Psychotherapy Research, October 2005; 15(4): 457
/469

Development and validation of the theistic spiritual outcome survey

P. SCOTT RICHARDS1, TIMOTHY B. SMITH1, MARION SCHOWALTER2,

MATTHIAS RICHARD2, MICHAEL E. BERRETT3, & RANDY K. HARDMAN3
1

Brigham Young University, 2University of Wurzburg, and 3Center for Change, Orem, Utah

(Received 29 September 2003; revised 2 July 2004; accepted 1 August 2004)

Abstract
The authors developed the Theistic Spiritual Outcome Survey (TSOS) to measure the spiritual outcomes of psychotherapy
from a theistic spiritual perspective. A 17-item version of the TSOS was found to have adequate reliability and validity in a
sample of college students. Three factors emerged from the analyses that corresponded to the Love of God, Love of Others,
and Love of Self subscales. Correlations with measures of psychological outcomes were statistically significant. In
subsequent analyses, the TSOS was administered over an 8-week period to a sample of inpatient women with eating
disorders and to two samples from inpatient psychological clinics in Germany. The analyses resulting from these studies
were also supportive of the psychometric properties and clinical sensitivity of the TSOS.

There is growing professional recognition of the

influences of religious faith and spirituality on human health and healing (Koenig, McCullough, &
Larson, 2001). In parallel with the increased research publications on this topic, numerous spiritually oriented psychotherapy approaches have been
developed (Miller, 1999; Richards & Bergin, 1997,
2000; Sperry & Shafranke, in press). As part of this
line of inquiry, researchers have called for specific
attention to the assessment of the religious and
spiritual outcomes of mental health treatment
(P. C. Hill & Pargament, 2003; Richards & Bergin,
1997; Worthington, Kurusu, McCullough, & Sanders, 1996).
There are several important reasons why researchers should assess how therapy affects clients religiousness and spirituality. Do secular psychotherapies undermine and weaken clients spiritual
beliefs and faith as some religious people have
claimed? Do spiritually oriented psychotherapies
strengthen clients faith and spirituality as others
have suggested? In light of the hesitancy of many
religiously committed people to seek psychotherapy
from secular therapists out of fear that it will
undermine their religious faith and values (Bergin,
1980; Richards & Bergin, 2000), it seems important
for psychotherapy researchers to address these
questions. Documenting that psychotherapy does
not necessarily undermine clients faith and spirituality could increase the trust leaders and members

of traditional religious communities have for mental

health professionals. It could also lead to greater
numbers of referrals from religious leaders and
increased mental health service utilization rates
from members of their congregations.
Furthermore, from an empirical perspective, assessing client spiritual outcomes is in line with
suggestions from leading psychotherapy researchers
who have advocated a multifaceted approach to
outcome assessment (e.g., Kazdin, 1994; Lambert
& Hill, 1994). With several extant multidimensional
models of mental health giving notable attention to
spirituality (e.g., Jung, 1933; Maslow, 1964; May,
1982; Myers, Sweeney, & Witmer, 2000), the need
to assess spiritual outcomes seems clear. Nevertheless, despite the widespread use of multidimensional assessments of clients psychological, social,
and physical functioning, the religious and spiritual
outcomes of psychotherapy have to date been largely
neglected (P. C. Hill & Pargament, 2003; Richards &
Bergin, 1997; Worthington et al., 1996).
A related reason why it is important to assess the
spiritual outcomes of mental health treatment is
to determine whether improvements in spiritual
well-being are associated with better treatment outcomes in other domains of clients functioning.
Some health professionals have suggested that improvements in faith and spirituality during treatment
may be associated with better and more enduring
emotional, social, familial, and work outcomes

Correspondence: P. Scott Richards, 340 MCKB, Department of Counseling Psychology, Brigham Young University, Provo, UT 84602.
E-mail: scott_richards@byu.edu
ISSN 1050-3307 print/ISSN 1468-4381 online # 2005 Society for Psychotherapy Research
DOI: 10.1080/10503300500091405

458

P. Scott Richards et al.

(Richards & Bergin, 1997). The possibility that

spiritual growth may act as a catalyst to help
promote and maintain other positive changes in
clients lives is intriguing. Spiritual outcome measures are needed to investigate this possibility.
The need for clinically valid spiritual
outcome measures
Perhaps the neglect of the spiritual dimension in
psychotherapy outcome research is not surprising
given the historical exclusion of religion and spirituality in the professions of psychology and psychotherapy (Bergin, 1980). Another reason for this
neglect is undoubtedly that there have been few
empirically proven and clinically useful religious and
spiritual outcome measures (Hall, Tisdale, Brokaw,
1994; P. C. Hill & Pargament, 2003; Richards &
Bergin, 1997). Fortunately, this situation may be
beginning to change.
C. H. Hill and Hoods (1999) handbook, Measures
of Religiosity, contains copies and critical reviews of
more than 120 measures of different types of
religiousness that have been developed in the psychology and sociology of religion fields, including
religious beliefs, practices, and attitudes; religious
orientation; religious development; religious commitment and involvement; religious experience;
religious and moral values; religious coping and
problem solving; spirituality and mysticism; God
concept; religious fundamentalism; views about
death and the afterlife; religious attributions; and
forgiveness. Many of these measures could be used
as therapy outcome measures, although few were
specifically designed or validated for this purpose.
Nevertheless, this resource should help make
religious and spiritual measures more easily accessible to psychotherapy outcome researchers and
clinicians.
A few religious and spiritual measures have been
used in psychotherapy outcome research. For example, the Spiritual Well-Being Scale (SWBS; Ellison, 1983; Ellison & Smith, 1991) has been used as a
therapy outcome measure in a number of studies.
The SWBS is appealing because it is brief and
correlates positively with a variety of physical and
mental health indicators. Another promising instrument is the Religious Commitment Inventory (RCI10; Worthington et al., 2003), which measures the
degree to which a person adheres to his or her
religious values, beliefs, and practices and uses them
in daily living (Worthington et al., 2003, p. 85).
Factor analyses have revealed that the RCI-10 is
composed of two factors: (a) intrapersonal religious
commitment and (b) interpersonal religious commitment. Evidence reported to date supports the

validity and clinical usefulness of the RCI-10 with

both Christian and non-Christian samples
(Worthington et al., 2003).
Several other religious and spiritual measures that
may prove useful as psychotherapy outcome measures are the Index of Core Spiritual Experiences
(Kass, Friedman, Lesserman, Zuttermeister, &
Benson, 1991), Spiritual Assessment Inventory
(Hall & Edwards, 1996), Spiritual Transcendence
Scale (Piedmont, 1999), Spiritual Health Inventory
(Veach & Chappel, 1992), and Spirituality Scale
(Jagers & Smith, 1996). More research in clinical
settings is needed with these measures.
In addition to validation studies of existing measures, new religious and spiritual measures created
specifically for psychotherapy outcome research are
needed. Such measures can be deliberately designed
to possess the qualities required of good outcome
instruments, such as brevity and sensitivity to
therapeutic change (Lambert & Hill, 1994).
In the remainder of this article, we describe the
development and validation of English and German
versions of the Theistic Spiritual Outcome Survey
(TSOS), a brief spiritual outcome questionnaire that
can be administered weekly in therapy to assess
clients perceptions of their spirituality. The TSOS is
theoretically grounded in a theistic view of spirituality described by Richards and Bergin (1997) that
includes several dimensions or components: (a) faith
in Gods existence, (b) feelings of reverence, love,
and closeness to God, (c) awareness of ones
spiritual identity and purpose as a creation of God,
(d) love for other people, including a desire to
promote their welfare, and (e) feelings of moral
congruence, worthiness, and self-acceptance. These
dimensions of spirituality can be found in the
teachings of the major theistic world religions,
including Judaism, Christianity, and Islam (Richards
& Bergin, 1997). The TSOS was not intended to
assess Eastern, transpersonal, humanistic, or other
nontheistic forms of spirituality.
Study 1
Item development
Using the theistic conception of spirituality described previously as a theoretical foundation,
P. Scott Richards wrote approximately 50 items
that he thought reflected various indicators of
theistic spirituality, including items that reflected
faith, love, and reverence toward God; life purpose
and meaning; love for other people; desires to serve,
help, and forgive others; and moral congruence,
worthiness, and self-acceptance. A relatively large
number of items were written with the intent that

Development and validation of the TSOS

sizable numbers of items might be dropped as a
result of psychometric considerations, lack of clarity,
or lack of ecumenical suitability. After writing the
item pool, P. Scott Richards then gave the items to
an internationally known researcher in the area of
psychotherapy outcomes and the psychology of
religion and to a professor of religion who is an
expert in the world religions. These two scholars
reviewed the items and made suggestions for item
pool refinement.
Method
Sample. A 34-item version of the TSOS was
administered to 344 research participants who were
recruited from multiple classes at a midsized university in the north-central region of the United
States (n /158) and at a large university in the
western region (n /185). Participants were offered
either complimentary movie passes or extra credit
toward their class grade as an incentive for their
participation. At the top of the TSOS, the following
instructions were provided: Please help us understand how you have been feeling spiritually this past
week, including today. Carefully read each item
below and fill in or mark the circle that best
describes how you felt. The items were based on a
5-point Likert-type scale; potential responses were
almost always, frequently, sometimes, rarely, and never.
The total sample was composed of 222 females
(64.5%) and 120 males (34.9%); two (0.6%)
participants did not report gender. The mean age
of the participants was 22.6 years (range / 17
/61
years). Most participants were European American
(n /292 [85%]), seven were African American
(2%), four were Latin American (1%), nine were
Asian American (3%), and 11 were Native American
(3%); 13 reported other ethnicities (4%) and eight
(2%) did not respond. Concerning religious affiliation, 41 of the participants were Roman Catholic
(14%), 15 were Methodist (4%), 33 were Lutheran
(10%), 187 were members of the Church of Jesus
Christ of Latter-Day Saints (LDS; 55%), 42 reported affiliation with other Christian denominations (14%), and 21 reported non-Christian
affiliations or no religious affiliation (7%). Regarding
perceptions of God, 304 (88%) participants indicated they believed in a personal God, 15 (4%)
stated they believed in an impersonal God, 16 (5%)
indicated they were agnostic, and nine (3%) did not
respond to this item.
Measures. In addition to the 34-item TSOS, the
participants completed the Outcome Questionnaire45 (OQ-45), a general measure of psychological
distress that has been frequently used in psychother-

459

apy outcome research. The OQ-45 provides a total

score and three subscale scores and has been shown
to be a reliable and valid instrument across many
studies (Lambert & Burlingame, 1996).
Results
Preliminary analyses and item reduction. To refine
the item composition of the TSOS, the 34 items
were subjected to exploratory factor analysis. Based
on Gorsuchs (1997) recommendations, principleaxis factoring was used, followed by promax rotation. Six factors had eigenvalues greater than one,
but analysis of the scree plot suggested that a fourfactor solution was optimal. Subsequently, four
factors were extracted and item communalities and
factor loadings were analyzed. Items that did not
have factor loadings above .40 on any of the four
factors were removed.
Conceptual analysis of item content was also used
to reduce the number of items. For example, all
negatively items were removed. When two items
shared similar content, the item that was more
clearly worded was retained. In all, these conceptual
and statistical analyses reduced the item pool by half,
resulting in a 17-item version of the TSOS.
Factor analyses. Data from the resulting 17-item
version were again subjected to factor analysis.
Three factors with eigenvalues greater than 1.0
were extracted and accounted for 61% of the
variance in the items. The first factor had an
eigenvalue of 7.55 (accounting for 44% of the
variance), the second an eigenvalue of 1.60 (accounting for 9.4% of the variance), and the third an
eigenvalue of 1.26 (accounting for 7.4% of the
variance). Item communalities and rotated factor
loadings are presented in Table I. All items loaded
significantly on one of the three rotated factors; six
items loaded on the first factor, six on the second
factor, and five on the third factor.
Item content was then examined to identify the
common meaning in the factors extracted. Items
that assessed participants feelings of love and
connectedness with God characterized the first
factor. Items that contained the words God and
spiritual loaded onto this factor, and the highest
factor loading was for Item 14: I felt Gods love.
The one item about prayer (communicating with
God) also loaded onto this factor, as did items about
having faith in and praising God. Thus, this first
factor was labeled the Love of God subscale.
Reliability analyses indicated that this scale had an
internal consistency (Cronbachs alpha) of .93.
Items loading on the second factor shared content
pertinent to ideal humanitarianism. Every item

460

P. Scott Richards et al.

Table I. Study 1: Item commonalities and rotated pattern matrix for Theistic Spiritual Outcome Survey items in the college student sample.
Rotated pattern matrix
Item
1. I had feelings of love toward others. (LO)
2. I felt there is a spiritual purpose for my life. (LG)
3. I felt good about my moral behavior. (LS)
4. I wanted to make the world a better place. (LO)
5. I felt peaceful. (LS)
6. I felt appreciation for the beauty of nature. (LO)
7. I felt like praying. (LG)
8. I felt spiritually alive. (LG)
9. I felt worthy. (LS)
10. My behavior was congruent with my values. (LS)
11. I felt love for all of humanity. (LO)
12. I had faith in Gods will. (LG)
13. I felt like helping others. (LO)
14. I felt Gods love. (LG)
15. I praised and worshipped God. (LG)
16. I felt forgiveness toward others. (LO)
17. I loved myself. (LS)

Communality

.33
.64
.55
.45
.47
.38
.68
.68
.55
.41
.50
.67
.51
.78
.73
.35
.37

.09
.73
.01

/.01
.01
/.12
.87
.62
.01
.08
.09
.84
.01
.93
.85
.07
/.07

.42
.06
.05
.75
.21
.62
.03
.02
/.13
.10
.66
/.10
.71
/.04
.08
.46
.01

.11
.05
.71

/.12
.53
.10

/.11
.26
.82
.51

/.19
.07
/.11
/.04
/.08
.11
.65

Note. Factor 1 was labeled love of God (LG), Factor 2 love of others (LO), and Factor 3 love of self (LS).

referred to either feelings or actions toward others.

The item with the highest loading (Item 11) reads:
I wanted to make the world a better place, and the
remaining items emphasize issues of forgiveness,
appreciation, and love toward others. Thus, this
factor was labeled the Love of Others subscale.
Reliability analyses indicated that this scale had an
internal consistency (Cronbachs alpha) of .80.
Items denoting self-acceptance and feelings of
moral worthiness characterized the third factor. An
item reflective of harmony between actions and
values also loaded on this scale, and the highest
loading was for Item 9: I felt worthy. The other
items denoted love for self. Thus, this factor was
labeled the Love of Self subscale, with the assumption that a personal sense of worth and worthiness
was also captured by this title. Reliability analyses
indicated that this scale had an internal consistency
(Cronbachs alpha) of .80.
Because men and women have been found to
differ in their spirituality orientations (Levin, Taylor,
& Chatters, 1994), subsequent analyses were conducted to verify that the obtained factor structure
would be replicated when data from the two genders
were examined separately. The results were very
consistent with the initial solution. For men, the
original factor structure was replicated exactly, and
all 17 items loaded significantly on the same factors
they had previously. For women, the analysis yielded
a fourth factor with an eigenvalue of exactly 1.0.
Examination of the pattern matrix revealed that this
fourth factor was accounted for by a single item
(Item 10): My behavior was congruent with my
values. Furthermore, it was noted that Item 1 (I

had feelings of love toward others) did not load

significantly on any of the four factors. All remaining
items loaded significantly on the same factors
identified in the initial analysis with the total sample.
When a three-factor solution was run with the data
from the female participants, Item 10 loaded significantly on Factor 3, as it had in the solution
generated from the data from the total sample.
Because more than half of the participants in this
study were LDS members, and past research has
indicated some differences in the spiritual orientations of people affiliated with LDS compared with
other Christian denominations (Jensen, Jensen, &
Wiederhold, 1993), it was important to verify that
the factor structure of the TSOS did not differ
between LDS and non-LDS participants. Results of
an analysis conducted with data from LDS participants were very similar to those conducted with the
total sample, except that Items 1 and 17 loaded
on the factor labeled love of God (the factor loadings
were .37 and .35, respectively) in addition to loading
significantly on the factors in which they had loaded
with the total sample (love of others and love of self,
respectively). For non-LDS participants, the results
also closely matched the original analysis, except
that Item 1 also loaded on love of self (loading /.38)
in addition to loading significantly on love of
others, as intended. Given the small magnitude of
the discrepant multiple loadings and problems
inherent with reducing the sample size, these
results served to confirm that the overall factor
structure reported previously was robust across
subgroups.

Development and validation of the TSOS

Correlational analyses. To assess the relationship
between the three subscales identified in the prior
analyses, interscale correlations were computed. As
may be seen in Table II, the subscales consistently
correlated with one another at about .59. The
magnitude of these correlations indicates that,
although the subscales clearly assess a similar core
construct (spirituality), they apparently assess somewhat different aspects of that construct.
To assess the relationship between spiritual outcome and psychiatric symptoms, the TSOS subscales were correlated with the total score and
subscales of the OQ-45. As may be seen in Table
II, the TSOS subscales were all significantly negatively correlated with the OQ-45. The Love of Self
subscale consistently correlated higher than the
other two TSOS subscales, reflecting a particular
overlap between Love of Self and mental health, as
expected. To assess whether the items assessing
feelings of moral self-acceptance and feelings of
inner peace were accounting for the high relationship
between the Love of Self subscale and the OQ-45,
individual items from that subscale were correlated
with the OQ-45. Results indicated that the magnitude of the correlation with the OQ-45 was very
similar for all items (range / /.34 to /.40). No
single item or set of items was identified as being
primarily responsible for the high degree of overlap
between the subscale and scores on the OQ-45.
Thus, a conceptual contribution of the TSOS is that
perceptions of moral behavior are potentially relevant to mental health. Overall, the correlational
analyses were supportive of the convergent and
discriminant validity of the TSOS.
Study 2
Although Study 1 served to validate the TSOS in a
nonclinical sample, the TSOS is intended for use as
a clinical outcome instrument. Therefore, additional
Table II. Study 1: Intercorrelations of Theistic Spiritual Outcome
Survey (TSOS) total score and subscales and correlations with
OQ-45 in the college student sample.

Scale
TSOS
Love of God
Love of Others
Love of Self
Outcome Questionnaire
OQ-45 total
Symptom Distress
Relationship Conflict
Social Role Conflict

TSOS
total

Love of
God

Love of
Others

.91
.81
.81

.58
.58

.59

/.33
/.38
/.55
/.38

/.25
/.28
/.45
/.31

/.20
/.23
/.39
/.26

Note. All correlations are significant at p B/.0001.

Love of
Self

/.48
/.54
/.61
/.45

461

work was needed to test its reliability and validity

with a clinical sample. To address this, the 17-item
TSOS was used as a weekly outcome measure in a
study of women who were receiving inpatient treatment for eating disorders. The complete findings of
this study have been reported in more detail elsewhere (Richards, Hardman, & Berrett, 2001). Here
we report only those findings that are relevant to the
development of the TSOS.
Method
Treatment facility and program. The treatment unit
is a private facility located in the western United
States that provides both inpatient and outpatient
treatment for women with eating disorders. The
multidisciplinary inpatient treatment staff includes
physicians, psychologists, social workers, dietitians,
and nurses. Patients participate in a variety of
individual, group, family, and experiential therapies
to ensure comprehensive treatment and progress
toward recovery.
Participants. Included in this study were 122
women suffering from anorexia nervosa (n /42;
34.4%), bulimia nervosa (n /47; 38.5%), or eating
disorder not otherwise specified (n /33; 27%). A
high percentage of participants (n /94; 77%) suffered from a psychiatric comorbid diagnosis. The
average length of stay in the inpatient treatment
program for patients was 68 days.
Participants came from 19 different states; the
majority were from Utah (n /62; 50.8%), California
(n /14; 11.5%), Idaho (n /7; 5.7%), and Colorado
(n /6; 4.9%). The ages of the participants ranged
from 13 to 52 years (M /21.2 years, SD /6.6
years). Most were European Americans (n /119;
97.5%). The majority of participants were LDS
(n /84; 68.9%); eight (6.5%) were Protestant, seven
(5.7%) were Catholic, two (1.6%) were Jewish, and
18 (14.8%) did not specify a religious affiliation but
did endorse personal beliefs in spirituality.
Procedures and treatment conditions. When participants were admitted to the center, they were
informed about the purpose and possible benefits
and risks of the study. After consenting to participate, they were randomly assigned to one of three
treatment groups, each led by a doctoral-level
psychologist. In addition to participating in the
regular intensive inpatient treatment program, patients in all three groups attended their weekly
treatment condition group sessions. Patients in
Treatment Group 1 attended a weekly 60-min
spiritual growth group and read a self-help workbook, which included scriptural and other spiritual

462

P. Scott Richards et al.

readings and educational materials about topics such

as God, spiritual identity, grace, forgiveness, repentance, faith, prayer, and meditation. Patients
in Treatment Group 2 attended a weekly 60-min
cognitive therapy group and read a self-help
workbook that describes a variety of cognitive
and behavioral techniques. Patients in Treatment
Group 3 attended a weekly 60-min emotional
support group and were invited to bring up any
topics for discussion that were of concern to them
each week.
Outcome measures. Several outcome measures were
administered at admission and discharge (posttreatment), including the Eating Attitudes Test (EAT;
Garner & Garfinkel, 1979) to assess symptoms
associated with anorexia nervosa and bulimia nervosa (e.g., restricting, binging, purging); Body Shape
Questionnaire (BSQ; Cooper, Taylor, Cooper, &
Fairburn, 1987) to assess concerns about body shape
and feelings of self-consciousness and shame about
ones body; OQ-45 (Lambert & Burlingame, 1996);
Multidimensional Self-Esteem Inventory (MSEI;
Epstein & OBrien, 1983) to assess patients global
self-esteem, moral self-approval, and lovability; and
SWBS (Ellison, 1983) to assess patients religious
and existential well-being. The TSOS was administered during the first 8 weeks of treatment.

Results
Correlations with measures of religiousness and
spiritual well-being. The Pearson correlations of the
TSOS with the Religious Orientation Scale (ROS)

and SWBS subscales are given in Table III. Overall,

the patterns of correlations between measures provided evidence that the TSOS is measuring similar
but not identical constructs as the ROS and Religious Well-Being (RWB) subscale, both of which
assess a mature, committed religious and spiritual
orientation (C. H. Hill & Hood, 1999). The total
TSOS and TSOS Love of God subscale were
significantly and positively correlated (.37 and .53,
respectively) with the intrinsic subscale of the ROS.
This indicates that those who are intrinsically
(devoutly) committed to their religious beliefs tend
to feel more love and closeness to God. The finding
that the total TSOS and TSOS subscales were not
significantly correlated with the extrinsic subscale of
the ROS provides some evidence concerning the
divergent validity of the TSOS, given that extrinsic
religiousness is viewed as an instrumental, less
devout form of religious involvement (C. H. Hill &
Hood, 1999).
Considerable evidence supports the construct
validity of the SWBS as a measure of spiritual wellbeing (e.g., Ellison, 1983); thus, the relatively high
correlations of the total TSOS with the SWBS
subscales provide convergent validity evidence for
the TSOS. The high correlation of the TSOS Love
of God subscale with the RWB subscale is particularly noteworthy in light of the fact that the RWB
subscale also measures peoples feelings of connection with God. Similarly, as expected, the Love of
Self (moral self-acceptance) subscale of the TSOS
correlated most highly with the Existential WellBeing subscale, which assesses whether people feel a
sense of life purpose, direction, and satisfaction.

Table III. Study 2: Correlations of Theistic Spiritual Outcome Survey total score and subscales with religious and psychological measures in
eating-disordered patient sample.
Scale

TSOS total

Love of God

ROS
Intrinsic
Extrinsic

.37***
.15

.53***
.18

SWBS
Religious Well-Being
Existential Well-Being

.57***
.49***

.71***
.43***

Outcome Questionnaire
OQ-45 total
Symptom Distress
Relationship Conflict
Social Role Conflict

/.35***
/.31***
/.37***
/.23*

EAT
BSQ
Global Self-Esteem (MSEI)

/.01
/.27**
.33***

Love of Others

.05
/.04

Love of Self

.17
.13

.19*
.33***

.29**
.45***

/.18
/.15
/.27**
/.10

/.31***
/.28**
/.31***
/.23*

/.44***
/.43***
/.35***
/.29**

.04
/.21*
.26**

.08
/.07
.23*

/.18
/.38***
.37***

Note. TSOS/Theistic Spiritual Outcome Survey; ROS/Religious Orientation Scale; SWBS/Spiritual Well-Being Scale; EAT /Eating
Attitudes Test; BSQ/Body Shape Questionnaire; MSEI/Multidimensional Self-Esteem Inventory.
*p B/.05. **p B/.01. ***p B/.001.

Development and validation of the TSOS

Correlations with psychological functioning. The
Pearson correlations of the TSOS with various
measures of psychological functioning are presented
in Table III. Generally, the TSOS manifested low
but significant correlations with the measures of
psychological functioning administered to this clinical sample. Those who scored higher on the TSOS
tended to be less psychologically disturbed. As
expected, the Love of Self (moral self-acceptance)
subscale of the TSOS correlated somewhat more
highly with the OQ-45, the BSQ, and the MSEI
subscales than did the other TSOS subscales.
Furthermore, the finding that the TSOS Love of
Self subscale correlated quite highly with the MSEI
Moral Self-Approval scale (r /.49, p B/.001) provides some evidence that the Love of Self subscale
measures in part peoples feelings of moral worthiness or congruence.
Rates of improvement over time on the Theistic
Spiritual Outcome Survey. A hierarchical linear modeling (HLM; Bryk & Raudenbush, 1992) analysis for
all eight weeks of TSOS total scores revealed that the
main effect for week was statistically significant,
F (1, 88)/45.4, p B/.0001, which indicates that
collectively the patients spirituality improved at a
significant rate during the first eight weeks of
treatment. The Week / Group interaction effect
on the TSOS for Weeks 1 through 8 was not
statistically significant, F (2, 87) /1.68, p /.19,
which indicates that the relative rates of improvement between the treatment groups across all eight
weeks of treatment did not significantly differ.
However, the Week /Group interaction effect on

Table IV. Weekly Theistic Spiritual Outcome Survey (TSOS)

Means, standard deviations, and standardized mean differences by
treatment conditions in eating-disordered patient sample (Study
2).

da

TSOS total
Spiritual
.58
M
45.3 46.4 48.9 49.9 49.3 51.4 52.9 51.9
SD
11.9 11.3 10.3 9.2 10.3 10.3 8.3 10.8
Cognitive
.23
M
41.0 43.7 42.7 40.7 40.3 41.6 41.6 43.2
SD
9.4 11.1 8.8 10.4 8.5 10.2 13.3 9.8
Support
M
SD

the TSOS for Weeks 1 through 4 was statistically

significant, F (2, 88) /3.12, p B/.05, which indicates
that the relative rates of improvement between the
treatment groups during the first four weeks of
treatment differed. The HLM slopes indicated that
TSOS scores of the patients in the spirituality group
(slope /2.23) increased more rapidly during the first
four weeks of treatment compared with those of
patients in the cognitive (slope /0.42) and support
(slope /1.09) groups.
The standardized mean score differences on
the TSOS between the three treatment conditions
(Table IV) indicate that patients in the spirituality
group on average experienced larger increases in
their TSOS scores during the first eight weeks of
treatment than did patients in the other two treatment groups. These results were consistent across all
three TSOS subscales, so data for the subscales are
not presented. The TSOS total scores for each
treatment group also indicate that there was not a
problem with ceiling effect because the mean scores
remained below or slightly above the theoretical
midpoint of the TSOS scale even after eight weeks
of treatment.
Study 3
To further validate the TSOS for its intended use as
a clinical outcome instrument, a third study was
conducted, this time in Germany. This study was
conducted in two inpatient treatment centers and
provided the opportunity to validate a German
language version of the TSOS as well as to crossvalidate some of the psychometric properties of the
English TSOS. The complete findings of this study
have been reported elsewhere (Schowalter, Richard,
Senst, & Murken, in press). Here we summarize only
those findings that are relevant to the validation of
the TSOS.
Method

Week
Measure/
group

463

.32
39.1 40.9 40.9 41.8 45.6 42.5 45.2 43.1
11.8 11.1 12.0 14.8 14.8 16.9 14.5 13.0

Note. The range of possible scores on the TSOS total scale is 17 to

85, with a midpoint of 51.
a
Standardized mean difference from Week 1 to Week 8.

TSOS translation. A native German doctorallevel psychologist translated the 17-item English
version of the TSOS into German. A native English
speaker then verified the accuracy of the translation
by retranslating the German version of the
TSOS back into English. Once the accuracy of the
German translation of the TSOS was verified, the
TSOS- German translation (TSOS-GT) was then
used in a comparative inpatient treatment outcome
study.
Treatment clinics and participants. Two inpatient
psychiatric clinics in Germany participated in the
treatment outcome study. Psychotherapists in both

464

P. Scott Richards et al.

clinics had been trained in either psychodynamic or

behavioral therapy and used these approaches in
their treatment programs.
Two hundred eighty patients in Clinic 1 participated in the study. Clinic 1 was a spiritually oriented
treatment center. All of the therapists were Christians, and spiritual interventions were integrated into
the treatment program (e.g., daily worship sessions,
prayer, scriptural discussions, teaching about religious concepts), along with psychodynamic and
behavioral interventions. Regarding religious affiliation, 12.1% of patients in Clinic 1 were Protestant,
30.4% were Roman Catholic, 41.3% were evangelical Christians, and 16.2% were of some other
nonspecified tradition. More than 84% of these
patients indicated on a rating scale that their
personal religious beliefs were very important or
extremely important to them, whereas only 2.9%
said their beliefs were less important or not
important.
The average age of Clinic 1 patients was 39 years
(SD /10.9 years). Two hundred eight (74.3%) of
Clinic 1 patients were women; 72 (25.7%) were
men. In regard to psychiatric diagnosis, 10.4% of the
patients in Clinic 1 had schizophrenia (nonacute),
44.9% had depression, 29.3% had a neurotic disorder (anxiety disorder, obsessive
/compulsive disorder, somatoform disorder), 6.8% had an eating
disorder, and 8.6% had a personality disorder. The
average length of treatment for patients in Clinic 1
was 52.2 days (SD /17.1 days).
One hundred eighty-five patients in Clinic 2 also
participated in the study. Clinic 2 was a secular
treatment program. The therapists were diverse in
terms of their religious affiliation and orientation,
and spiritual interventions were not used in the
treatment program. In regard to religious affiliation,
45.4% of patients in Clinic 2 were Protestant, 35.1%
were Roman Catholic, 1.6% were evangelical Christians, 3.8% were nonspecified Christian, 1.6% were
non-Christian, and 12.5% said they belonged to no
religious denomination. Only 21.1% of these patients indicated that their personal religious beliefs
were very important or extremely important to
them, whereas 50.8% said their religious beliefs were
less important or not important.
The average age of Clinic 2 patients was 43.0 years
(SD /10.4 years); 132 (71.4%) were women and 53
(28.6%) were men. Regarding psychiatric diagnosis,
one (0.5%) Clinic 2 patient had schizophrenia
(nonacute), 74 (40%) had depression, 75 (40.5%)
had a neurotic disorder (anxiety disorder, obsessive
/
compulsive disorder, somatoform disorder), 22
(11.9%) had an eating disorder, and five (2.7%)
had a personality disorder; the diagnosis for eight

(4.4%) patients was not available. The average

length of treatment for patients in Clinic 2 was
44.9 days (SD /13.3).
Outcome measures. Several outcome measures were
administered at admission and discharge (posttreatment), including the Symptom Checklist (SCL-90R; Derogatis, 1983), Freiburg Personality Inventory
(FPI), and TSOS-GT. The SCL-90-R is a widely
used 90-item self-report symptom inventory. It has
nine primary symptom dimensions (e.g., somatization, depression, anxiety, hostility) and three global
indexes of distress, of which the most important is
the Global Severity Index (GSI). The German
translation of SCL-90-R has adequate reliability,
and evidence supporting its validity has been reported (Franke, 1995). The FPI is a German
personality questionnaire that assesses general wellbeing and satisfaction with life. The FPI has
adequate reliability and evidence supporting its
validity (Schowalter et al., in press). Two items
written by the researchers that inquired about the
importance of the patients religious beliefs were also
rated at admission and discharge. Item 1 asked
patients to rate on a 6-point scale How important
are your personal religious beliefs? Item 2 asked the
patients to rate on a 6-point scale Do your personal
religious beliefs serve as a source of consolation and
strength?
Results
Factor analyses. Because of linguistic and cultural
differences between Germany and the United States,
it was important to factor analyze the German
translation of the TSOS to see how consistent its
factor structure was with the English TSOS. Three
factors with eigenvalues greater than 1.0 were
extracted that accounted for 61.7% of the variance
in the items. The first factor accounted for 40% of
the variance, the second factor for 13.5%, and the
third factor for 8.2%. The TSOS-GT items that
loaded on each factor exactly corresponded with the
items that loaded on each factor for the English
version of the TSOS. The correlations between the
TSOS subscales were comparable to those reported
in Study 1 for the English version of the TSOS.
The correlation between the Love of God and
Love of Others subscales was .45 (p B/.01); between
the Love of God and Love of Self subscales, .48
(p B/.01); and between the Love of Others and Love
of Self subscale, .55 (p B/.01).
Internal consistency reliabilities. The Cronbach alpha reliability estimates for the total TSOS-GT was

Development and validation of the TSOS

465

Table V. Correlations of Theistic Spiritual Outcome Survey total score and subscales with religious and psychological measures in
german psychiatric clinics (Study 3).
Scale
Importance of Personal Belief
Beliefs a Source of Consolation
SCL-90-R Global Severity Index
FPI Satisfaction With Life

TSOS total
NA
NA
/.24**
.32**

Love of God

Love of Others

.64***
.73***
/.08
.16**

Love of Self
/.01
.25**
/.49***
.47**

.10*
.27**
/.14**
.35**

Note. TSOS/Theistic Spiritual Outcome Survey; NA /not applicable; SCL-90-R/Symptom Checklist-90-Revised; FRI/Freiburg
Personality Inventory.
*p B/ .05. **p B/.01. ***p B/.001.

.90. The internal consistency reliability estimates for

the TSOS-GT subscales were as follows: .91 for the
Love of God (closeness to God) subscale, .76 for the
Love of Others (love of humanity) subscale, and .86
for the Love of Self (moral self-acceptance) subscale.
These reliabilities are comparable to those obtained
for the English version of the TSOS.
Correlations with importance of personal religious
belief. Table V presents the correlations of the
TSOS with patients ratings of the importance of
their personal religious beliefs. The TSOS Love of
God subscale was significantly and positively correlated (.64) with participants rating of importance of
personal religious beliefs. This indicates that those
who view their religious beliefs as highly important
tend to feel more love and closeness to God.
The TSOS subscales were also significantly and
positively correlated with participants ratings of the
degree to which their personal religious beliefs are a
source of consolation and strength. The TSOS
Love of God subscale was most strongly correlated
with this item. The positive correlations between the
TSOS subscales and this item provided some
evidence that the type of theistic spirituality assessed
by the TSOS is a positive resource in peoples lives.
Correlations with psychological functioning. The
Pearson correlations of the TSOS with the SCL90-R and FPI are given in Table V. The TSOS and
the TSOS subscales were negatively associated with
the SCL-90-Rs GSI index, ranging from
/.08 to

/.49. Interestingly, the Love of Self (moral selfacceptance) subscale of the TSOS correlated more
highly with the SCL-90-Rs GSI index than did the
other TSOS subscales. The TSOS and the TSOS
subscales were positively associated with the FPI
Satisfaction With Life subscale. Again, the Love of
Self subscale of the TSOS correlated most highly
with this measure. Consistent with the findings of
Study 2, the TSOS and TSOS subscales manifested
low but statistically significant positive associations
with the measures of psychological functioning. We
interpreted this correlation pattern as evidence that

the TSOS is assessing a meaningful construct in its

own right. It captures aspects of general religiosity
and psychosocial functioning as the moderate correlations with the various subscales show but is not
identical to these constructs.
Treatment outcome results. Table VI presents the
pre- and posttreatment mean TSOS-GT scores for
Clinic 1 and Clinic 2. It can be seen that patients in
the spiritually oriented treatment center (Clinic 1)
had larger increases on all three of the TSOS-GT
subscales than did those in the secular treatment
center (Clinic 2). The TSOS total scores for each
group also indicate that there was not a problem with
ceiling effect because the mean scores remained
below the theoretical midpoint of the TSOS scale
even after treatment.
General discussion
The three studies reported here provide initial
evidence that the TSOS is a potentially useful
outcome measure for psychotherapy research. Study
Table VI. Study 3: TSOS pre- and posttreatment means, standard
deviations, and standardized mean differences in german treatment clinics.
Spiritually oriented clinic
Scale
TSOS total
M
SD
Love of God
M
SD
Love of Others
M
SD
Love of Self
M
SD

Pre

Post

37.4
12.0

44.9*
11.7

14.5
5.3

17.3*
5.0

13.1
4.2

14.8*
4.0

9.8
4.5

12.8*
4.2

Secular clinic
Pre

Post

39.8
11.8

40.3
10.7

12.7
5.7

12.1
5.5

15.7
4.2

15.3
3.6

11.1
4.3

12.7*
3.9

.63

d
.04

/.11

.49

/.10

.41

.69

.39

Note. TSOS/Theistic Spiritual Outcome Survey.

*p B/ .01.

466

P. Scott Richards et al.

1 provided evidence concerning the factor structure,

reliability, and validity of the TSOS with a normal
young adult sample. Study 2 provided similar
evidence with a sample of clinical inpatients.
Furthermore, Study 2 provided some evidence
that the TSOS is a sensitive measure of some theistic
outcomes of a spirituality group for eating-disordered inpatients. Study 3 provided initial information about a German language version of the TSOS
and in so doing also provided some cross-validation
evidence concerning the psychometric properties
and clinical usefulness of the TSOS. Study 3 also
provided further evidence that the TSOS is a
sensitive measure of some theistic outcomes of a
spiritually based treatment program. Both Studies 2
and 3 provided some initial evidence that the TSOS
may not be susceptible to problems related to ceiling
effects.
The TSOS does not measure all dimensions of
theistic spirituality, but the findings of these three
studies provide some evidence that it validly measures several of them, including love of God, love of
others, and moral self-acceptance (Richards & Bergin, 1997). Fortunately, as mentioned, there are
many other measures that assess other aspects of
theistic spirituality that may be of interest to
researchers. The process of investigating and clarifying the relationship of the TSOS with other measures of theistic religiousness and spirituality was
begun in the studies reported previously. The TSOS
Love of God subscale is positively associated with
intrinsic religiousness and the RWB subscale of the
SWBS, two widely used measures of theistic spirituality. The TSOS Love of Others and Love of Self
subscales are also positively, but relatively moderately, associated with religious well-being. Additional
studies are needed of the TSOS and other spirituality measures before their meaning and value in
relation to each other will be fully understood.
In addition to the evidence pertaining to the
psychometric properties of the TSOS, the findings
of the three studies reported here also have some
implications for theory and practice. First and foremost, the fact that a measure of spirituality correlated significantly with several validated measures of
psychological functioning is worthy of note. The
magnitude of correlations observed here is similar to
that of other widely investigated correlates of mental
health, such as genetic predisposition (Kendler,
Gardner, & Prescott, 1999) and cognitive irrational
beliefs (McDermut, Haaga, & Bilek, 1997). Our
findings, therefore, add evidence to the growing
body of research that has shown there are often

positive associations between spirituality and various

indicators of mental health (Koenig et al., 2001).
Second, the finding that the items comprising the
Love of Self subscale, which essentially assessed
moral self-acceptance, correlated significantly with
all measures of psychological outcome is particularly
noteworthy. As stated earlier, the data showed that
moral self-acceptance may be a meaningful correlate
of mental health. Although this conceptualization
clearly fits within a theistic model of mental health
(Richards & Bergin, 1997), secular theories of
personality and psychopathology have little to say
about the role of moral behavior. Congruence of
behavior with personal moral values is a topic
warranting additional work.
Finally, the findings of Studies 2 and 3 support the
notion that clinical treatment, especially spiritually
oriented treatment, can improve patients spiritual
outcomes, including their feelings of love for God,
others, and self. This possibility potentially expands
the conceptualization of the role of psychotherapists.
In cases in which clients desire improved spiritual
outcomes, such work can apparently be undertaken
with success (McCullough, 1999). It is also noteworthy, in light of fears expressed by some religious
people that psychotherapy may undermine faith and
spirituality (Richards & Bergin, 1997, 2000), that
patients scores in Studies 2 and 3 did not significantly decline during treatment. This was true even
of patients who received solely secular forms of
psychological treatment. Findings such as these
will, it is hoped, help allay fears about psychological
treatment in religious communities and may help
leaders and members in religious communities be
more inclined to use psychological services when
there is a need (Richards & Bergin, 2000).
A number of limitations of the current studies
need to be acknowledged. First, there was an overrepresentation of one religious group (LDS) in the
first two studies. Second, racial and ethnic diversity
of the samples in all three studies was lacking. Third,
the sample size in the second study was not large
enough to conduct a factor analysis of the TSOS
with that particular clinical sample. Fourth,
although the factor analyses reported in this study
were supportive of the construct validity of the
TSOS, a confirmatory factor analysis would be
useful to further assess the degree to which the
factor structure and item loadings of the TSOS
conform to theoretical expectations. Fifth, the TSOS
needs to be assessed with additional clinical populations, religious denominations, age groups, and
racial-ethnic groups. The TSOS is intended to be a
measure of theistic spiritual outcomes with Chris-

Development and validation of the TSOS

tian, Jewish, and Muslim individuals, but more data
are needed to determine whether it can be validly
used with clients from all of these traditions. Sixth,
in Study 3, participants were not randomly assigned
to the two treatment clinics and so the two patient
samples were undoubtedly nonequivalent. Thus,
differences in spirituality scores between the two
clinics at the conclusion of treatment cannot be
attributed with confidence to the spiritually based
treatment. Seventh, as with most self-report outcome measures, the TSOS may be susceptible to
social desirability and demand characteristics. Research is needed to assess when and to what extent
this may be a problem. Eighth, the TSOS items are
quite generic, having little grounding in specific,
traditional outcomes of psychotherapy. The TSOS
may, therefore, be useful as a multipurpose theistic
spiritual outcome measure but may lack the specificity desired in some situations. Finally, additional
normative data for the TSOS and TSOS-GT are
needed so that researchers and clinicians can more
confidently interpret TSOS scores for individuals
and groups. Such data would also make it more
feasible to calculate clinical cutoff scores to assist in
such interpretations.
Although further empirical work is needed in the
validation of the TSOS, the studies here tentatively
support its use as a clinical outcome scale and
research instrument. The neglect of the religious
and spiritual outcomes of treatment in the psychotherapy outcome field needs to change (P. C.
Hill & Pargament, 2003; Richards & Bergin, 1997).
We hope that this instrument will make it more
feasible for psychotherapy outcome researchers to
investigate the spiritual outcomes of psychotherapy
with their theistic clients. Such research will be
valuable for mental health professionals because it
will provide an empirical basis from which to
respond to questions and concerns from clients
and the public about the religious and spiritual
effects of psychotherapy.

References
Bergin, A. E. (1980). Psychotherapy and religious values. Journal
of Consulting and Clinical Psychology, 48 , 75
/105.
Bryk, A. S., & Raudenbush, S. W. (1992). Hierarchical linear
models: Applications and data analysis methods . Newbury Park,
CA: Sage.
Cooper, P. J., Taylor, M., Cooper, Z., & Fairburn, C. G. (1987).
The development and validation of the Body Shape Questionnaire. International Journal of Eating Disorders , 6 , 48
/494.
Derogatis, L. R. (1983). SCL-90: Administration, scoring, and
procedures manual for the revised version . Baltimore: Clinical
Psychometric Research.
Ellison, C. W. (1983). Spiritual well-being: Conceptualization and
measurement. Journal of Psychology and Theology, 11 , 330
/340.

467

Ellison, C. W., & Smith, J. (1991). Toward an integrative measure

of health and well-being. Journal of Psychology and Theology, 19 ,
35
/48.
Epstein, S., & OBrien, E. J. (1983). The Multidimensional Selfesteem Inventory. Odessa, FL: Psychological Assessment Resources.
Franke, G. H. (1995). Die Symptom-Checkliste von Derogatis .
Weinheim: Beltz.
Garner, D. M., & Garfinkel, P. E. (1979). The Eating Attitudes
Test: An index of the symptoms of anorexia nervosa. Psychological Medicine , 9 , 273
/279.
Gorsuch, R. (1997). Exploratory factor analysis: Its role in item
analysis. Journal of Personality Assessment , 68 , 532
/560.
Hall, T. W., & Edwards, K. J. (1996). The initial development and
factor analysis of the Spiritual Assessment Inventory. Journal of
Psychology and Theology, 24 , 233
/246.
Hall, T. W., Tisdale, T. C., & Brokaw, B. F. (1994). Assessment of
religious dimensions in Christian clients: A review of selected
instruments for research and clinical use. Journal of Psychology
and Theology, 22 , 395
/421.
Hill, C. H., & Hood, R. W. (1999). Measures of religiosity.
Birmingham, AL: Religious Education Press.
Hill, P. C., & Pargament, K. I. (2003). Advances in the
conceptualization and measurement of religion and spirituality.
American Psychologist , 58 , 64
/74.
Jagers, R. J., & Smith, P. (1996). Further examination of the
Spirituality Scale. Journal of Black Psychology, 23 , 429
/442.
Jensen, L. C., Jensen, J., & Wiederhold, T. (1993). Religiosity,
denomination, and mental health among young men and
women. Psychological Reports , 72 , 1157
/1158.
Jung, C. G. (1933). Modern man in search of a soul . New York:
Harcourt, Brace, & World.
Kass, J. D., Friedman, R., Lesserman, J., Zuttermeister, P., &
Benson, H. (1991). Health outcomes and a new index of
spiritual experience. Journal for the Scientific Study of Religion ,
30 , 203
/211.
Kazdin, A. E. (1994). Methodology, design, and evaluation in
psychotherapy research. In A. E. Bergin & S. L. Garfield (Eds),
Handbook of psychotherapy and behavior change (4th ed.) (pp.
19
/71). New York: Wiley.
Kendler, K. S., Gardner, C. O., & Prescott, C. A. (1999). Clinical
characteristics of major depression that predict risk of depression in relatives. Archives of General Psychiatry, 56 , 322
/327.
Koenig, H. G., McCullough, M. E., & Larson, D. B. (2001).
Handbook of religion and health . New York: Oxford University
Press.
Lambert, M. J., & Burlingame, G. M. (1996). The Outcome
Questionnaire . Stevenson, MD: American Professional Credentialing Services.
Lambert, M. J., & Hill, C. E. (1994). Assessing psychotherapy
outcomes and processes. In A. E. Bergin & S. L. Garfield (Eds),
Handbook of psychotherapy and behavior change (4th ed.). New
York: Wiley.
Levin, J., Taylor, R. J., & Chatters, L. M. (1994). Race and gender
differences in religiosity among older adults: Findings from four
national surveys. Journals of Gerontology, 49 , S137
/S145.
Maslow, A. (1964). Religion, values, and peak experiences . Columbus: Ohio State University Press.
May, R. (1982). The will and spirit: A contemplative psychology. San
Francisco, CA: Harper & Rowe.
McCullough, M. E. (1999). Research on religion-accommodative
counseling: Review and meta-analysis. Journal of Counseling
Psychology, 46 , 92
/98.
McDermut, J., Haaga, D. A. F., & Bilek, L. (1997). Cognitive bias
and irrational beliefs in major depression and dysphoria.
Cognitive Therapy & Research , 21 , 459
/476.

468

P. Scott Richards et al.

Miller, W. R. (1999). Integrating spirituality into treatment: Resources

for practitioners . Washington, DC: American Psychological
Association.
Myers, J. E., Sweeney, T. J., & Witmer, J. M. (2000). The wheel of
wellness counseling for wellness: A holistic model for treatment
planning. Journal of Counseling and Development , 78 , 251
/266.
Piedmont, R. L. (1999). Does spirituality represent the sixth
factor of personality? Spiritual transcendence and the fivefactor model. Journal of Personality, 67 , 985
/1013.
Richards, P. S., & Bergin, A. E. (1997). A spiritual strategy for
counseling and psychotherapy. Washington, DC: American Psychological Association.
Richards, P. S., & Bergin, A. E. (Eds.). (2000). Handbook of
psychotherapy and religious diversity. Washington, DC: American
Psychological Association.
Richards, P. S., Hardman, R. K., & Berrett, M. E. (2001, August).
Evaluating the efficacy of spiritual interventions in the treatment of
eating disorder patients: An outcome study. Paper presented at the
109th Annual Convention of the American Psychological
Association, San Francisco, CA.
Schowalter, M., Richard, M., Senst, R., & Murken, S. (in press).
Die Integration von Religiositat in der psychotherapeutischen
Behandlung bei religiosen Patienten */ein Klinikvergleich [The
integration of religious elements into inpatient psychotherapy
for religious patients */A comparison of two clinics]. Zeitschrift
fu
r Klinische Psychologie, Psychiatrie und Psychotherapie .
Sperry, L., & Shafranske, E. P. (2004). Spiritually-oriented
psychotherapy: Contemporary approaches . Washington, DC:
American Psychological Association.
Veach, T. L., & Chappel, J. N. (1992). Measuring spiritual health:
A preliminary study. Substance Abuse , 13 , 139
/147.
Worthington, E. L., Jr., Kurusu, T. A., McCullough, M. E., &
Sanders, S. J. (1996). Empirical research on religion and
psychotherapeutic processes and outcomes: A ten-year review
and research prospectus. Psychological Bulletin , 119 , 448
/487.
Worthington, E. L., Jr., Wade, N. G., Hight, T. L., Ripley, J. S.,
McCullough, M. E., Berry, J. W., et al. (2003). The Religious
Commitment Inventory */10: Development, refinement, and
validation of a brief scale for research and counseling. Journal of
Counseling Psychology, 50 , 84
/96.

Zusammenfassung
Entwicklung und Validierung des theistisch
spirituellen Ergebnisbogens
Die Autoren entwickelten den theistisch spirituellen Ergebnisbogen (Theistic Spiritual Outcome Survey
[TSOS]), um die spirituellen Ergebnisse von Psychotherapie aus einer theistisch spirituellen Sichtweise zu messen.
Eine 17-Itemversion des TSOS zeigte angemessene Werte
fur Zuverlassigkeit und Gultigkeit bei einer Stichprobe von
Collegestudenten. Drei Faktoren wurden extrahiert und
als Gottesliebe, Liebe zu anderen und Selbstliebe benannt.
Es gab signifikante Korrelationen mit den Maen anderer
psychologischer Tests. In den nachfolgenden Analysen
wurde TSOS uber einen Zeitraum von acht Wochen einer
Stichproben essgestorter Patientinnen gegeben und an
zwei Stichproben stationarer Patienten in psychotherapeutischer Behandlung in Deutschland. Die Ergebnisse dieser
Untersuchungen weisen ebenfalls auf gute psychometrische Eigenschaften und eine klinische Sensibilitat von
TSOS hin.

Resume
Developpement et validation du Theistic Spiritual
Outcome Survey
Les auteurs ont developpe le Theistic Spiritual Outcome
Survey (TSOS) pour mesurer les issues spirituelles de la
psychotherapie dun point de vue theiste et spirituel. Une
version en 17 items du TSOS a montre une fidelite et une
validite adequates dans un echantillon detudiants universitaires. Trois facteurs emergeant des analyses correspondaient aux sous-echelles de lAmour de Dieu,
lAmour des autres et lAmour de soi. Des correlations
avec des mesures du resultat psychologique etaient statistiquement significatives. Dans des analyses consecutives, le
TSOS a ete administre pendant 8 semaines a` un echantillon de femmes hospitalisees souffrant de troubles de
lalimentation ainsi qua` 2 echantillons issus de cliniques
psychologiques en Allemagne. Les analyses resultant de
ces etudes soutenaient egalement les proprietes psychometriques et la sensibilite clinique du TSOS.

Resumen
Desarrollo y validacion del relevamiento testico
espiritual
Los autores desarrollaron el Relevamiento del Resultado
Teista Espiritual (TSOS) para medir los resultados espirituales de la psicoterapia desde una perspectiva espiritual
testa. Se considero que la version de 17-tems de la TSOS
tena una confiabilidad y validez adecuadas para una
muestra de estudiantes univesitarios. De los analisis
emergieron tres factores que correspondieron a las subescalas de Amor a Dios, Amor a Otros y Amor a s mismos.
Las correlaciones con las mediciones de los resultados
psicologicos fueron estadsticamente significativs. En analisis subsiguientes, se administro el TSOS por un perodo
de ocho semanas a una muestra de pacientes femeninas
internadas por desordenes alimentarios y a dos muestras
de internadas en clnicas psicologicas de Alemania. Los
analisis resultantes de estos estudios tambien confirmaron
las propiedades psicometricas y la sensibilidad clnica del
TSOS.

Resumo
Desenvolvimento e Validacao do Inventario de
Resultado Testa-Espiritual
Os autores desenvolveram o Inventario de Resultado
Testa-Espiritual (IRTE) para avaliar os resultados espirituais da psicoterapia a partir de uma perspective testaespiritual. A versao de 17itens do IRTE foi considerada
adequada em termos de fidelidade e validade numa
amostra de estudantes universitarios. A partir das analises
emergiram tres factores que correspondem a`s subescalas
de Amor de Deus, Amor de Outros e Amor-proprio. As
correlacoes com as medidas psicopatologicas foram estatisticamente significativas. Em analises subsequentes, o
IRTE foi administrado durante um perodo de 8 semanas a
uma amostra de mulheres internadas com perturbacao
alimentar e a duas amostras de pacientes internados em

Development and validation of the TSOS

clnicas na Alemanha. As analises resultantes destes
estudos corroboraram as propriedades psicometricas e a
sensibilidade clnica do IRTE.

Sommario
Sviluppo e convalida dei risultati dei sondaggi
teistico spirituali
Gli autori hanno sviluppato la TSOS per misurare i
risultati spirituali di una psicoterapia da una prospettiva
spirituale teistica. Una versione di 17 item del TSOS e`
stata trovata per avere laffidabilita` e validita` sufficiente in
un campione di studenti delluniversita`. Tre fattori sono
emersi dalle analisi che corrispondevano alle sottoscale :
lamore di Dio, lamore di altri, lamore di se`. Le
correlazioni con le misure dei risultati psicologici erano

469

statisticamente significative. Nelle analisi successive, il

TSOS e` stato amministrato per un periodo di otto
settimane ad un campione di donne ricoverate con i
disturbi del comportamento alimentare ed a 2 campioni
di pazienti ricoverate in cliniche psicologiche in Germania.
Le analisi che derivano da questi studi supportavano le
proprieta psicometriche e la sensibilita` clinica del TSOS.