How Religious Coping is Used Relative to Other Coping Strategies Depends on the Individual's Level of Religiosity and Spirituality

Page 1

ORI GIN AL PA PER

How Religious Coping is Used Relative to Other CopingStrategies Depends on the Individual’s Levelof Religiosity and Spirituality

Christian U. Krageloh • Penny Pei Minn Chai •

Daniel Shepherd • Rex Billington

Published online: 2 November 2010� Springer Science+Business Media, LLC 2010

Abstract Results from empirical studies on the role of religiosity and spirituality in

dealing with stress are frequently at odds, and the present study investigated whether level

of religiosity and spirituality is related to the way in which religious coping is used relative

to other coping strategies. A sample of 616 university undergraduate students completed

the Brief COPE (Carver in Int J Behav Med 4:92–100, 1997) questionnaire and was

classified into groups of participants with lower and higher levels of religiosity and spir-

ituality, as measured by the WHOQOL-SRPB (WHOQOL-SRPB Group in Soc Sci Med

62:1486–1497, 2006) instrument. For participants with lower levels, religious coping

tended to be associated with maladaptive or avoidant coping strategies, compared to

participants with higher levels, where religious coping was more closely related to prob-

lem-focused coping, which was also supported by multigroup confirmatory factor analysis.

The results of the present study thus illustrate that investigating the role of religious coping

requires more complex approaches than attempting to assign it to one higher order factor,

such as problem- or emotion-focused coping, and that the variability of findings reported

by previous studies on the function of religious coping may partly be due to variability in

religiosity and spirituality across samples.

Keywords Coping strategies � Religious coping � Religiosity � Spirituality � Brief COPE �WHOQOL-SRPB

Introduction

The psychological health benefits of religious involvement are very well documented

(Koenig et al. 2001; Seybold and Hill 2001), but the mechanisms by which these effects

emerge are still unclear. One way in which religion can exert a positive influence on psy-

chological well-being is through its function as a resource in dealing with stress. Compared

C. U. Krageloh (&) � P. P. M. Chai � D. Shepherd � R. BillingtonDepartment of Psychology, Faculty of Health & Environmental Sciences, Auckland Universityof Technology, North Shore Campus, Private Bag 92006, Auckland 1142, New Zealande-mail: [email protected]

123

J Relig Health (2012) 51:1137–1151DOI 10.1007/s10943-010-9416-x

Page 2

to the considerable amount of research that has been published during the last 40 years on

secular coping strategies, religious coping has only recently received more attention in the

psychological literature (Folkman and Moskowitz 2004). The function of religious coping

relative to these other coping strategies is still being debated, and its assignment to higher

order coping factors remains unclear. Tamres et al. (2002), for example, noted that religious

coping could be clearly classified neither as problem nor as emotion-focused coping—a

popular distinction between coping strategies that defines the former as activities directly

addressing the source of the stressor, while the latter are responses aimed to change one’s

own reactions to the stressors (Lazarus and Folkman 1984).

Coping strategies are commonly assessed using a variety of self-report instruments, with

specific ones available to assess religious coping. The RCOPE (Pargament et al. 2000), for

example, is a multidimensional inventory that asks about a very broad range of coping

responses that involve the use of religious concepts. The questionnaire classifies religious

coping as either positive or negative, where positive coping includes forgiveness and

seeking of spiritual support, and negative religious coping includes spiritual discontent and

punishing God reappraisals. This level of detail and sub-classification of religious coping is

obviously not provided in general coping inventories. The COPE questionnaire (Carver

et al. 1989), for example, includes religious coping as one of 15 sub-scales, but it is assessed

by using four items only and not divided further into different types of religious coping. In

the Brief COPE (Carver 1997), the abbreviated version of the COPE, religious coping

remains one of 14 sub-scales, but is assessed by only two items. The limitations of such

general questionnaires to inform about the role of religious coping have been mentioned

(Lavery and O’Hea 2010; Schottenbauer et al. 2006b), but questionnaire length remains a

major consideration in the design and use of survey-based research (Carver 1997).

Because of its brevity the Brief COPE (Carver 1997) can be used easily in conjunction

with other questionnaires. The author of the scale even recommended the use of selected

sub-scales when researchers have very focused interests or need to restrict participant

response burden to a minimum. This flexibility is no doubt the reason for the popularity of

the instrument, which has been used in a wide range of populations, including people

suffering from migraine (Radat et al. 2009), parents of children with end-stage renal

disease (Zelikovsky et al. 2007), emergency workers (Cicognani et al. 2009), genetic

counselors (Udipi et al. 2008), and international university students (Miyazaki et al. 2008).

For the initial development of the Brief COPE, Carver (1997) conducted a principal

component analysis on data collected from a community sample of hurricane survivors and

reported that the two items from the sub-scale turning to religion loaded onto a unique

factor. Based on the recommendation of Carver to use the Brief COPE flexibly, researchers

often conduct a factor analysis to determine how to analyze coping scores in their par-

ticular sample (e.g., Radat et al. 2009; Zelikovsky et al. 2007). While some studies con-

firmed Carver’s (1997) results of turning to religion forming a separate factor (Cicognani

et al. 2009; Miyazaki et al. 2008), results from other studies vary tremendously, with

reports of the religious items loading onto higher order factors, such as together with the

two items from the denial sub-scale (Hastings et al. 2005), the items from the substanceuse sub-scale (Paukert et al. 2009), the items from the behavioral disengagement, denial,and self-blame sub-scales (Liu and Iwamoto 2007), or religious coping items not loading

onto any factor (Welbourne et al. 2007). Other researchers have taken the approach of

conducting exploratory factor analyses by treating sub-scale total scores as indicators

instead of individual items, but results here are equally variable. Farley et al. (2005), for

instance, reported that turning to religion loaded together with positive reframing,

acceptance, and humor, while Schottenbauer et al. (2006a) found that religious coping

1138 J Relig Health (2012) 51:1137–1151

123

Page 3

loaded together with emotional support and instrumental support. Findings of turning toreligion forming its own factor are also reported (Weininger et al. 2006; Wood and

Rutterford 2006), as are findings of turning to religion failing to have sufficiently high

factor loading scores (Lawrence and Fauerbach 2003; Yang et al. 2008).

Results from exploratory factor analyses such as those mentioned above affect not only

the way coping data are analyzed in those studies, but also any general conclusions that the

researchers might derive about the role and function of religious coping. While it is

certainly valid to argue that low factor loadings of turning to religion with other coping

strategies could indicate that religious coping is a unique and independent strategy (Cook

and Heppner 1997; Hudek-Knezevic et al. 1999), the reasons for the large range of

reported factor structures still require further investigation. Kershaw et al. (2004) found

that religious coping loaded differently depending on whether an exploratory factor

analysis was conducted for patients with breast cancer or for their family caregivers. For

patients, the turning to religion sub-scale loaded together with active coping, planning,

positive reframing, acceptance, and emotional support, while for caregivers religious

coping loaded together with humor, self-distraction, venting, behavioral disengagement,denial, and substance use. This indicates that the characteristics of the sample can have a

substantial effect on the way religious coping is related to other coping strategies.

So far, no study has systematically explored how religious and spiritual beliefs affect

the use of religious coping relative to the secular coping strategies that are assessed by the

Brief COPE (Carver 1997). The varied reports of the role of religious coping in relation to

secular coping methods and the unclear assignment of religious coping to a higher order

factor structure might certainly be a reflection of differences in the role that religion and

spirituality play in different groups of people. In individuals with a religious affiliation,

religious orientation has been found to affect the manner in which religion is used as a

coping resource. An intrinsic religious orientation is more likely to be associated with

problem solving, and an extrinsic orientation more likely to be linked with cognitive

avoidance strategies (Aguilar-Vafaie and Abiari 2007; Pargament et al. 1992). However, in

countries like New Zealand, where more than 30% of the population does not self-identify

with a particular religious denomination (Perrott 2007), the intrinsic–extrinsic religious

orientation differentiation excludes a substantial proportion of people. Therefore, together

with the Brief COPE, participants in the present study completed the 32-item WHOQOL-

SRPB questionnaire (WHOQOL-SRPB Group 2006), which assesses the strength of a

person’s spiritual, religious, and personal beliefs. The advantage of this instrument is that it

is not limited to religiosity only, but also accounts for spiritual and personal beliefs that are

independent of any formal religious denomination. Moreover, the development of this

instrument occurred through an international collaborative project of 18 countries, and the

WHOQOL-SRPB is therefore considered cross-culturally valid and applicable to a wide

range of belief systems. If religious coping indeed plays a different role depending on the

degree of the spiritual and religious beliefs of an individual, it might be found to correlate

differently with the use of other, secular, coping strategies and may therefore also affect to

which higher order factor an exploratory factor analysis might assign religious coping.

Method

Participants

A convenience sample of university undergraduate students at a variety of different fac-

ulties at a university in New Zealand provided data on coping, demographic characteristics

J Relig Health (2012) 51:1137–1151 1139

123

Page 4

and spiritual, religious and personal beliefs. A total of 712 participants completed the

questionnaire, of which 679 were valid. The mean age of the participants was 22.83 years

(SD = 6.88), and 179 respondents indicated that they were male, compared to 494 who

indicated that they were female. Slightly less than 50% of the participants (N = 330) stated

that they had a religious affiliation.

Instruments

Brief Cope

Participants were given the dispositional form of the Brief COPE (Carver 1997) ques-

tionnaire, which asked them to rate to what extent they usually use a range of strategies to

cope with stressful events. Although four ordinal response categories may be an acceptable

minimum to conduct factor analyses (Lozano et al. 2008), increasing the number to five

categories has been found to lead to clear gains in accuracy of model fits (Johnson and

Creech 1983; Lozano et al. 2008), and for that reason the present study used a five-point

Likert scale instead of the four-point scale that is commonly used for the Brief COPE. No

data were imputed for the Brief COPE, but instead a very stringent inclusion criterion was

used according to which only data from participants who completed all question items

were included.

WHOQOL-SRPB

The 32-item questionnaire asks respondents to rate on a five-point Likert scale their extent

of spiritual, religious or personal beliefs (WHOQOL-SRPB Group 2006). Scores are

presented both as total scores and as one of the following eight facets: spiritual connection,

meaning of life, awe, wholeness, spiritual strength, inner peace, hope, and faith. The facet

awe, for example, includes one question that asks ‘‘To what extent are you able to

experience awe from your surroundings? (e.g., nature, art, music)’’, and one question from

the facet inner peace asks ‘‘To what extent do you feel peaceful within yourself?’’. When

at least three items of a facet were answered, the missing item was imputed by the facet

mean. If more than two items in a facet were missing, no facet score was calculated for that

participant. A participant’s total WHOQOL-SRPB score was only calculated when all facet

scores were available.

Procedure

The research project was introduced to university undergraduate students at the end of class.

Participants were alerted to standard ethical guidelines surrounding research using paper-

and-pencil questionnaire formats, such as anonymity and the right to withdraw or not

choosing to participate. A closed carton box with a slot was provided for participants to

return questionnaires. This research study was approved by the university ethics committee.

Data Analysis

Data analyses were conducted using the program Statistics Package for Social Sciencesversion 16.0 (SPSS 2008). The large sample size of the study afforded parametric Pear-

son’s r bivariate analyses of the association between religious coping and other secular

1140 J Relig Health (2012) 51:1137–1151

123

Page 5

coping strategies measured by the Brief COPE. These were conducted separately for

groups of participants classified according to whether they had obtained scores higher or

lower than the mean on the WHOQOL-SRPB. There were no a priori reasons why the

sample should have been divided in any particular way, and the sample was categorized

into groups of participants with WHOQOL-SRPB scores above and below the mean in

order to maintain a sufficiently large sample size for the factor analyses. To test whether

group differences in the associations of religious coping with other strategies might also

affect an extracted higher order factor solution, exploratory factor analyses were conducted

separately for the different groups and on the overall dataset. Analyses used sub-scale total

scores as indicators, a technique also known as extension analysis (Bernstein and Teng

1989), which is more robust than conducting factor analyses at item level. Factors were

extracted using the maximum likelihood method of factor extraction as this is one of the

most widely used methods, and factor rotations were conducted using promax. The number

of factors that was extracted for each analysis was based solely on the Kaiser-Guttman

criterion of number of eigenvalues larger than 1.00. Additional criteria, such as inspection

of the scree plot or interpretability of the factor solution were not utilized, since they

require subjective interpretations, and the goal of the present study was to investigate how

the characteristics of the sample determine the factor solutions of the different groups.

To provide a gauge of the adequacy of the extracted solutions, confirmatory factor

analyses were also conducted, using LISREL version 8.8 (Joreskog and Sorbom 1993).

Factor structures identified by the exploratory factor analyses were simplified, reducing

heavy cross-loadings across factors, and then tested for the two groups separately using

confirmatory factor analysis, as well as using a multigroup configural invariance test (Horn

and McArdle 1992; Meredith 1993). When data are ordinal in nature, researchers are

advised to use an asymptotically distribution-free (ADF) method with polychoric correla-

tions and asymptotic co-variance matrices (Joreskog 1990). The method of diagonally

weighted least squares is a suitable ADF method for small to moderate sample sizes (Flora

and Curran 2004) and was therefore selected for the present study. A set of indices was used

to determine the goodness-of-fit of the alternative factor solutions. Since chi-square

becomes inflated with increases in sample size (Marsh et al. 1988), model fits were eval-

uated using root mean square error of approximation (RMSEA), comparative fit index (CFI)

and standardized root mean square residual (SRMR). Using the frequently quoted guide-

lines by Hu and Bentler (1999), model fits were considered acceptable if RMSEA \ 0.06,

CFI [ 0.90 and SRMR \ 0.08.

Results

With the criterion that a full set of data be available for the Brief COPE (Carver 1997), the

sample size was reduced to 616. The mean total WHOQOL-SRPB score was higher for

female (106.07) than for male (105.65) participants, but this difference was not statistically

significant (t(603) = -0.18, P [ .05). There was a very small but statistically significant

correlation between age and WHOQOL-SRPB scores (r = 0.09, P \ .05). The majority of

kurtosis and skewness values for the Brief COPE sub-scales were within the negligible

range of -1.00 to 1.00 (Muthen and Kaplan 1985), and those outside this range were still

well below the levels at which deviations from normality become problematic (Curran

et al. 1996). For that reason, all sub-scales were included in the subsequent data analyses.

Table 1 shows the correlations between all different coping strategies. Apart from a few

correlations between highly related strategies, such as active coping and planning, most

J Relig Health (2012) 51:1137–1151 1141

123

Page 6

Tab

le1

Co

rrel

atio

ns

bet

wee

nth

e1

4co

pin

gsu

b-s

cale

sco

res

of

the

Bri

efC

OP

E(C

arv

er1

99

7)

Copin

gsu

b-

scal

eA

ctiv

eco

pin

gP

lannin

gP

osi

tive

refr

amin

gA

ccep

tance

Hum

or

Turn

ing

tore

ligio

nE

moti

onal

support

Inst

rum

enta

lsu

pport

Sel

f-dis

trac

tion

Den

ial

Ven

ting

Subst

ance

use

Beh

avio

ral

dis

engag

emen

tS

elf-

bla

me

Act

ive

copin

g1

Pla

nnin

g0.6

3**

1

Posi

tive

refr

amin

g0.4

3**

0.4

7**

1

Acc

epta

nce

0.4

0**

0.3

6**

0.4

8**

1

Hum

or

0.1

3**

0.0

9*

0.2

2**

0.2

7**

1

Turn

ing

tore

ligio

n0.1

4**

0.1

8**

0.2

2**

0.1

2**

-0.0

11

Em

oti

onal

support

0.2

7**

0.2

7**

0.2

7**

0.2

1**

0.0

9*

0.2

8**

1

Inst

rum

enta

lsu

pport

0.2

5**

0.3

1**

0.2

6**

0.1

6**

0.1

3**

0.1

7**

0.6

3**

1

Sel

f-dis

trac

tion

0.0

40.1

0*

0.1

4**

0.1

1**

0.0

50.0

40.1

6**

0.2

5**

1

Den

ial

-0.0

60.0

2-

0.0

3-

0.0

9*

-0.0

20.1

6**

0.0

80.1

4**

0.2

8**

1

Ven

ting

0.0

70.1

7**

0.1

1**

0.0

00.0

60.1

6**

0.2

2**

0.2

7**

0.2

7**

0.4

6**

1

Subst

ance

use

-0.1

4**

-0.0

6-

0.1

0*

-0.0

50.0

6-

0.1

7**

-0.1

5**

-0.0

70.1

4**

0.1

7**

0.1

9**

1

Beh

avio

ral

dis

engag

emen

t-

0.1

2**

-0.0

6-

0.1

4**

-0.0

60.0

00.0

3-

0.0

50.0

50.2

3**

0.4

5**

0.3

2**

0.3

7**

1

Sel

f-bla

me

-0.0

40.0

7-

0.0

9*

-0.0

50.0

40.0

10.0

20.1

0*

0.2

6**

0.3

3**

0.4

4**

0.2

6**

0.4

9**

1

*P

\.0

5;

**

P\

.01

1142 J Relig Health (2012) 51:1137–1151

123

Page 7

correlations were weak to moderate. The first eight strategies, sometimes labeled adaptivestrategies (Meyer 2001), were generally correlated with each other, as were the mal-adaptive strategies. Turning to religion was weakly correlated with active coping, plan-ning, positive reframing, acceptance, emotional support, instrumental support, denial and

venting, and negatively correlated with substance use.

Table 2 shows correlations between turning to religion and the remaining 13 Brief

COPE sub-scales, calculated separately for participants with a WHOQOL-SRPB score

below the mean and above the mean. To provide a comparison of groups of participants

that differ to a larger extent in terms of religiosity and spirituality, participants were also

grouped into those with a score that was at least one standard deviation above the mean and

one standard deviation below the mean. Since WHOQOL-SRPB scores correlated sig-

nificantly with age, partial correlations were calculated, controlling for age as a con-

founding variable. For participants with a WHOQOL-SRPB score above the mean, turningto religion was correlated with planning only, and therefore, the correlation was very weak.

For participants with a WHOQOL-SRPB score that was more than one standard deviation

above the mean, religious coping was not correlated with planning, but instead with activecoping, positive reframing and acceptance. For participants with lower WHOQOL-SRPB

scores, a very different picture emerged. Turning to religion was not correlated with any of

the adaptive coping strategies, except for emotional support and instrumental support for

participants with a score below the mean. Instead, turning to religion was correlated with

Table 2 Partial correlations (controlling for age) between scores on the turning to religion sub-scale withthose of the other 13 coping sub-scales of the Brief COPE (Carver 1997)

More than one standarddeviation belowthe mean (N = 101)

Below the mean(N = 303)

Above the mean(N = 307)

More than one standarddeviation abovethe mean (N = 111)

Active coping -0.08 -0.03 0.11 0.20*

Planning 0.01 0.07 0.16** 0.14

Positivereframing

0.08 0.02 0.10 0.28**

Acceptance -0.02 -0.02 0.07 0.20*

Humor -0.07 -0.04 0.01 -0.07

Emotionalsupport

0.15 0.15* 0.10 0.14

Instrumentalsupport

0.14 0.13* 0.02 0.01

Self-distraction 0.06 0.10 -0.01 -0.07

Denial 0.45** 0.37** 0.07 0.06

Venting 0.38** 0.29** 0.09 0.12

Substance use 0.10 0.05 -0.07 -0.13

Behavioraldisengagement

0.35** 0.35** -0.04 0.01

Self-blame 0.13 0.17** 0.03 -0.00

Results are shown separately for participants who obtained a WHOQOL-SRPB (WHOQOL-SRPB Group2006) at least one standard deviation below the mean, below the mean, above the mean, and one standarddeviation above the mean

* P \ .05; ** P \ .01

J Relig Health (2012) 51:1137–1151 1143

123

Page 8

the maladaptive strategies denial, venting and behavioral disengagement, as well as

self-blame for participants with a WHOQOL-SRPB score below the mean.

An exploratory factor analysis of the Brief COPE sub-scales conducted on the whole

dataset suggested a four-factor solution, which accounted for 43.71% of the variance. The

factor loadings are shown in Table 3. Except for turning to religion, where all factor

loadings are shown, only loadings above 0.30 are shown for the other sub-scales, with the

primary factor loadings shown in bold. The six so-called maladaptive (Meyer 2001) coping

strategies all loaded onto the same factor without cross-loadings, except for venting, which

had a secondary loading with the factor that included turning to religion, emotional supportand instrumental support. Two other factors were active coping and planning, and positivereframing, acceptance, and humor, with a substantial amount of cross-loadings.

The comparison of the results from exploratory factor analyses conducted separately for

participants with a total WHOQOL-SRPB score lower than the mean and those higher than

the mean are shown in Table 4. In both cases, a four-factor solution was extracted, which

accounted for 46.73% of the variance for the former group, and only 38.69% for the latter

group. As with the analyses of the overall dataset (Table 3), the maladaptive strategies

loaded together with one factor (Table 4). The factor loadings of the other sub-scales for

both groups were also very similar to those of the overall dataset: Active coping and

planning formed one factor, as did positive reframing, acceptance, and humor, and lastly,

emotional support and instrumental support. The only differences were related to religious

coping. Turning to religion loaded with the maladaptive strategies for participants with a

WHOQOL-SRPB score lower than the mean, and failed to load for participants with a

score higher than the mean (Table 4). The factor loading scores of the latter group were not

even near the 0.30 criterion for any of the four extracted factors.

Confirmatory factor analyses investigated the adequacy of the factor solutions extracted

by the preceding exploratory factor analyses. Given the large amount of cross-loadings of

the first five coping strategies (Table 4), the factor structure was reduced to three factors,

Table 3 Factor loading patterns from an exploratory factor analysis conducted using the overall dataset

Factor 1 Factor 2 Factor 3 Factor 4

Active coping 0.72 0.48 0.34

Planning 0.88 0.43 0.38

Positive reframing 0.57 0.61 0.36

Acceptance 0.48 0.78

Humor 0.35

Turning to religion (0.22) (0.15) 0.32 (0.07)

Emotional support 0.34 0.85

Instrumental support 0.36 0.74

Self-distraction 0.39

Denial 0.61

Venting 0.33 0.60

Substance use 0.42

Behavioral disengagement 0.70

Self-blame 0.66

Only factor loadings above 0.30 are shown, and primary loadings are highlighted in bold. For turning toreligion, all factor loadings are shown, but those below 0.30 are shown in parentheses

1144 J Relig Health (2012) 51:1137–1151

123

Page 9

comprising (a) active coping, planning, positive reframing, acceptance and humor, (b)

emotional support and instrumental support, and (c) self-distraction, denial, venting,

substance use, behavioral disengagement and self-blame. The first factor was labeled

active, the second emotional and the third maladaptive. The goodness-of-fit parameters

were compared when fitting turning to religion to each of these three factors and conducted

separately for different groups of participants, divided according to whether the partici-

pants had obtained a WHOQOL-SRPB below or above the mean. Additionally, tests of

configural invariance (Horn and McArdle 1992; Meredith 1993) across these groups were

conducted, imposing no invariance constraints on the factor loadings, factor correlations

and error variances. Adequacy of goodness-of-fit parameters of these multigroup confir-

matory analyses would indicate that the factor structure of the two groups is equivalent.

As shown in Table 5, the results of fitting turning to religion to active, emotional or

maladaptive coping differed for the two groups of participants that were divided according

to whether their obtained total WHOQOL-SRPB scores were either below or above the

mean. The configural invariance tests showed that the model fits were best when religious

coping was excluded. CFI was at 0.90, although RMSEA and SRMR were above the

respective cut-off values of 0.06 and 0.08 to indicate acceptable fits. Fitting the multigroup

model with turning to religion assigned to any of the three factors resulted in slightly worse

fit indices, indicating that the group differences were minimized when religious coping was

excluded. However, group differences in assignment of religious coping to the factor

structure were evident when comparing the results of the individual group fits. For par-

ticipants with higher WHOQOL-SRPB scores, fit indices for religious coping being part of

the active factor indicated equally good if not better fits compared to when religious coping

was excluded. Assigning turning to religion to emotional coping resulted in a slightly

Table 4 Factor loading patterns from an exploratory factor analysis conducted separately for participantswith a total WHOQOL-SRPB (WHOQOL-SRPB Group 2006) score lower than the mean and those higherthan the mean

WHOQOL-SRPB score below themean (N = 303)

WHOQOL-SRPB score above themean (N = 307)

Factor 1 Factor 2 Factor 3 Factor 4 Factor 1 Factor 2 Factor 3 Factor 4

Active coping 0.75 0.47 0.31 0.77 0.37

Planning 0.89 0.40 0.35 0.73 0.43 0.31

Positive reframing 0.59 0.61 0.32 0.47 0.75

Acceptance 0.48 0.77 0.45 0.52

Humor 0.45

Turning to religion (0.01) (-0.05) (0.18) 0.42 (0.18) (0.09) (0.10) (0.04)

Emotional support 0.32 0.78 0.81

Instrumental support 0.32 0.81 0.32 0.31 0.74

Self-distraction 0.33 0.38 0.43

Denial 0.64 0.61

Venting 0.40 0.63 0.61

Substance use 0.45 0.30

Behavioral disengagement 0.80 0.60

Self-blame 0.65 0.65

Only factor loadings above 0.30 are shown, and primary loadings are highlighted in bold. For turning toreligion, all factor loadings are shown, but those below 0.30 are shown in parentheses

J Relig Health (2012) 51:1137–1151 1145

123

Page 10

worse fit, and the worst was when it was assigned to maladaptive coping. For participants

with lower levels of religiosity and spirituality, in contrast, the opposite pattern was

noticeable, where the best fit was for excluding religion, followed by assigning religious

coping to maladaptive coping, emotional coping and active coping. As an additional check,

the pattern of standardized residuals was regarded, as a wrongly assigned indicator would

result in negative residuals with the other indicators of that factor (Anderson and Gerbing

1988). Although residuals were mostly not significantly different from zero, the pattern

was consistent with the interpretation that for participants with lower levels of religiosity

and spirituality, religious coping was least aligned with active and most with maladaptivecoping, whereas the opposite was the case for participants with higher levels of religiosity

and spirituality.

It may be of interest to the reader to note that the clear differences in patterns of

associations between religious coping and secular coping strategies by participants cate-

gorized as having higher versus lower WHOQOL-SRPB scores were not present when

participants were categorized according to whether they had a religious affiliation or not.

Instead, when partial correlations were calculated separately for the two groups, a very

similar pattern emerged (Table 6). For both groups, religious coping was positively cor-

related with active coping, planning, positive reframing, emotional support, instrumentalsupport and venting. For participants with a religious affiliation, religious coping was also

positively correlated with acceptance and negatively correlated with substance use, while

for participants without a religious affiliation, there was also a significant positive corre-

lation between religious coping and denial.

Table 5 Degrees of freedom (df), goodness-of-fit indices (RMSEA, CFI and SRMR) for confirmatoryfactor analyses assigning turning to religion to either active, emotional, or maladaptive coping, and whenturning to religion was not included

Turning to religion with df v2 RMSEA CFI SRMR

Active coping

WHOQOL-SRPB above mean 74 255.757 0.068 0.914 0.073

WHOQOL-SRPB below mean 74 379.028 0.093 0.897 0.106

Configural invariance 148 634.785 0.081 0.876 0.106

Emotional coping

WHOQOL-SRPB above mean 74 260.980 0.069 0.911 0.075

WHOQOL-SRPB below mean 74 376.645 0.093 0.898 0.097

Configural invariance 148 637.625 0.081 0.876 0.097

Maladaptive coping

WHOQOL-SRPB above mean 74 265.640 0.070 0.909 0.077

WHOQOL-SRPB below mean 74 342.309 0.086 0.911 0.085

Configural invariance 148 607.949 0.078 0.885 0.085

Turning to religion not included

WHOQOL-SRPB above mean 62 235.671 0.071 0.918 0.075

WHOQOL-SRPB below mean 62 285.482 0.086 0.920 0.085

Configural invariance 124 521.153 0.079 0.900 0.085

Analyses were conducted separately for participants with WHOQOL-SRPB (WHOQOL-SRPB Group 2006)scores below the mean and those above, as well as configural invariance tests on both groups. Values of fitindices are shown with three decimal places to highlight differences between the different fits

1146 J Relig Health (2012) 51:1137–1151

123

Page 11

Discussion

The present study investigated whether level of religiosity and spirituality is related to the

way in which religious coping is used relative to other coping strategies. Participants’

scores on the sub-scale turning to religion from the Brief COPE (Carver 1997) ques-

tionnaire were correlated with their scores on secular coping strategies. Very different

patterns emerged when data were analyzed separately for participants with lower scores on

the WHOQOL-SRPB compared with those who had obtained higher scores. While for the

latter group religious coping was only significantly correlated with planning, for the other

group, turning to religion was correlated significantly with a range of other strategies, such

as the two emotion-focused strategies emotional support and instrumental support, as well

as most of the maladaptive coping strategies (Table 2). As a consequence, the exploratory

factor analyses also produced different results for the two groups. While for the overall

dataset turning to religion loaded together with emotional support and instrumental sup-port (Table 3), it failed to load onto any factor for participants with higher levels of

religiosity and spirituality, and for those with lower level, religious coping loaded together

with maladaptive coping strategies, such as self-distraction, denial and behavioral disen-gagement (Table 4). The results from the confirmatory factor analyses support the con-

clusion that the factor structure differs for these two groups of participants. For participants

with higher levels of spirituality, religious coping was most related to active coping

strategies and least related to maladaptive ones, while the exact opposite was the case for

participants with lower levels of spirituality.

The findings of the present study may help explain the very diverse reports of the role of

religious coping in relation to other coping strategies. These reports range from findings

that turning to religion loads with the active strategies positive reframing, acceptance and

humor (Farley et al. 2005), with both active and emotional coping strategies (Bean et al.

2009; Ebert et al. 2002), only emotional strategies (Schottenbauer et al. 2006a), or mal-

adaptive strategies, such as with behavioral disengagement, denial and self-blame (Liu and

Iwamoto 2007) or denial only (Hastings et al. 2005). Religious coping failing to load

(Lawrence and Fauerbach 2003; Kershaw et al. 2008; Yang et al. 2008), or forming its own

Table 6 Partial correlations(controlling for age) betweenscores on the turning to religionsub-scale with those of the other13 coping sub-scales of the BriefCOPE (Carver 1997)

Results are shown separately forparticipants who indicated thatthey had a religious affiliationand those that did not

* P \ .05; ** P \ .01

No affiliation(N = 321)

Religious affiliation(N = 295)

Active coping 0.15** 0.18**

Planning 0.18** 0.28**

Positive reframing 0.23** 0.22**

Acceptance 0.09 0.18**

Humor 0.08 0.01

Emotional support 0.21** 0.26**

Instrumental support 0.16** 0.18**

Self-distraction 0.08 -0.00

Denial 0.19** 0.08

Venting 0.16** 0.15*

Substance use -0.03 -0.14*

Behavioral disengagement 0.08 0.01

Self-blame -0.05 0.04

J Relig Health (2012) 51:1137–1151 1147

123

Page 12

factor (Saroglou and Anciaux 2004; Weininger et al. 2006; Wood and Rutterford 2006), is

also frequently reported. The fact that religious coping loaded differently in the present

study for groups classified by their WHOQOL-SRPB scores suggests that variability in the

reported factor structures across different studies may be partly due to variability of the

extent of religiosity and spirituality of the participants sampled.

A large number of studies using the Brief COPE use the individual sub-scale scores

(e.g., Li and Lambert 2007; Pritchard and McIntosh 2003; Taylor et al. 2004), but when

coping is analyzed as higher order factors, such as emotion-focused coping, the way in

which composite scores are calculated is frequently based on a prior exploratory factor

analysis of the specific sample. While this practice reduces the overall likelihood that

religious coping is assigned to an inadequate factor, it still makes the assumption that the

extracted factor structure applies to the entire sample in the same manner. The present

study showed that, at least in respect to religious coping, this assumption cannot always be

justified. Kershaw et al. (2004) also reported that the factor structure was different,

depending on whether the factor analysis was conducted for caregivers or care recipients

with breast cancer, and one might need to question whether other studies would equally

have found such variations within their samples. Studies making comparisons of sub-

groups within their sample are therefore advised to ascertain that the intended comparisons

are valid by demonstrating that religious coping aligns with other factors in the same

manner for all the groups that are to be compared.

The results of the present study also illustrate that investigating the role of religious

coping requires more complex answers than simply assigning it to one factor, such as

active or emotion-focused coping. As shown in the present study, religious coping tends to

be used in a more problem-focused manner by individuals with higher levels of religiosity

and spirituality. This implies the use of one’s religion and spirituality to assist gathering or

focusing resources on solving the problem that causes the individual stress. For individuals

with lower levels of spirituality, in contrast, religion and spirituality might instead be

associated with avoidant or escapist strategies such as denying the problem or seeking

distraction that helps avoid directly addressing the source of the problem. As with indi-

viduals with an extrinsic religious orientation (Aguilar-Vafaie and Abiari 2007; Pargament

et al. 1992), these may be desperate religious coping strategies called upon predominantly

in times of stress, unlike those utilized by individuals with an intrinsic orientation, where

religious coping is not driven by immediate personal needs, but tends to be more deeply

rooted in religious belief systems.

For the purposes of the present study, the term maladaptive coping was used, although

the usefulness of the distinction between adaptive and maladaptive strategies is certainly

contentious (Lazarus 2000; Zuckerman and Gagne 2003). Effectiveness of coping

responses depends on the situational context, and supposedly adaptive coping strategies are

not necessarily always effective, and maladaptive ones ineffective (David et al. 2006).

Future research is necessary to explore further how exactly religious coping interacts with

the use of other strategies, particularly different types of religious coping, such as positive

and negative religious coping (Pargament et al. 2000). Further work is also required to

investigate specifically how spiritual and existential contemplations are used as coping

resources in the nonreligious. As Hwang et al. (in press) highlighted, this group is very

diverse and is therefore likely to have a wide range of personal, spiritual and existential

beliefs that may be utilized in dealing with stress. This type of research is especially

important in countries like New Zealand, where a substantial proportion of people are not

directly affiliated with a particular religion (Perrott 2007). The results shown in Table 6

demonstrate that religious affiliation is not a good indicator of the way in which religious

1148 J Relig Health (2012) 51:1137–1151

123

Page 13

coping is utilized in relation to other strategies, unlike level of spiritual, religious and

personal beliefs (WHOQOL-SRPB), which are very clearly linked to different patterns of

associations (Table 2).

As also previously reported by Hsu et al. (2009), more than 50% of university students

do not consider themselves to be part of a religious denomination, unlike the 30% reported

in the 2006 New Zealand national census (Perrott 2007). Generalizations of the findings of

the present study are therefore limited by the fact that data were collected from a con-

venience sample of university undergraduate students. Furthermore, the dispositional form

of the Brief COPE (Carver 1997) was administered, which asked the participants how they

usually respond to a stressful situation, as opposed to asking them how they dealt with

stress in a specific situation. The frequency and types of strategies that individuals report to

use generally can differ significantly from the frequency and types individuals report to use

in specific stressful situations. Gillen (2006), for instance, found such discrepancies in

individuals who had a stroke, although turning to religion was one of the strategies that

was relatively stable, and scores of the situational and dispositional formats correlated

positively. However, future research may investigate whether the differences in the way

that religious coping relates to other coping strategies for individuals with lower and higher

levels of religiosity and spirituality are also present when inquiring about specific stressful

events, and not only when assessing coping using a dispositional format.

References

Aguilar-Vafaie, M. E., & Abiari, M. (2007). Coping response inventory: Assessing coping among Iraniancollege students and introductory development of an adapted Iranian coping response inventory (CRI).Mental Health, Religion & Culture, 10, 489–513.

Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice: A review and rec-ommended two-step approach. Psychological Bulletin, 103, 411–423.

Bean, M. K., Gibson, D., Flattery, M., Duncan, A., & Hess, M. (2009). Psychosocial factors, quality of life,and psychological distress: Ethnic differences in patients with heart failure. Progress in Cardiovas-cular Nursing, 24, 131–140.

Bernstein, I. H., & Teng, G. (1989). Factoring items and factoring scales are different: Spurious evidence formultidimensionality due to item categorization. Psychological Bulletin, 105, 467–477.

Carver, C. S. (1997). You want to measure coping but your protocol’s too long: Consider the Brief COPE.International Journal of Behavioral Medicine, 4, 92–100.

Carver, C. S., Scheier, M. F., & Weintraub, J. K. (1989). Assessing coping strategies: A theoretically basedapproach. Journal of Personality and Social Psychology, 56, 267–283.

Cicognani, E., Pietrantoni, L., Palestini, L., & Prati, G. (2009). Emergency workers’ quality of life: Theprotective role of sense of community, efficacy beliefs and coping strategies. Social IndicatorsResearch, 94, 449–463.

Cook, S. W., & Heppner, P. P. (1997). A psychometric study of three coping measures. Educational andPsychological Measurement, 57, 906–923.

Curran, P. J., West, S. G., & Finch, J. F. (1996). The robustness of the test statistics to nonnormality andspecification error in confirmatory factor analysis. Psychological Methods, 1, 16–29.

David, D., Montgomery, G. H., & Bovbjerg, D. H. (2006). Relations between coping responses and opti-mism-pessimism in predicting anticipatory psychological distress in surgical breast cancer patients.Personality and Individual Differences, 40, 203–213.

Ebert, S. A., Tucker, D. C., & Roth, D. L. (2002). Psychological resistance factors as predictors of generalhealth status and physical symptom reporting. Psychology, Health, & Medicine, 7, 363–375.

Farley, T., Galves, A., Dickinson, L. M., & de Jesus Diaz Perez, M. (2005). Stress, coping, and health: Acomparison of Mexican immigrants, Mexican-Americans, and Non-hispanic whites. Journal ofImmigrant Health, 7, 213–220.

Flora, D. B., & Curran, P. J. (2004). An empirical evaluation of alternative methods of estimation forconfirmatory factor analysis with ordinal data. Psychological Methods, 9, 466–491.

J Relig Health (2012) 51:1137–1151 1149

123

Page 14

Folkman, S., & Moskowitz, J. T. (2004). Coping: Pitfalls and promise. Annual Review of Psychology, 55,745–774.

Gillen, G. (2006). A comparison of situational and dispositional coping after a stroke. Occupational Therapyin Mental Health, 22, 31–59.

Hastings, R. P., Kovshoff, H., Brown, T., Ward, N. J., Espinosa, F. D., & Remington, B. (2005). Copingstrategies in mothers and fathers of preschool and school-age children with autism. Autism, 9, 377–391.

Horn, J. L., & McArdle, J. J. (1992). A practical and theoretical guide to measurement invariance in agingresearch. Experimental Aging Research, 18, 117–144.

Hsu, P. H.-C., Krageloh, C. U., Shepherd, D., & Billington, R. (2009). Religion/spirituality and quality oflife of international tertiary students in New Zealand: An exploratory study. Mental Health, Religion &Culture, 12, 385–399.

Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventionalcriteria versus new alternatives. Structural Equation Modeling, 6, 1–55.

Hudek-Knezevic, J., Kardum, I., & Vukmirovic, Z. (1999). The structure of coping styles. A comparativestudy of Croatian sample. European Journal of Personality, 13, 149–161.

Hwang, K., Hammer, J. H., & Cragun, R. T. (in press). Extending religion-health research to secularminorities: Issues and concerns. Journal of Religion and Health.

Johnson, D. R., & Creech, J. C. (1983). Ordinal measures in multiple indicator models: A simulation studyof categorization error. American Sociological Review, 48, 398–407.

Joreskog, K. G. (1990). New developments in LISREL: Analysis of ordinal variables using polychoriccorrelations and weighted least squares. Quality & Quantity, 24, 387–404.

Joreskog, K. G., & Sorbom, D. (1993). LISREL 8: Structural equation modelling with the SIMPLIS com-mand language. Hillsdale, NJ: Erlbaum Associates.

Kershaw, T. S., Mood, D. W., Newth, G., Ronis, D. L., Sanda, M. G., Vaishampayan, U., & Northouse, L. L.(2008). Longitudinal analysis of a model to predict quality of life in prostate cancer patients and theirspouses. Annals of Behavioral Medicine, 36, 117–128.

Kershaw, T., Northouse, L., Kritpracha, C., Schafenacker, A., & Mood, D. (2004). Coping strategies andquality of life in women with advanced breast cancer and their family caregivers. Psychology andHealth, 19, 139–155.

Koenig, H. G., McCullough, M. E., & Larson, D. B. (2001). Handbook of religion and health. Oxford:Oxford University Press.

Lavery, M. E., & O’Hea, E. L. (2010). Religious/spiritual coping and adjustment in individuals with cancer:Unanswered questions, important trends, and future directions. Mental Health, Religion & Culture, 13,55–65.

Lawrence, J. W., & Fauerbach, J. A. (2003). Personality, coping, chronic stress, social support and PTSDsymptoms among adult burn survivors—A path analysis. Journal of Burn Care and Rehabilitation, 24,63–72.

Lazarus, R. S. (2000). Toward better research on stress and coping. American Psychologist, 55, 665–673.Lazarus, R. S., & Folkman, S. (1984). Stress appraisal and coping. New York: Springer.Li, J., & Lambert, V. A. (2007). Coping strategies and predictors of general well-being in women with breast

cancer in the People’s Republic of China. Nursing and Health Sciences, 9, 199–204.Liu, W. M., & Iwamoto, D. K. (2007). Conformity to masculine norms, Asian values, coping strategies, peer group

influences and substance use among Asian American men. Psychology of Men & Masculinity, 8, 25–39.Lozano, L. M., Garcıa-Cueto, E., & Muniz, J. (2008). Effect of the number of response categories on the

reliability and validity of rating scales. Methodology, 4, 73–79.Marsh, H. W., Balla, J. R., & McDonald, R. P. (1988). Goodness-of-fit indexes in confirmatory factor

analysis: The effect of sample size. Psychological Bulletin, 103, 391–410.Meredith, W. (1993). Measurement invariance, factor analysis and factorial invariance. Psychometrika, 58,

525–543.Meyer, B. (2001). Coping with severe mental illness: Relations of the Brief COPE with symptoms, func-

tioning, and well-being. Journal of Psychopathology and Behavioral Assessment, 23, 265–277.Miyazaki, Y., Bodenhorn, N., Zalaquett, C., & Ng, K.-M. (2008). Factorial structure of Brief COPE for

international students attending US colleges. College Student Journal, 42, 795–806.Muthen, B., & Kaplan, D. (1985). A comparison of some methodologies for the factor analysis of non-

normal Likert variables. British Journal of Mathematical and Statistical Psychology, 38, 171–189.Pargament, K. I., Koenig, H. G., & Perez, L. M. (2000). The many methods of religious coping: Devel-

opment and initial validation of the RCOPE. Journal of Clinical Psychology, 56, 519–543.Pargament, K. I., Olsen, H., Reilly, B., Falgout, K., Ensing, D. S., & van Haitsma, K. (1992). God help me

(II): The relationship of religious orientations to religious coping with negative life events. Journal forthe Scientific Study of Religion, 31, 504–513.

1150 J Relig Health (2012) 51:1137–1151

123

Page 15

Paukert, A. L., LeMaire, A., & Cully, J. A. (2009). Predictors of depressive symptoms in older veterans withheart failure. Age & Mental Health, 13, 601–610.

Perrott, K. (2007). Atheism and religious diversity. Aotearoa Ethnic Network Journal, 2(2), Article 8.Pritchard, M. E., & McIntosh, D. N. (2003). What predicts adjustment among law students? A longitudinal

panel study. The Journal of Social Psychology, 143, 727–745.Radat, F., Lanteri-Minet, M., Nachit-Ouinekh, F., Massiou, H., Lucas, C., Pradalier, A., et al. (2009). The

GRIM2005 study of migraine consultation in France. III: Psychological features of subjects withmigraine. Cephalalgia, 29, 338–350.

Saroglou, V., & Anciaux, L. (2004). Liking sick humor: Coping styles and religion as predictors. Humor, 17,257–277.

Schottenbauer, M. A., Klimes-Dougan, B., Rodriguez, B. F., Arnkoff, D. B., Glass, C. R., & Lasalle, V. H.(2006a). Attachment and affective resolution following a stressful event: General and religious copingas possible mediators. Mental Health, Religion & Culture, 9, 448–471.

Schottenbauer, M. A., Rodriguez, B. F., Glass, C. R., & Arnkoff, D. B. (2006b). Religious coping researchand contemporary personality theory: An exploration of Endler’s (1997) integrative personality theory.British Journal of Psychology, 97, 499–519.

Seybold, K. S., & Hill, P. C. (2001). The role of religion and spirituality in mental and physical health.Current Directions in Psychological Science, 10, 21–24.

SPSS, Inc. (2008). SPSS for windows, version 16.0. Chicago, USA: SPSS, Inc.Tamres, L. K., Janicki, D., & Helgeson, V. S. (2002). Sex differences in coping behavior: A meta-analytic

review and an examination of relative coping. Personality and Social Psychology Review, 6, 2–30.Taylor, S. E., Sherman, D. K., Kim, H. S., Jarcho, J., Takagi, K., & Dunagan, M. S. (2004). Culture and

social support: Who seeks it and why? Journal of Personality and Social Psychology, 87, 354–362.Udipi, S., Veach, P. M., Kao, J., & LeRoy, B. S. (2008). The psychic costs of emphatic engagement:

Personal and demographic predictors of genetic counselor compassion fatigue. Journal of GeneticCounseling, 17, 459–471.

Weininger, C. F., Shalev, A. Y., Ofek, H., Freedman, S., Weissman, C., & Einav, S. (2006). Posttraumaticstress disorder among hospital surgical physicians exposed to victims of terror: A prospective, con-trolled questionnaire survey. Journal of Clinical Psychiatry, 67, 890–896.

Welbourne, J. L., Eggerth, D., Hartley, T. A., Andrew, M. E., & Sanchez, F. (2007). Coping strategies in theworkplace: Relationships with attributional style and job satisfaction. Journal of Vocational Behavior,70, 312–325.

WHOQOL-SRPB Group. (2006). A cross-cultural study of spirituality, religion, and personal beliefs ascomponents of quality of life. Social Science and Medicine, 62, 1486–1497.

Wood, R. L. L., & Rutterford, N. A. (2006). Demographic and cognitive predictors of long-term psycho-social outcome following traumatic brain injury. Journal of the International NeuropsychologicalSociety, 12, 350–358.

Yang, H.-C., Brothers, B. M., & Andersen, B. L. (2008). Stress and quality of life in breast cancerrecurrence: Moderation or mediation of coping? Annals of Behavioral Medicine, 35, 188–197.

Zelikovsky, N., Schast, A. P., & Jean-Francois, D. (2007). Parent stress and coping: Waiting for a child toreceive a kidney transplant. Journal of Clinical Psychology in Medical Settings, 14, 320–329.

Zuckerman, M., & Gagne, M. (2003). The COPE revised: Proposing a 5-factor model of coping strategies.Journal of Research in Personality, 37, 169–204.

J Relig Health (2012) 51:1137–1151 1151

123All in-text references underlined in blue are linked to publications on ResearchGate, letting you access and read them immediately.