Discussion: Correlation and Bivariate Regression

TheNatureContactQuestionnaire.pdf

Home >Mathematics homework help >Statistics homework help >Discussion: Correlation and Bivariate Regression

Work 40 (2011) 411–423 411 DOI 10.3233/WOR-2011-1253 IOS Press

The Nature Contact Questionnaire: A measure of healthy workplace exposure

Erin Largo-Wighta,∗, W. William Chenb, Virginia Doddc and Robert Weilerb aDepartment of Public Health, University of North Florida, Gainesville, FL, USA bDepartment of Health Education and Behavior, University of Florida, Jacksonville, FL, USA cDepartment of Community Dentistry and Behavioral Sciences, University of Florida, Gainesville, FL, USA

Received 5 December 2009

Accepted 4 April 2010

Abstract. Understanding and promoting healthy workplaces is an important and growing area of interest in occupational health. Nature contact is a central component to the study of and promotion of healthy places. Previous findings suggest that nature contact influences health via stress appraisal process. Currently, there are no known comprehensive valid and reliable measures of nature contact, which presents obstacles to research and worksite health promotion. Objective: This study was designed to develop and test an instrument to measure nature contact at work, entitled the Nature Contact Questionnaire (NCQ), 16-item self-reported checklist to measure actual exposure. Participants: A sample of 503 (30% response rate) office staff completed the questionnaire. Methods: Office staff were sent an email with a link to the electronic survey twice, two weeks apart. Results: Content and construct validity (KMO = 0.68), internal consistency (Alpha = 0.64), and test-retest reliability (r = 0.85, p < 0.01) were established. Conclusions: The NCQ is the first known comprehensive, reliable and valid survey to measure nature contact, which allows research to compare forms of nature contact to best inform practice and design of healthy places.

Keywords: Employee stress, healthy places, work environment, nature contact, worksite health promotion

1. Introduction

The causal relationship between stress and the three leading causes of death in the US has been well doc- umented [2,9,35]. The abundance of stress-health ev- idence points to two physiological mechanisms. First, perceived stress results in nervous and endocrine sys- tem arousal through neural and chemical pathways. Increased heart rate, respiratory changes, and overall physiological mobilization occur seconds after a threat is perceived [3,57]. Second, perceived stress has been shown to suppress the immune system [10] through both cellular and humeral immune responses such as

∗Address for correspondence: Erin Largo-Wight, PhD, CHES, Department of Public Health, University of North Florida, 1 UNF Drive, Jacksonville, FL 32224, USA. Tel.: +1 904 620 2037; Fax: +1 904 620 1035; E-mail: [email protected].

decreased natural killer (NK) cell activity [37] and sec- retary immunologlobin A (S-IgA) levels [57]. The out- comes of chronic stress (and related co-morbidities) has societal impacts on health care costs and produc- tivity [2,51].

Americans today are more stressed than in the past and work is attributed as a major cause [28,51]. Of- fice staff, in particular, are a priority stress population according to the demand-control model and previous findings with high pscychological demands and low re- sources [44,45]. In addition, office employees repre- sent more than 70% of the American workforce [46]. Stress is a critical health issue among office staff.

1.1. Healthy workplaces

The study and promotion of healthy workplaces is important. An unhealthy workplace appears to cause

412 E. Largo-Wight et al. / The Nature Contact Questionnaire: A measure of healthy workplace exposure

poor employee health through biological and stress- related pathways. Biologically, building-related air pollutants and toxins can promote illnesses and respira- tory conditions, such as Sick Building Syndrome [38, 58]. Office environments with “environmental psycho- logical demands”, such as crowding and noise at work, increase the likelihood of perceived stress [7,29,42,49, 55].

Healthy workplaces, on the other hand, are free of these unhealthy biological and stress-related charac- teristics. In addition, healthy offices are places with health-promoting qualities such as the availability of healthy choices, safety, environmental sustainability and stewardship, and the opportunity for nature con- tact at work [17,19,20,23,24,32,62,67]. Understanding and promoting healthy workplaces is a critical aspect of effective worksite health promotion. “Environmen- tal modifications are believed to be important additions to worksite health promotion programs (WHPPs)” [17, p. 61]. It is now widely believed that worksite health promotion should go beyond education and include en- vironmental changes.

1.2. Nature contact at work

This study was designed to examine nature contact, one important component of a healthy place, at work. There is a vast literature that demonstrates the rela- tionship between nature contact and human health out- comes [24]. Nature contact has been shown to be health-promoting in many settings, including office set- tings. In general, nature contact is achieved when an individual is exposed to natural elements. Nature con- tact has been defined as the interaction between humans and animals, plants, landscape views, and outdoors [21, 24].

Outdoor work breaks, a form of nature contact at work, has previously been studied. Researchers found that relaxing and enjoying nature was related to less work stress compared to active leisure such as chal- lenging, exciting, and goal-driven activities. “Orga- nizations, if they are committed to employee health and well-being, may now need to consider the benefits of providing recuperative-therapeutic opportunities in much the same way as they have considered and pro- vided opportunities for employees to engage in social, sporting, and health-related activities” [66, p. 100].

Indoor plants in the office, a type of nature contact, have also been studied. In addition to serving as a “flexible and attractive biofiltration system. . . that can be used in any indoor space” [71, p. 7], indoor plants

may reduce stress among employees. Larsen and oth- ers [41] tested the effects of plants in an office space on health outcomes and found that participants in the high plant condition (22 plants in 140 square feet office space) reported greater well-being, greater perceived office attractiveness, and comfort than the other condi- tions [41].

Office windows provide the employee with the op- portunity to look outdoors and have natural sunlight; both of which are examples of nature contact at work. Some researchers have commented that office windows are valued status symbols because people and soci- eties intuitively understand the health benefits of win- dows [31,32]. Others proposed that windowless rooms in public places should be outlawed because they make occupants suffer [36]. Findings suggest that people prefer rooms with windows that cover about a quarter of the wall space [13].

Previous findings indicate that brief window view- ings of natural elements at home and work were as- sociated with wellbeing and stress reduction [31,32]. One of the first studies on the heath-related importance of outdoor views took place in a prison. Moore [48] found that prisoners with an outside view from their cell window of rolling farmland and trees had 24% less sick visits than prisoners with a courtyard view from their cell window over the 11 year study. There were no other differences in cell design and prisoners were randomly assigned cells upon entry. Although an office is not a prison cell, these findings may inform healthy worksites because occupants of both spend day-time hours in the relative settings. In addition to the out- door view, windows provide sunlight. Sunlight is a pre- ferred, attractive, and practical source of light for work environments [6]. Beauchemin and Hays [4] exam- ined the relationship between bright natural light rooms (direct sunlight from window) and dim natural light rooms (structure or position preventing direct sunlight from window) and found that health outcomes were significantly better for those in the bright room.

The health promoting mechanism of nature contact has been explained by natural scientists and theoritsts. E.O. Wilson’s evolutionary-based biophilia hypothesis contends that nature is innately relaxing, healing, and stress relieving for present-day humans because of the positive relationship related to survival between man and nature in primitive times [70]. Similarly, envi- ronmental restoration theories contend that nature con- tact is healthful because it restores cognitive resources necessary to cope with stress, thus reducing stress and protecting against stress-related health aliments [33,34, 68].

E. Largo-Wight et al. / The Nature Contact Questionnaire: A measure of healthy workplace exposure 413

1.3. Significance

Nature contact at work is important to study. In contrast to other important factors that influence per- ceived stress, such as personality, problem solving abil- ity, or job demands, enhancing nature contact at work is relatively easy, simple, and practical. Adding a plant or nature photography to one’s office or going outside for a work break instead of the break room are simple ways to increase nature contact, which have implica- tions for reducing stress and promoting stress-related health [24]. Cultivating a healthy workplace with pur- poseful opportunities of nature contact is an important and practical approach to reduce employee stress and promote health.

Although it is apparent that nature contact is an im- portant aspect of a healthy workplace, there is a need for future research in measurement. Nature contact can take many forms and previous researchers have opera- tionalized nature contact differently. For example, re- searchers have measured nature contact in the follow- ing ways: exercise on nature trails (e.g. [5]), plants in a workplace (e.g. [41]), fish tank in a dentist wait- ing room (e.g. [67]), windows and natural light in a classroom (e.g. [39]), nature CDs and photography in a hospital (e.g. [14]), and gardening (e.g. [8]). In addi- tion, researchers have measured perceived restorative qualities of natural settings [26].

However, there is no instrument for measuring an individual’s actual exposure to nature contact at work (or any setting). Currently, no published comprehen- sive nature contact measurements are available. Future research is needed in instrument development and test- ing. A comprehensive measure of nature contact would address an important gap in the literature and enable the research comparison of nature contact forms to inform practice recommendations. It is important to develop and test an instrument to enable future occupational health researchers to study the effects of the workplace on employee stress. A nature contact instrument would allow future researchers to answer important research questions.

First, a nature contact instrument would enable re- searchers to assess cross-sectional relationships be- tween nature contact at work and stress and stress- related health outcomes. For example, researchers may examine the relationship between the amount of nature contact at work and employee stress. The following re- search question could be addressed: “Is nature contact at work associated with employee stress and or health outcomes?” These findings, including the significance

and the effect size of the relationship, would be espe- cially important to worksite health promoters working to cultivate a healthy workplace.

Second, a nature contact instrument would allow researchers to assess the efficacy of a workplace- improvement intervention. The following research question, for example, could be addressed with a na- ture contact measurement taken pre and post a behav- ior change intervention: “Did an environmental edu- cation program result in healthier workplace exposures among employee participants?” These research find- ings would be especially important to health promoters working on workplace improvement efforts.

Third, a comprehensive nature contact instrument would allow researchers to make comparisons about workplace exposures. Kuo emphasized the need to de- velop a comprehensive measure of nature contact in order to study “which forms or doses of nature en- hance effectiveness and which do not” [40, p. 9]. Re- searchers may examine the relationships among differ- ent forms of nature contact at work on employee stress and health. For example, the following research ques- tion could be addressed with a comprehensive measure of nature contact at work: “Which nature contact work exposure (outdoor work breaks, indoor plants, or nature photography) has the strongest correlation to employ- ee stress?” This comparison would be especially im- portant to health promoters trying to prioritize efforts by determining if one form of nature contact is more healthful than another.

1.4. Purpose

This study was designed to develop and test a com- prehensive, reliable, and valid survey to measure ac- tual nature contact experienced at work. The survey was designed to quantify and count the number of ob- jective nature exposures, experiences, and features at work in a given week. We classified nature contact into three forms or categories to enable research compar- isons and inform practice recommendations. Outdoor nature contact was defined as interaction with natural elements outdoors, such as gardening, spending time in an urban park [8,25]. Indoor nature contact was de- fined as contact with natural elements within a building, such as landscape view from window, natural light, live plants, fish tank [31,42]. Indirect nature contact was defined as interaction with abstract representations of natural elements within a building, such as photographs of a natural landscape and recorded nature sounds [14].

414 E. Largo-Wight et al. / The Nature Contact Questionnaire: A measure of healthy workplace exposure

2. Methods

2.1. Sample

A census of office staff at a Southeastern University were invited to participate in the study (N = 1,622). Inclusion criteria required participants to be full-time, mostly desk-bound office staff. Thirteen job codes, such as secretary and office clerk, fit the criteria and were invited to participate. Desk bound office staff were chosen because they have stable work environ- ments, which is important for reliable measurement. Electronic informed consent was obtained from all par- ticipants; participation was anonymous and voluntary.

2.2. Item development

The first step in item development was reviewing the multidisciplinary literature on the health benefits of na- ture contact among all populations. Most of nature con- tact’s health benefits may be construed as stress-related such as clinical recovery, less incidences of illness, or restored cognitive abilities [24,33,68]. Therefore, the item development and selection literature review fo- cused on nature contact that was related to stress and stress-related outcomes. Nature contact that has previ- ously been shown to be health promoting or stress re- ducing was included. Each survey item was supported by direct or closely related previous findings supporting its health impact.

Devised and modified items were organized into one of the three categories of nature contact at work defined by the researchers. In this study, we classified nature contact at work into three categories for organization and comparison: outdoor (outdoor breaks, lunch, de- livery); indoor (window view, natural light, fish tank, live plants); and indirect (artificial plants, nature sounds on CD, nature photography or art).

The literature was reviewed to determine how na- ture contact has previously been measured. Although there was no developed tool to measure everyday nature contact, researchers have developed individual ques- tion items to measure specific forms of nature contact such as availability of windows, quality of view, and number of indoor plants. Appropriate, previously used items that measured nature contact were adopted and modified [1,14,31,32,39,41,67,69]. Additional items were devised to measure actual nature contact previ- ously shown to promote health.

2.3. Scaling responses

The survey was designed to quantify the actual and objective contact with nature features and experiences at work. Continuous response options were used be- cause numerical data offers the most precise measure- ment to reduce possible measurement error [15]. “Ze- ro, one, two, three, four, and five or more” were the re- sponse options for part one of the NCQ. Non-applicable (“N/A”) was also used for the four outdoor nature con- tact items to represent no lunch anywhere, no breaks anywhere, and no delivery anywhere. The remaining response options for items related to exposure were ranges of percentages including “0, 1–20%, 21–40%, 41–60%, 61–80%, and 81–100%.”

2.4. Scoring

The final instrument designed to quantify nature con- tact at work was titled Nature Contact Questionnaire (NCQ) (Appendix A). The NCQ was comprised of 16 items with six numerical response options. The NCQ assessed three forms of nature contact at work including indirect, indoor, and outdoor nature contact. The NCQ results in a total nature contact score and three subtotals for each of the forms of nature contact.

To calculate subtotals and totals, the participant’s continuous responses were treated as ordinal data and assigned a value. “Zero” and “0%” categories of re- sponses represent no nature contact and were assigned a value of 1. “One” and “1–20%” responses represent low nature contact and were assigned a value of 2 and so on. The “N/A” response option was available for the four outdoor nature contact items and was assigned a value of 0. The “N/A” suggested that the participant did not eat lunch at all in any environment, whereas the “zero” response option suggested that the participant did not eat lunch outside.

Four items, items a-d, represented outdoor nature contact; five items, items e, g-i, m-n, represented in- door nature contact; and five items, items f, j-l, o-p, represented indirect nature contact. The subtotals were calculated by adding the assigned ordinal response op- tion value for each of the associated domain items. The summed scores represented the nature contact subtotal where higher numbers represent more nature contact in that form or domain. (Note: If future researchers plan to compare forms of nature contact, a conversion factor needs to be applied to outdoor nature contact be- cause it has one less item than the other two subtotals [Conversion factor: X*4 = 5 = 1.25)]. The outdoor

E. Largo-Wight et al. / The Nature Contact Questionnaire: A measure of healthy workplace exposure 415

nature contact subtotal should be multiplied by the con- version factor in order to compare the subtotals). Total nature contact score was calculated by adding the three subtotals.

2.5. Data collection and analysis

Data were collected using a cross-sectional, web- based survey design. We sent an email invitation along with the web-link to access the online survey to the census [12,53,54]. The participants took approximate- ly 10–15 minutes to complete the online survey. We repeated the data collection two weeks later to assess the Nature Contact Questionnaire test-retest reliabili- ty. Data analysis procedures were conducted using the Statistical Software for Social Sciences (SPSS) version 16. Content and face validity, internal consistency, test- retest reliability, and construct validity were assessed.

3. Results

3.1. Sample characteristics

The majority of the participants were women (92.9%) and whites (82.5%). The mean age of the par- ticipants was 42 years old with a standard deviation of 12 years. Approximately half of the participants attend- ed some college or technical school (47.5%), reported earning between $25,001–$35,000 per year (49.5%), and reported being married (54.4%).

The response rate was about 30% (N = 503) for the first data collection. Of those responders, 83% partic- ipated in the second data collection for test-retest re- liability analysis of the Nature Contact Questionnaire (N = 401, 25% of total census). There were no signif- icant socio-demographic differences in second survey responders and non-responders.

3.2. NCQ profile

The instrument was designed to measure total nature contact and three subscales of nature contact, outdoor, indoor, and indirect. The sub-scores and overall total score were calculated by converting the continuous re- sponses into ordinal responses and adding the respons- es. See Table 1 for a summary of the nature contact scores with higher numbers representing more nature contact at work.

The frequencies and percentages of each of the 16 na- ture contact items with the actual continuous response

Table 1 Nature contact questionnaire (NCQ) scores∗

NCQ score Minimum Maximum M sd

Outdoor Nature Contact 0.0∗ 24.0 7.6 3.7 Indoor Nature Contact 6.0 26.0 12.0 5.6 Indirect Nature Contact 6.0 26.0 9.7 3.9

Total Nature Contact 14.0 63.0 29.3 8.6 ∗Expressed as M mean and sd standard deviation.

options are reported in Tables 2 through 4. The fre- quencies and percentages of outdoor nature contact re- sponse options, with responses ranging from 0–5 or more including non-applicable (N/A), are presented in Table 2. The N/A response option was designated for participants who did not have a break, lunch, or deliv- ery task in any setting and was assigned a value of 0 for the ordinal conversion. The majority of the sample took breaks and ate lunch, but did not do so outside (59% and 67%, respectively). About a third of the par- ticipants took at least one weekly work break (35%) and ate lunch (30%) outdoors. A majority of the sam- ple reported at least one outdoor delivery task (62%) and 18% exercised outdoors during their lunch break at least once over the last week.

Indoor and indirect nature contact item profiles are reported in Tables 3 and 4. Table 3 displays frequencies and percentages of items with response options ranging from 0–5 or more. Indoor nature contact items with the 0–5 or more response option were e, g, h, and i and indirect nature contact items with the 0–5 or more re- sponse option were f, j, k, and l (Table 3). About half of the participants reported having at least one of the fol- lowing in their primary work space: live plant (53%), window that leads to the outdoors (53%), nature paint- ing (50%), and professional or personal photograph of nature (52%).

Table 4 displays frequencies and percentages of items with response options ranging 0–100% of the time. Indoor nature contact items with the 0–100% re- sponse option were m and n and indirect nature contact items with the 0–100% response option were o and p (Table 4). Over half of the sample reported zero per- centage of the day with sunlight (52%) and time with unobstructed view outside (54%). About 15% of the participants reported spending at least some time dur- ing the day watching nature images on a monitor or T.V. screen.

Three out of the 16 NCQ items were excluded from future analyses because over 95% of the responses fell in the zero response option. Fish tanks (h), animal/pets (i), and nature sounds (o) items were not included in the NCQ or future analyses. These items had very low

416 E. Largo-Wight et al. / The Nature Contact Questionnaire: A measure of healthy workplace exposure

Table 2 Outdoor NCQ items

Items N/A 0 1 2 3 4 5 or more

a. Outdoor work breaks N 28 234 37 30 19 13 38 Valid Percent∗ 7.0 58.6 9.3 7.5 4.8 3.3 9.5

b. Outdoor lunch breaks N 13 268 39 19 24 14 23 Valid Percent∗ 3.3 67.0 9.8 4.8 6.0 3.5 5.8

c. Outdoor exercise during lunch N 15 312 25 17 13 3 15 Valid Percent∗ 3.8 78.0 6.3 4.3 3.3 0.8 3.8

d. Outdoor delivery N 7 147 77 72 35 22 40 Valid Percent∗ 1.8 36.8 19.3 18.0 8.8 5.5 10.0

∗Adjusted for missing data.

Table 3 Indoor (e,g-i) and indirect (f,j-l) NCQ items with response options 0 to 5 or more

Items 0 1 2 3 4 5 or more

e. Live plants/flowers N 186 76 52 35 16 35 Valid Percent∗ 46.5 19.0 13.0 8.8 4.0 8.8

f. Artificial plants/flowers N 279 63 34 7 8 9 Valid Percent∗ 69.8 15.8 8.5 1.8 2.0 2.3

g. Windows that lead to outdoors N 187 106 38 24 15 30 Valid Percent∗ 46.8 26.5 9.5 6.0 3.8 7.5

h. Fish tank N 389 10 – 1 – – Valid Percent∗ 97.3 2.5 – 0.3 – –

i. Animals/pets N 392 4 2 – 1 1 Valid Percent∗ 98.0 1.0 0.5 – 0.3 0.3

j. Realistic nature paintings/art N 203 80 52 22 12 31 Valid Percent∗ 50.8 20.0 13.0 5.5 3.0 7.8

k. Professional or personal nature photography N 191 87 49 14 12 47 Valid Percent∗ 47.8 21.8 12.3 3.5 3.0 11.8

l. Realistic nature carvings/sculptures N 311 51 16 9 4 9 Valid Percent∗ 77.8 12.8 4.0 2.3 1.0 2.3

∗Adjusted for missing data.

variability in response distribution and were excluded from the future analysis [43].

Outdoor nature contact items (Table 2) were items a, b, c, and d. Based on the acceptable distribution of these responses, all of the items were included in future analysis. Indoor nature contact items (Tables 3–4) were items e, g, h, i, m, and n. Two of the six indoor nature contact items, fish tank (h) and animal (i) exposures, were excluded from future analyses because of the low variability in response distribution. Over 95% of the participants reported the zero response option to these items. Indirect nature contact items (Tables 3–4) were items f, j, k, l, o, and p. One of the six indirect nature

contact items, time spent listening to nature sounds in the office (o), was excluded from future analyses because of the low variability in response distribution. Over 95% of the participants reported the zero percent response option to this item.

3.3. NCQ content and face validity

Content and face validity was assessed by an expert panel. A seven member expert panel reviewed the orig- inal items and conducted a second review of the revised items. Quantitative and qualitative responses were col-

E. Largo-Wight et al. / The Nature Contact Questionnaire: A measure of healthy workplace exposure 417

Table 4 Indoor (m,n) and indirect (o,p) NCQ items with response options 0% to 100% of time

Items 0% 1–20% 21–40% 41–60% 61–80% 81–100%

m. Time with sunlight N 209 40 19 21 20 91 Valid Percent∗ 52.3 10.0 4.8 5.3 5.0 22.8

n. Time with unobstructed view outside N 217 34 10 13 15 111 Valid Percent∗ 54.3 8.5 2.5 3.3 3.8 27.8

o. Time listening to nature sounds N 382 10 7 1 – – Valid Percent∗ 95.5 2.5 1.8 0.3 – –

p. Time watching nature images on monitor N 342 38 12 2 2 4 Valid Percent∗ 85.5 9.5 3.0 0.5 0.5 1.0

∗Adjusted for missing data.

lected and analyzed to assess and improve NCQ’s con- tent and face validity.

Quantitative data was presented in the form of per- centage of experts that agree that the Nature Contact Questionnaire was focused, brief, clear, readable, and assuring. Each item, direction, and response option was assessed. Any percentage less than 70% was con- sidered problematic. The results indicated that most re- viewers believed the preliminary Nature Contact Ques- tionnaire was focused, brief, clear, readable, and as- suring. However, there were concerns reflected in the lower percentages related to item and direction clarity and adequacy of response options.

Qualitative data were classified based on a phe- nomenological approach; statements were clustered or grouped based on commonality [65]. The comments were grouped based on their content in the following areas: time frame of the questionnaire, response op- tions, wording and clarity, layout and formatting, con- tent, and scoring and data analyses.

The panel’s quantitative and qualitative feedback and suggestions were used to make modifications to the NCQ. Specifically, three of the nineteen items were eliminated, the response options were changed from a Likert-type scale to a continuous scale, the ques- tionnaire’s time-frame was changed from over the past month to the last week (Monday-Friday), and the di- rections and items were modified for improved clarity. A second review of the revised NCQ suggested there were concerns related to layout such as font size and color. Formatting and layout suggestions were used to make the final visual modifications. There were no expressed content concerns at the second review.

3.4. NCQ construct validity

Because outdoor, indoor, and indirect nature contact were sub-categories of nature contact conceptualized

by the researchers for this instrument and nature contact has never been previously measured comprehensively, it was important to establish construct validity. Prin- cipal component analysis (PCA) was used to collapse objective indicators of nature contact into a set of more parsimonious indices and to test if “outdoor, indoor, and indirect” were meaningful factors or indices within nature contact [56].

The Kaiser-Meyer-Olkin (KMO) measuring of sam- pling adequacy was 0.68 for NCQ items. Principal component analysis (PCA) with a varimax rotation method revealed three factors (Table 5). The factors were defined by a standard assignment rule including: a) inclusion criteria that items load at 0.6 or higher to the component, b) exclusion criteria that the item did not load more than 0.4 on any other component, and c) more than one loaded item is necessary for factor classification [18]. Highlighted items in Table 5 loaded on component.

Outdoor breaks, outdoor lunch, and outdoor exercise represent factor three. This factor closely reflects the hypothesized subtotal in this study titled outdoor na- ture contact. Number of windows, time spent with sun- light, and time with unobstructed view outdoors repre- sent factor one. This factor closely reflects the hypoth- esized subtotal in this study titled indoor nature con- tact. Nature photography, nature paintings, and nature carvings represent factor two. This factor closely re- flects the hypothesized subtotal in this study titled in- direct nature contact. Component four was not a factor based on the standard assignment rule, specifically the criteria that more than one item is necessary for factor classification.

3.5. NCQ test-retest reliability

A bivariate Pearson Product correlation between to- tal and subtotals at time I and time II were run to de-

418 E. Largo-Wight et al. / The Nature Contact Questionnaire: A measure of healthy workplace exposure

Table 5 Results of NCQ item principal component analysis

NCQ Item 1 2 3 4

Frequency of outdoor Break – – 0.81 – Frequency of outdoor lunch – – 0.69 −0.17 Frequency of outdoor exercise – 0.14 0.68 0.20 Frequency of outdoor delivery – 0.36 – −0.50 Number of live life plants 0.29 0.26 0.11 0.35 Number of artificial plants – 0.14 – 0.78 Number of windows leading outdoors 0.79 – – 0.14 Number of nature paintings – 0.71 – 0.29 Number of nature photography – 0.75 – −0.10 Number of nature carvings 0.12 0.62 – – Time with sunlight 0.90 – – – Time with unobstructed view 0.86 – – – Time viewing images on monitor – 0.43 0.39 0.23

Table 6 Test-retest reliability coefficients (r) and associated R2 for NCQ between time I (TI) and time II (TII)∗

NCQ Time I Time II r R2

M (sd) M (sd) TI – TII TI – TII

Outdoor Nature Contact Subtotal 7.6 (3.7) 7.9 (3.6) 0.75∗∗ 0.56 Indoor Nature Contact Subtotal 9.9 (5.6) 9.9 (5.7) 0.90∗∗ 0.81 Indirect Nature Contact Subtotal 8.6 (3.8) 8.7 (3.7) 0.79∗∗ 0.62 Total Nature Contact 26.2 (8.6) 26.5 (8.3) 0.84∗∗ 0.71 ∗Expressed as M mean, sd standard deviation. ∗∗p < 0.01 (two tailed).

termine the temporal stability of the survey items [52]. Because participants may remember their responses and try to emulate in order to appear consistent [16], we conducted the test-retest data collections two weeks apart to minimize this effect. See Table 6 for a summary of the NCQ’s reliability coefficients.

3.6. Internal consistency

Internal consistency was measured for NCQ total and the three subscales with Cronbach’s Alpha. See Table 4 for the NCQ’s total, outdoor subtotal, indoor subtotal, and indirect subtotal respective Alpha levels. Table 7 reports the internal consistency for the NCQ total and subtotals as they were developed based on theoretical assumptions and previous findings. The frequencies of outdoor breaks, outdoor delivery, outdoor exercise, and outdoor lunch were included in the outdoor nature con- tact subtotal analysis. The number of live plants, the number of windows, time with sunlight, and time with unobstructed outdoor view were included in the indoor nature contact subtotal analysis. The number of arti- ficial plants, nature photographs, paintings, carvings, and time spent watching nature images were included in the indirect nature contact subtotal analysis.

See Table 8 for the NCQ’s total, outdoor subtotal, indoor subtotal, and indirect subtotal respective Alpha levels based on the PCA findings conducted in this study. Table 5 reports the internal consistency for the NCQ total and subtotals based on the data reduction findings of the PCA. Ten of the 13 NCQ items load- ed on a factor and thus the 10 items were included in the internal consistency analyses reported in Table 8. Based on the loadings frequencies of outdoor breaks, outdoor exercise, and outdoor lunch were included in the outdoor nature contact subtotal analysis. The fre- quency of outdoor delivery did not load on the outdoor nature contact factor, or any factor, and was not includ- ed. The number of windows, time with sunlight, and time with unobstructed outdoor view were included in the indoor nature contact subtotal analysis. The num- ber of live plants did not load on the indoor nature con- tact factor and was not included. The number of nature photographs, paintings, and carvings were included in the indirect nature contact subtotal analysis. The num- ber of artificial plants and time spent watching nature images and listening to nature sounds did not load on the indirect nature contact factor, or any factor, and was not included.

E. Largo-Wight et al. / The Nature Contact Questionnaire: A measure of healthy workplace exposure 419

Table 7 Cronbach’s Alpha internal consistency based on theory∗

NCQ N Items M sd Variance Alpha

Outdoor Nature Contact Subtotal 4 8.0 3.8 14.5 0.47 Indoor Nature Contact Subtotal 4 9.8 5.6 31.4 0.73 Indirect Nature Contact Subtotal 5 8.8 3.9 15.2 0.56

Total Nature Contact 13 26.5 8.8 76.9 0.64 ∗Expressed as M mean, sd standard deviation.

Table 8 Cronbach’s Alpha internal consistency based on factor analysis∗

NCQ N Items M sd Variance Alpha

Outdoor Nature Contact Subtotal 3 5.6 3.4 11.8 0.61 Indoor Nature Contact Subtotal 3 7.5 5.0 25.2 0.81 Indirect Nature Contact Subtotal 3 5.9 3.2 10.5 0.58

Total Nature Contact 10 22.5 8.3 68.8 0.65 ∗Expressed as M mean, sd standard deviation.

4. Discussion

Studying the workplace is a critical aspect of occupa- tional health research. There is a growing recognition that cultivating healthy workplaces is an important as- pect of worksite health promotion [17]. Previous find- ings suggest that nature contact is health-promoting in many settings, including work. Although nature con- tact at work is an essential aspect of healthy work- places, there are reseach gaps that were addressed by this study. Until now, there was no comprehensive mea- sure of actual nature contact experienced at work (or any setting). This study was designed to develop and test a comprehensive measure of nature contact at work. The Nature Contact Questionnaire (NCQ), developed and tested in this study, will enable future occupational health researchers to study the effects of the workplace environment on employee stress. These findings will inform practical efforts to increase the opportunity for nature contact at work, an important aspect of healthy workplaces [24,40].

The findings suggest that the instrument developed for this study, the Nature Contact Questionnaire (NCQ), is a valid and reliable instrument to quantify actual na- ture contact in an office setting. The Nature Contact Questionnaire is the first known comprehensive, reli- able, and valid measure of nature contact. Because the NCQ is comprehensive with three subscales of nature contact, it also adds to applied and basic science by al- lowing researchers and practitioners to compare forms and doses of nature contact.

The NCQ’s validity indicator (KMO = 0.68) was above the standard of 0.5 for scales that have evi- dence of construct validity [56]. The principle com-

ponent analysis (PCA) revealed three distinct factors that corresponded almost identically to our hypothe- sized subscale categories of outdoor, indoor, and in- direct subtotals. These findings provided additional support for the proposed comprehensive nature contact construct validity as well as the hypothesized subcate- gories of nature contact. Although the PCA revealed three distinct factors congruent with our hypothesized subscales, there were several items that did not load on any of the factors as expected. It is not clear why these four items did not load on a NCQ factor. Fu- ture research could be conducted with special attention to the items that were expected to contribute based on previous findings, but did not load on any factor in this study. The wording of these items could be refined and further tested.

The NCQ’s test-retest reliability coefficient (r = 0.84) was far above the standard of 0.5 [63] and the conservative standard of 0.6 [11] for stable survey in- struments. Based on the findings, the NCQ can be considered stable over a two week time period.

The NCQ’s internal consistency (alpha = 0.64) was satisfactory based on standards for exploratory scales. Basic and applied research should reach an alpha be- tween 0.7 and 0.9 [50]. There are two potential con- clusions to these findings.

First, internal consistency may not be a relevant in- dicator of reliability for the NCQ. Internal consistency reliability (Alpha) is a measure of the similarity of in- strument’s items to each other as measured by average correlations. Internal consistency is not an appropri- ate measure “questionnaires that seek to obtain factual information about the occurrence of unrelated events or behaviors would not be theoretically amenable to

420 E. Largo-Wight et al. / The Nature Contact Questionnaire: A measure of healthy workplace exposure

internal consistency reliability assessment” [64, p. 3]. Since the NCQ measures objective counts of nature contact rather than a subjective construct or concept of nature contact, internal consistency may not be a relevant measure [26]. The best alternative reliability indicator when internal consistency is inappropriate is test-retest reliability [64], which far exceeded standards for the NCQ.

The second conclusion is that NCQ is an exploratory scale and its internal consistency needs improvement and further testing. One strategy to increase internal reliability is to conduct a study with a larger number of participants. It has been suggested that there should be at least 5 participants for every item on the scale, and at least 200 participants for exploratory scales regardless of the number of items [18]. Although this study satis- fied this requirement, increasing the number of partic- ipants over 500 may enhance internal consistency.

In addition, the response options may also be im- proved. Increasing the number of response options for each item has been shown to increase internal consis- tency. However, in this study, there were five to six response options for each item and the distribution of responses revealed that most participants reported low exposures to nature contact. Therefore, it may prove more beneficial to change the response options to a low- er frequency scale [18,43,50,59]. Future studies may also convert the continuous response options to nomi- nal response options. For example, the responses could be classified as either plants or no plants. This strate- gy may still pose a problem to internal consistency if over 70% of the responses fall into one of the nominal response options [43].

Internal consistency may also be enhanced by as- sessing office staff at other universities or other set- tings. All participants in this study were office staff at a southeastern university. Although there are many benefits to a well defined study population, a narrow population may pose range restriction limitations [50, 60]. Future studies could assess a more varied popula- tion to decrease this likelihood and to increase NCQ’s internal consistency.

Criterion validity is another area of future research. Criterion validity is used to measure external validi- ty by comparing the newly developed tool to a previ- ously developed valid measure [47,61]. Because there are no other comprehensive nature contact instruments for comparison, criterion validity could be assessed by randomly visiting participant’s office space to compare their responses to the researcher’s evaluation of the space. The researcher’s evaluation and the participant’s

responses could be compared for an indication of the NCQ’s criterion validity.

In addition, future research could also include mod- ifications and application in other settings. First, items could be added to the NCQ based on new evidence. For example, open windows could be an item of indoor nature contact. Future studies could address if hearing outdoor sounds or feeling outdoor air and breezes are healthful and stress reducing. NCQ item development was based on previous direct or closely related (i.e. ar- tificial plant) quantitative findings. Second, NCQ may be applied and tested in other settings beyond the work- place office. Environments such as classrooms, hospi- tal rooms, and homes are examples of locations where the current NCQ could be applied and tested in future work.

5. Conclusions

The NCQ is an important contribution to occupa- tional health research and worksite health promotion practice. The NCQ could be used by researchers to ex- amine the relationship between NCQ and stress scores and or related health outcomes among employees. Re- searchers could also use the NCQ to statistically com- pare forms and doses of workplace exposures on stress outcomes. These research findings would help work- site health promoters increase the opportunity for na- ture contact at work, an important aspect of healthy workplaces. The NCQ met the a priori standards for validity and reliability of survey instruments. Based on the findings, it the NCQ is the first comprehensive, reliable, and valid measure for healthy nature contact exposure at work among office staff.

References

[1] K. Allen, Are pets a healthy pleasure? The influence of pets on blood pressure, Current Directions in Psychological Science 12(6) (2003), 236–239.

[2] J. Banks, M. Marmot, Z. Oldfield and J.P. Smith, Disease and disadvantage in the United States and in England, Journal of American Medical Association JAMA 295(17) (2006), 2037– 2045.

[3] A. Baum and D.M. Posluszny, Healthy psychology: Map- ping biobehavioral contributions to health and illness, Annual Review of Psychology 50 (1999), 147–163.

[4] K.M. Beauchemin and P. Hays, Sunny hospital rooms expedite recovery from severe and refractory depression, Journal of Affective Disorders 40 (1996), 49–51.

[5] M. Bodin and T. Hartig, Does the outdoor environment matter for psychological restoration gained through running? Psy- chology of Sport and Exercise 4 (2003), 141–153.

E. Largo-Wight et al. / The Nature Contact Questionnaire: A measure of healthy workplace exposure 421

[6] P. Boyce, C. Hunter and O. Howlet, The benefits of daylight through windows, (U.S. Department of Energy Award No. DE-FC26-02NT41497). Troy, NY: Lighting Research Center, 2003.

[7] A. Brennan, J.S. Chugh and T. Kline, Traditional versus open office design: A longitudinal field study, Environment and Behavior 34(3) (2002), 279–299.

[8] B. Cimprich, Development of an intervention to restore atten- tion in cancer patients, Cancer Nursing 16(2) (1993), 83–92.

[9] S. Cohen, E. Frank, W.J. Doye D.P. Dkoner, B.S. Rabin and J.M. Gwaltney, Types of stressors that increase susceptibility to the common cold in healthy adults, Health Psychology 17(3) (1998), 214–223.

[10] S. Cohen, G.E. Miller and B.S. Rabin, Psychological stress and antibody response to immunization: A critical review of the human literature, Psychosomatic Medicine 63 (2001), 7–18.

[11] J.W. Creswell, Educational Research: Planning, Conducting, and Evaluating Quantitative and Qualitative Research, Upper Saddle River, NJ: Pearson Education, 2005.

[12] E.M. Daley, R.J. McDermott, K.R. McCormack-Brown and K.J. Kittleson, Conducting web-based survey research: A les- son in internet designs, American Journal of Health Behavior 27(2) (2003), 116–164.

[13] A.S. Devlin and A.B. Arneill, Health care environments and patient outcomes: A review of the literature, Environment and Behavior 35(5) (2003), 665–694.

[14] G.B. Diette, N. Lechtzin, E. Haponik, A. Devrotes and H.R. Rubin, Distraction therapy with nature sights and sounds re- duces pain during flexible bronchoscopy: A complementary approach to routine analgesia, Chest 123(3) (2003), 941–948.

[15] D.A. Dillman, Mail and Internet Surveys: The Tailored Design Method 2nd edition, New York: John Wiley & Sons, Inc, 2000.

[16] D. Dooley, Social Research Methods, 4th edition, New Jersey: Prentice Hall, 2001.

[17] L.H. Engbers, M.N.M. van Poppel, M.C. Paw and W. van Mechelen, Worksite health promotion programs with envi- ronmental changes, American Journal of Preventive Medicine 29(1) (2005), 61–70.

[18] F.J. Floyd and K.F. Widamen, Factor analysis in the develop- ment and refinement of clinical assessment instruments, Psy- chological Assessment 7(3) (1995), 286–299.

[19] L.D. Frank and P.O. Engelke, The build environment and hu- man activity patterns: exploring the impacts of urban form on public health, Journal of Planning Literature 16(2) (2001), 202–218.

[20] H. Frumkin, Climate change and public health: Think- ing, communicating, acting, American Journal of Preventive Medicine 35(5) (2008), 403–410.

[21] H. Frumkin, Environmental Health, San Francisco: Jossey- Bass, 2005.

[22] H. Frumkin, White coats, green plants: Clinical epidemiology meets horticulture, Acta Horticulturae 639 (2004), 15–26.

[23] H. Frumkin, Healthy places: Exploring the evidence, Ameri- can Journal of Public Health 93(9) (2003), 1451–1456.

[24] H. Frumkin, Beyond toxicity: Human health and the natu- ral environment, American Journal of Preventative Medicine 20(3) (2001), 234–240.

[25] P. Grahn and U.A. Stigsdotter, Landscape planning and stress, Urban Forestry and Urban Greenery 2 (2003), 1–18.

[26] K.T. Han, A reliable and valid self-rating measure of the restorative quality of natural environments, Landscape and Urban Planning 64 (2003), 209–232.

[27] R.K. Henson, Understanding internal reliability estimates: A conceptual primer on coefficient alpha, Measurement and Evaluation in Counseling and Development 34 (2001), 177– 189.

[28] A.P. Horan, An effective workplace stress management inter- vention: Chicken Soup for the Soul at Work employee groups, Work 18 (2002), 3–13.

[29] H.S. Huang, M.M. Robertson and K.I. Chang, The role of en- vironmental control on environmental satisfaction, communi- cation, and psychological stress: Effects of office ergonomics training, Environment and Behavior 36(5) (2004), 617–637.

[30] K.N. Irvine and S.L. Warber, Greening healthcare: Practic- ing as if the natural environment really mattered, Alternative Therapies in Health and Medicine 8(3) (2002), 76–83.

[31] R. Kaplan, The nature of the view from home: Psychological benefits, Environment and Behavior 33(4) (2001), 507–542.

[32] R. Kaplan, The role of nature in the context of the workplace, Landscape and Urban Planning 26 (1993), 193–201.

[33] S. Kaplan, The restorative benefits of nature: Towards an integrative framework, Journal of Environmental Psychology 15 (1995), 169–182.

[34] S. Kaplan and R. Kaplan, Health, supportive environments, and the reasonable person model, American Journal of Public Health 93(9) (2003), 1484–1489.

[35] K.J. Karren, B.Q. Hafen, N.L. Smith and K.J. Frandsen, Mind- body health, 2nd ed., San Francisco: Benjamin Cummings, 2002.

[36] P. Keep, J. James and M. Inman, Windows in the intensive therapy unit, Anesthesia 35 (1980), 257–262.

[37] J.K. Kiecolt-Glaser and R. Glaser, Psychosocial moderators of immune function, Annals of Behavioral Medicine 9(2) (1987), 16–20.

[38] K. Kreiss, The sick building syndrome: Where is the epi- demiologic basis? American Journal of Public Health 80(10) (1990), 1172–1173.

[39] R. Kuller and C. Lindstein, Health and behavior of children in classrooms with and without windows, Journal of Environ- mental Psychology 12 (1992), 305–317.

[40] F.E. Kuo, Coping with poverty: Impacts of environment and attention in the inner city, Environment and Behavior 33(1) (2001), 5–34.

[41] L. Larsen, J. Adams, B. Deal, B.S. Kweon and E. Tyler, Plants in the workplace: The effects of plant density on productivity, attitudes, and perceptions, Environment and Behavior 30(4) (1998), 261–281.

[42] P. Leather, D. Beale, A. Santos, J. Watts and L. Lee, Outcomes of environmental appraisal of different hospital waiting areas, Environment and Behavior 35(6) (2003), 842–869.

[43] J.R. Masters, The relationship between number of response categories and reliability of likert-type questionnaires, Journal of Educational Measurement 11(1) (1974), 49–53.

[44] H. Mausner-Dorsch and W.E. Eaton, Psychological work en- vironments and depression: Epidemiologic assessment of demand-control model, American Journal of Public Health 9(11) (2000), 1765–1770.

[45] M. Meichior, N. Krieger, I. Kawachi, L.F. Berkman, I. Nied- hammer and M. Goldberg, Work factors and occupational class disparities in sickness absence: Findings from the GAZEL Co- hort Study, American Journal of Public Health 95(7) (2005), 1206–1212.

[46] M.J. Mendell, W.J. Fisk, K. Kreiss, H. Levin, D. Alexander, W.S. Cain et al., Improving the health of workers in indoor environments: Priority research needs for a national occupa-

422 E. Largo-Wight et al. / The Nature Contact Questionnaire: A measure of healthy workplace exposure

tional research agenda, Public Health Matters 92(9) (2002), 1430–1440.

[47] S. Messick, Validity of psycholgoical assessment: Validation of inferences form persons’ response and performances as scietific inquiry into score meaning, American Psychologists 50(9) (1995), 741–749.

[48] E.O. Moore, A prison environment’s effects on health car ser- vice demands, Journal of Environmental Systems 2(11) (1981), 17–34.

[49] D. Ouis, Annoyance from road traffic noise: A review, Journal of Environmental Psychology 21 (2001), 101–120.

[50] R.A. Peterson, A meta-analysis of cronbach’s coefficient al- pha, Journal of Consumer Research 21 (1994), 381–391.

[51] P.M. Peterson and J.F. Wilson, Work stress in America, Inter- national Journal of Stress Management 11(2) (2004), 91–113.

[52] D.J. Pittenger, Behavioral Research Design and Analysis, US: McGraw Hill, 2003.

[53] S.R. Porter, Raising response rates: What works, New Direc- tions for Institutional Research 121 (2004), 5–21.

[54] S.R. Porter, Pros and cons of paper and electronic surveys, New Directions for Institutional Research 121 (2004), 91–97.

[55] M. Raffaello and A. Maass, Chronic exposure to noise in industry: The effects on satisfaction, stress symptoms, and company attachment, Environment and Behavior 34(5) (2002), 651–671.

[56] L.G. Portney and M.P. Watking, Foundations of clinical re- search: Applications to practice, 2nd ed., US: Prentice Hall, 2000.

[57] P. Salovey, A.J. Rothman, J.B. Detweiler and W.T. Steward, Emotional states and physical health, American Psychologist 55(1) (2000), 110–121.

[58] J.M. Samet and J.D. Spengler, Indoor environments and health: Moving into the 21st century, American Journal of Public Health 93(9) (2003), 1489–1497.

[59] N. Schwarz, Self-reports: How the questions shape the an- swers, American Psychologists 54(2) (1999), 93–105.

[60] M.K. Shear, T.A. Brown, D.H. Barlow, R. Money, D.E. Sholomskas, S.W. Woods, J.M. Gorman and L.A. Papp, Mul-

ticenter collaborative panic disorder severity scale, American Journal of Psychiatry 154 (1997), 1571–1575.

[61] L.A. Sheppard, Evaluating test validity, Review of research in education, volume 19, (Chp 9). American Educational Re- search Association, 1993.

[62] S. Srinivansan, L.R. O’Fallon and A. Dearry, Creating healthy communities, healthy homes, healthy people: Initiating a re- search agenda on the built environment and public health, American Journal of Public Health 93(9) (2003), 1446–1450.

[63] D.L. Streiner and G.R. Norman, Health Measurement Scales: A Practical Guide to Their Development and Use, 2nd ed., NY: Oxford Press, 1996.

[64] O.L. Strickland, When is internal consistency reliability as- sessment inappropriate? Journal of Nursing Measurement 7(1) (1999), 3–4.

[65] S. Thorne, Data anlaysis in qualitative research, Evidence Based Nursing 3 (2000), 68–70.

[66] L. Trenberth, P. Dewe and F. Walkey, Leisure and its role as a strategy for coping with work stress, International Journal of Stress Management 6(2) (1999), 89–103.

[67] R.S. Ulrich, Wellness by design: Psychologically supportive patient surroundings, Group Practice Journal 40(4) (1991), 10–19.

[68] R.S. Ulrich, R.F. Simons, B.D. Losito, E. Florito, M.A. Miles and M. Zelson, Stress recovery during exposure to natural and urban environments, Journal of Environmental Psychology 11 (1991), 201–230.

[69] R.S. Ulrich, R.F. Simons and M.A. Miles, Effects of envi- ronmental simulations and television on blood donor stress, Journal of Architectural and Planning Research 20(1) (2003), 38–47.

[70] E.O. Wilson, Biophilia: The Human Bond with Other Species, Harvard University Press, Cambridge, MA, 1984.

[71] R.A. Wood, Improving the Indoor Environment for Health, Well-Being, and Productivity, Paper presented at the meeting of the Greening Cities: A New Urban Ecology, Australian Technology Park, Sydney, Australia, April 2003.

E. Largo-Wight et al. / The Nature Contact Questionnaire: A measure of healthy workplace exposure 423

Appendix

Nature Contact Questionnaire

Directions Part 1: The following questions are about your contact with the outside environment near your office. (Choose N/A if you did NOT have a work break or lunch break last week).

1. Last week (Monday-Friday), how many times did you. . . 0 1 2 3 4 5 or more N/A a. spend a morning or afternoon work break outside? b. eat your lunch outside? c. exercise outside during your lunch break? d. go outside for work related task such as delivery or errands?

Directions Part 2–4: The following questions are about your indoor work environment (primary work space only). When completing the rest of this survey, please do not consider indoor spaces that you sometimes visit during your work day. Instead, only consider your primary work space, the one indoor space that you spend most of your day at work.

2. How would you describe your primary work space?

– Private office – Shared office – Reception area (faced to assist or greet visitors) – Private cubicle – Open design without offices or cubicle dividers between employees (non reception area) – Other (specify)

3. Last week (Monday-Friday), how many do you have in your primary work space? 0 1 2 3 4 5 or more e. live plants or flower arrangements f. artificial plants or flower arrangements g. windows (including those on doors) that lead directly to the outdoors h. fish aquariums with live fish i. animals or pets j. paintings or drawings that represent realistic natural scenes or animals (such as open fields, trees,

flowers, rain drop, ocean, cats, etc) k. personal or professional photographs of natural scenes or animals l. realistic sculptures or carvings of animals or natural elements (such as a bird or sunset)

4. In your primary work space last (Monday-Friday), what percentage of the time did you. . .

0% 1–20% 21–40% 41–60% 61–80% 81–100%

m. have sunlight lighting your space? n. have an unobstructed view outside (blinds open and not drawn)? o. listen to recorded nature sounds (such as ocean waves)? p. watch images of natural places or animals on a T.V. or monitor?

posted to a listserv without the copyright holder's express written permission. However, users may print,

download, or email articles for individual use.