What makes you think that in particular?’’’

b. ‘Oh no. You have seemed okay lately. What is causing it?’’’

c. ‘Well, what are they going to do about it?’’’

d. ‘You are always depressed, though. What’s the difference?’’’

e. ‘Are you okay? Did you go to a doctor? I didn’t know you weren’t feeling

okay.’’’

f. ‘Oh. How long have you been feeling like this? Why didn’t you say anything

to me?’’’

2. Conventional responses

a. ‘I’m so sorry to hear that. Well, we need to focus on ways to keep your spirits

high. Do you know what makes you sad?’’’

b. ‘I’m sorry to hear that. How long have you been feeling this way? Is there any-

thing I can do to help?’’’

c. ‘Is there anything that I need to know about that you haven’t told me other

than that? What do you need from me that can help you?’’’

d. ‘Really? Why do you think you were diagnosed with depression? Is something

wrong? Well, if you need anything, let me know for sure.’’’

e. ‘Really? I’m so sorry. Do you want to talk about it? Did the doctor say any

specifics as to what is causing your depression?’’’

f. ‘What’s on your mind? You know that you can talk with me. Dude, seriously

what’s up?’’’

162 A. M. Scott et al.

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13

3. Rhetorical responses

a. ‘Oh my gosh. Are you serious? You know I’m always here for you if you need

anything or if you want to talk. I know this is a hard time for you, and I want

the best for you. We can work through this together.’’’

b. ‘How are you feeling now? I’m really glad you decided to tell me this. I just

want you to know that no matter what happened, I’ll always be by your side.

This won’t change our relationship and you can always come to me. Is there

anything I can do for you right now?’’’

c. ‘Wow, that’s kind of scary. But the good thing about it is you’ve been diag-

nosed, which means that you’ll start getting treatment. If you had just gone

through the rest of your life without knowing, there’s no telling what would

have happened. You might have hit a real low, and you wouldn’t have known

how to get out of it. I know you know, but I’m going to tell you anyway: I’m

here for you, and if you need to talk about anything I’m here. Even if you’re

just feeling sad and you don’t know why, I’ll listen to you while you sort it

out. You know I love you, and you’re going to be alright, seriously.’’’

d. ‘I am glad that you were willing to tell me that. I appreciate the fact that you

want me to know. I think that you being willing to go to the doctor and ask

for help is a good decision. They will be able to get you help and help you

manage those feelings with the proper care. You know that you can come

to me and I will try to help too, or at least be a good person to listen.’’’

e. ‘I’m sorry this happened. Is there anything I can do? I’ll be here for any

emotional, physical, mental support you may need. Just know that I’ll always

still be your friend throughout this time, no matter what happens.’’’

f. ‘I’m so sorry. I cannot even imagine how difficult that was to hear. Don’t

worry, though. We’ll get through this together. I’ll be here for you no matter

what. If you need anything—seriously anything—just let me know.’’’

Western Journal of Communication 163

D ow

nl oa

de d

by [

W es

te rn

I lli

no is

U ni

ve rs

ity ]

at 0

6: 53

1 9

N ov

em be

r 20

13