See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/232541865

Factorial and Construct Validity of the Sibling Relationship Questionnaire

Article  in  European Journal of Psychological Assessment · January 2010

DOI: 10.1027/1015-5759/a000037

CITATIONS READS

23 5,698

4 authors, including:

Ron H J Scholte William Martin Van der Veld

Radboud University Radboud University
259 PUBLICATIONS   6,811 CITATIONS    53 PUBLICATIONS   798 CITATIONS   

SEE PROFILE SEE PROFILE

Rutger C M E Engels
Trimbos Institute / Utrecht University
780 PUBLICATIONS   24,023 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Substance use View project

Efficacy of Art Therapy in patients with personality disorders cluster B/C. A randomized controlled trial. View project

All content following this page was uploaded by Ron H J Scholte on 01 June 2014.

The user has requested enhancement of the downloaded file.

 M. M. S. Derkman
European JournalofetPsychological
al.: Factorial Assessment
and Co nstruct
©2010;
2010Validity
Vol.
Hogrefe ofPublishing
the SRQ
26(4):277–283

Original Article

Factorial and Construct

Validity of the Sibling
Relationship Questionnaire
Marleen M. S. Derkman, Ron H. J. Scholte, William M. Van der Veld,
and Rutger C. M. E. Engels
Radboud University Nijmegen, The Netherlands

Abstract. The psychometric properties of the shortened version of the Sibling Relationship Questionnaire (SRQ; Furman & Buhrmester,
1985) were examined in a sample of 428 Dutch adolescents, aged 13 to 16. In order to examine the dimensions Warmth/Closeness and
Conflict, as proposed by Furman and Buhrmester, we first conducted confirmatory factor analyses. Second, we investigated the construct
validity by correlating Warmth/Closeness and Conflict with internalizing and externalizing behaviors and the quality of relationship with
parents. Third, we determined the internal reliability of Warmth/Closeness and Conflict and the underlying qualities by computing
Cronbach’s αs. The results confirmed the dimensions Warmth/Closeness and Conflict as major aspects of sibling relationships. Moreover,
the construct validity of the instrument and internal consistency for these dimensions and the underlying qualities proved to be good. Our
findings underscore the psychometric properties of the SRQ, in that it appears to be a valid and reliable measure to assess Warmth/Close-
ness and Conflict in sibling relationships.

Keywords: SRQ, CFA, validity, internalizing and externalizing behaviors

In adolescents’ daily lives, siblings play important roles. In Warmth and Conflict
fact, sibling relationships are often among an individual’s
longest lasting relationships. According to Blyth, Hill, and Warmth and conflict are basic aspects to describe sibling
Smith Thiel (1982), more than two-thirds of adolescents relationships (e.g., Kim, McHale, Osgood, & Crouter,
with siblings listed their siblings as significant persons in 2006; Scharf, Shulman, & Avigad-Spitz, 2005) or to pre-
their lives and, on average, 77% of adolescents’ siblings dict adjustment in adolescence (e.g., Kim, McHale, Crou-
were listed as important persons. Moreover, siblings rated ter, & Osgood, 2007; Stocker, 1994; Stocker et al., 2002).
each other as important sources of companionship, intima- Warmth and conflict were also used to investigate sibling
cy, and nurturance (Lempers & Clark-Lempers, 1992). So, relationship trajectories in adolescence (e.g., Feinberg, Mc-
by being an influential role model, and a source of help, Hale, Crouter, & Cumsille, 2003) and to assess the quality
emotional support, and companionship, siblings might of sibling relationships (e.g., Richmond et al., 2005). These
have major and unique effects on each other’s development studies indicate that warmth and conflict represent impor-
(Furman & Buhrmester, 1985). Therefore, sibling relation- tant aspects of the quality of sibling relationships and, giv-
ships might be significant in predicting adolescents’ emo- en their importance, we focused on these two aspects in the
tional and behavioral adjustment (Noller, 2005). present study.
To assess aspects of the sibling relationship, Furman and Furman and Buhrmester’s study (1985) indicates that the
Buhrmester (1985) developed the Sibling Relationship dimension Warmth/Closeness consists of seven underlying
Questionnaire (SRQ), which has been applied in a number qualities (i.e., Intimacy, Prosocial Behavior, Companion-
of studies (e.g., Richmond, Stocker, & Rienks, 2005; ship, Affection, Similarity, Admiration of the sibling, and
Stocker, Burwell, & Briggs, 2002; Van der Vorst, Engels, Admiration by the sibling) and that the dimension Conflict
Meeus, Dekovi2, & Van Leeuwe, 2007). However, there is can be divided into three underlying qualities (i.e., Quar-
little research on the factorial and construct validity of the reling, Antagonism, and Competition). So, according to
SRQ. Therefore, in this study, we aimed to test the theoret- Furman and Buhrmester (1985), underlying the SRQ is a
ical model underlying the SRQ by conducting confirmatory second-order factor model. The first-order factors are the
factor analyses (CFA) and examining its construct validity qualities and the second-order factors are the overarching
and internal reliability. dimensions (i.e., Warmth/Closeness and Conflict). To our

© 2010 Hogrefe Publishing European Journal of Psychological Assessment 2010; Vol. 26(4):277–283
DOI: 10.1027/1015-5759/a000037
278 M. M. S. Derkman et al.: Factorial and Construct Validity of the SRQ

knowledge, CFA that test the theoretical model of the SRQ Conflict are frequently reported. However, detailed infor-
have not yet been reported. Therefore, the first aim of this mation about the internal reliabilities of the 10 underlying
study was to investigate the structures of the dimensions qualities is currently lacking. The third aim of our study
Warmth/Closeness and Conflict and the underlying quali- was, therefore, to examine the internal reliability of the two
ties, by conducting CFA in a sample of Dutch adolescents. dimensions Warmth/Closeness and Conflict and of the 10
underlying qualities.
In sum, the SRQ was administered to a large sample of
Construct Validity Dutch siblings in order to (1) test the factor structure, (2)
test the construct validity, and (3) test the internal consis-
The second aim of our study was to investigate the con- tency of the two dimensions Warmth/Closeness and Con-
struct validity of the dimensions Warmth/Closeness and flict and the 10 underlying qualities. To examine whether
Conflict. Moser and Jacob (2002) investigated the con- Warmth/Closeness and Conflict are universal aspects of
struct validity by correlating the scores on the SRQ with sibling relationships, these aims were also tested over time.
the scores on the Family Environment Scale. They con-
cluded that the construct validity of the SRQ was adequate.
We aimed to test to what extent Warmth/Closeness and
Conflict were associated with internalizing and externaliz- Method
ing behaviors. Previous studies have found associations be-
tween sibling relationships and internalizing and external- Participants and Procedure
izing behaviors. For example, Stocker (1994) found that
warmth in sibling relationships was associated with higher For the present study, data were collected as part of a broad-
self-worth and with fewer conduct problems and feelings er longitudinal study called “Family and Health”, in which
of loneliness. Also, more conflicts in sibling relationships different socialization processes underlying various health
have been related to adolescents depressed moods and ex- behaviors in adolescence were examined (for details of the
ternalizing behavior (Kim, Hetherington, & Reis, 1999; sample selection, see Van der Vorst et al., 2007). The ad-
Stocker et al., 2002). Therefore, we expected that in sibling dresses of families with at least two children aged between
relationships Warmth/Closeness would be negatively relat- 13 to 16 were selected from 22 municipalities in the Neth-
ed and Conflict would be positively related to internalizing erlands. Of the approximately 5000 families that received
and externalizing behaviors. a letter, 885 families agreed to participate by returning the
To investigate the construct validity further, we exam- response forms. The remaining families were contacted by
ined whether Warmth/Closeness and Conflict were associ- phone to establish whether they fulfilled the inclusion cri-
ated with the quality of relationship with parents. In the teria: They had to be full families with two biologically
literature, the congruence hypothesis is commonly men- related adolescent children who were neither twins nor
tioned to describe possible links between family relation- mentally or physically disabled. Families who did not ful-
ships (e.g., Noller, 2005). This hypothesis proposes that fill these criteria were excluded from the study; as a result
qualities of different family relationships are positively as- 765 families remained. Because of financial constraints we
sociated with each other, so that a high quality parent-child were only able to select a maximum of 428 families, with
relationship is related to a high quality sibling relationship. an equal number of adolescents from lower education, in-
Following the congruence hypothesis, we, thus, expected termediate general education, and preparatory college and
more Warmth/Closeness and less Conflict in sibling rela- university education. All participants gave their written in-
tionships to be associated with high quality relationships formed consent.
with parents. Trained interviewers visited the families at their homes
for the first time between November 2002 and April 2003
(T1) and for the second time 1 year later (T2). In the second
Internal Reliability wave, 416 families were still included in the study. During
these home visits each family member filled in the exten-
Our third aim was to investigate the internal reliability of sive questionnaire without discussing its questions or an-
the dimensions Warmth/Closeness and Conflict and their swers with each other. When all four family members had
underlying qualities. Several studies have reported on the completed the questionnaires, each family received 30
internal reliability of the SRQ (e.g., Richmond et al., 2005; Euros for their participation. At the end of the study five
Stocker et al., 2002). For example, the original study of travel-cheques of 1000 Euros each were raffled between
Furman and Buhrmester (1985) reported Cronbach’s αs the families that took part in the entire “Family and Health”
that exceeded .70 for all 10 underlying qualities, with the study.
exception of the quality Competition. Moser and Jacob At the beginning of the project, the mean age of the
(2002) reported Cronbach’s αs of .90 and .68 for the di- mothers and fathers was 43.82 (SD = 3.57) and 46.18 (SD =
mensions Warmth/Closeness and Conflict, respectively. 4.00), respectively. The mean age of the older and younger
Cronbach’s αs of the dimensions Warmth/Closeness and siblings was 15.22 (SD = .60; range 14 to 17 years) and

European Journal of Psychological Assessment 2010; Vol. 26(4):277–283 © 2010 Hogrefe Publishing
 M. M. S. Derkman et al.: Factorial and Construct Validity of the SRQ 279

13.36 (SD = .50; range 13 to 15 years), respectively. Of the and depression was .51, p < .01). Higher scores on inter-
older adolescents, 98.1% were of Dutch origin and of the nalizing behavior indicated more internalizing problems.
younger adolescents 96.3% were of Dutch origin. Further,
53% of the older and 47.7% of the younger siblings were
male. Externalizing Behavior

To test the construct validity, we measured externalizing

Measures behavior. Siblings filled in 13 items about whether they had
been involved in various delinquent activities, such as
Sibling Relationship Questionnaire (SRQ) stealing and vandalism, in the past 12 months (Scholte, En-
gels, De Kemp, Harakeh, & Overbeek, 2007). The possible
The SRQ is a self-report questionnaire to assess dimensions answers included 0 times, 1–3 times, 4–6 times, 7–12
of sibling relationships (Furman & Buhrmester, 1985). To times, and more than 12 times. Cronbach’s α for this scale
create a Dutch version of the original English question- was .77. Higher scores on this scale indicated more exter-
naire, the SRQ was first translated into Dutch and then back nalizing behavior.
into English by two researchers of the Radboud University
Nijmegen. They discussed the translation with a language
consultant and with that feedback the final translation of Quality of Relationship with Parents
the SRQ was established (see Van der Vorst et al., 2007).
The Dutch items of the SRQ are available from the authors. To assess the quality of relationship with parents, the par-
The first dimension of the SRQ is Warmth/Closeness (15 ticipants filled in 12 items of the Inventory of Parent and
items). An example of a Warmth/Closeness item is, “How Peer Attachment for both their fathers and mothers (IPPA;
much do you and this sibling do nice things for each other?” Armsden & Greenberg, 1987). The participants answered
The second dimension is Conflict (6 items). An example of items such as “My mother/father accepts me the way I am,”
a Conflict item is, “How much do you and this sibling dis- on a 6-point scale, ranging from 1 (never) to 6 (always).
agree and quarrel with each other?” The siblings responded Internal reliability for the quality of the relationship with
to each item on a 5-point scale, ranging from 1 (hardly at father and mother was good, with Cronbach’s αs of .85 and
all) to 5 (extremely much), assessing the extent to which .82, respectively. In order to create a score of the quality of
certain features were characteristic of their sibling relation- the relationship with parents, 8 items had to be recoded and
ship. A higher score on Warmth/Closeness and Conflict in- a mean score of the quality of relationship with father and
dicated respectively more warmth/closeness and conflict in relationship with mother was computed (Pearson correla-
the sibling relationship. tion of relationship with father and mother was .60, p <
.01). Higher scores on the quality of relationship with par-
ents indicated higher quality of relationships between par-
Internalizing Behavior ticipants and their parents. This scale was used to test the
construct validity of the SRQ.
Internalizing behavior was used to test the construct valid-
ity of the SRQ. To measure internalizing behavior, a sum
score was computed of self-esteem and depression. Both Strategy for Analyses
siblings answered 10 items about their self-esteem (Rosen-
berg Self-esteem Scale; Rosenberg, 1965) and 6 items The factorial validity was tested in two steps. In the first
about their depressive feelings (Kandel & Davies, 1982). step we evaluated the factor structure underlying the SRQ.
An example of a self-esteem item is “Overall, I am satisfied To avoid dependency in the data, we selected one sibling
with myself.” Siblings answered these items on a 4-point from each family using a randomization procedure result-
scale that ranged from 1 (does not fit me at all) to 4 (fits me ing in a sample of N = 401. In the second step, we tested
well). Five items were recoded; higher scores indicated whether the factorial structure found in the first step had
lower self-esteem. The scale is internally reliable, with a measurement invariance over time (T1 vs. T2). This was
Cronbach’s α of .86. To measure depression, the partici- accomplished with multigroup factor analysis. It is impor-
pants answered on a 5-point scale ranging from 1 (never) tant to establish measurement invariance when the same
to 5 (always) whether certain feelings, such as feeling ner- group is compared across time (Kuhn & Holling, 2009;
vous and tensed, occupied them during the last 12 months. Meredith, 1993). We tested both the metric invariance and
Higher scores on this scale indicated more depressive feel- scalar invariance, by putting across-group constraints on
ings. Internal reliability for the depression scale was good, the factor model. Since the factorial structure underlying
with a Cronbach’s α of .78. We computed mean scores of the SRQ is a second-order factor model, scalar invariance
self-esteem and depression in order to get a score on inter- cannot be tested in a straightforward way as is the case for
nalizing behavior (Pearson correlation between self-esteem first-order factor models. We followed the suggestions of

© 2010 Hogrefe Publishing European Journal of Psychological Assessment 2010; Vol. 26(4):277–283
280 M. M. S. Derkman et al.: Factorial and Construct Validity of the SRQ

Figure 1. Standardized estimates* belonging to the first- and second-order factor loadings of the SRQ. Note. *All estimates
are significant (p < .05).

Byrne and Stewart (2006) and, thus, assumed that the first- Data Preparation
order factors had means of zero.
Current practice to evaluate model fit is by the use of fit Listwise deletion was used to deal with missing data. We
indices, specifically the RMSEA. Recent studies have, how- analyzed the covariance matrix, estimated with Prelis2
ever, shown that fit indices with fixed critical values (e.g., (Jöreskog & Sörbom, 1996b) under the assumption that the
RMSEA, GFI) cannot control for type I and type II errors data were continuous. Estimation of the factor models was
(Barret, 2006; Marsh, Hau, & Wen, 2004; Saris, Satorra, Van done with the maximum likelihood procedure in Lisrel 8.8
der Veld, 2009). As a consequence, correct theories are reject- (Jöreskog & Sörbom, 1996a), even though some variables
ed and incorrect theories are accepted in unknown rates. An were not normally distributed. However, robustness studies
alternative procedure, the detection of misspecifications, has indicated that the so-called “quasi-maximum likelihood”
recently been suggested by Saris and colleagues (2009). estimator was robust under quite general conditions (e.g.,
Since the procedure of Saris et al. (2009) is a relatively new Satorra, 1992).
approach we will also report the χ² test and other fit measures.
To examine the second aim – the construct validity of
Warmth/Closeness and Conflict – we added internalizing Results
and externalizing behaviors and the quality of relationship
with parents to the model in Lisrel 8.8 (Jöreskog & Sörbom, Factorial Validity
1996a) and computed correlations between these variables
and Warmth/Closeness and Conflict. To investigate the The measurement model we tested is depicted in Figure 1.
third aim of our study – the internal reliability of the two The output indicated that there were three serious misspec-
dimensions Warmth/Closeness and Conflict and the under- ifications, that is, two correlated errors (between Items 14
lying qualities – we computed Cronbach’s αs. and 15, and between Items 30 and 31; to be consistent with

European Journal of Psychological Assessment 2010; Vol. 26(4):277–283 © 2010 Hogrefe Publishing
 M. M. S. Derkman et al.: Factorial and Construct Validity of the SRQ 281

Table 1. Relationship of warmth/closeness and conflict Table 2. Internal consistency for the dimensions warmth/
with internalizing behavior, externalizing behav- closeness and conflict and the 10 underlying qual-
ior, and the quality of relationship with parents ities
Correlation between r Dimensions Cronbach’s Qualities Cronbach’s α/
Warmth/closeness and α Pearson
correlation
Internalizing behavior –.16*
Warmth/ .94 Intimacy .87
Externalizing behavior –.23* Closeness
Relationship with parents .38* Prosocial behavior .77
Conflict and Companionship .73
Internalizing behavior .13* Affection .87
Externalizing behavior .16* Similarity .76
Relationship with parents –.26* Admiration of sibling .74
Warmth/closeness and conflict –.40* Admiration by sibling .78
Internalizing and externalizing behavior .15* Conflict .93 Quarreling .80
Note. *p < .05. Antagonism .84
Competition .56
Furman and Burhmester’s study, we used the original num-
bering for the items) and one correlated disturbance term
(between the first-order factors Admiration of Sibling and SRQ. Finally, Table 1 also presents the correlation between
Admiration by Sibling). After introducing these parame- Warmth/Closeness and Conflict. As expected, this was a
ters, no other relevant misspecifications were found in the negative correlation that was stronger than any other cor-
model. The global fit measures for this model were: NT- relation in the table, which also supports construct validity.
χ² = 436 (df = 177), RMSEA = 0.061, sRMR = 0.054, The global fit measures for this model were: NT-χ² = 565
NNFI = 0.98, GFI = .91, and CFI = .98. (df = 234), RMSEA = 0.060, sRMR = 0.057, NNFI = 0.97,
The first step was to establish the factor structure under- GFI = .89, and CFI = .97.
lying the SRQ. The completely standardized estimates of The third aim of this study was to assess the internal
the first- and second-order factor loadings of the SRQ are consistency (Cronbach’s α) of the 10 qualities and the two
presented in Figure 1. Most of the first-order factor load- overarching dimensions of the SRQ. Internal consistency
ings were high, with loadings above .70. The factor load- proved to be good, with all Cronbach’s αs for the qualities
ings on Items 9 (Companionship), 17 (Prosocial Behavior), – except for Competition – ranging from 0.73 to 0.87 (see
and 30 (Admiration of sibling) were satisfactory, with fac- Table 2). Note that most of the qualities were measured by
tor loadings above .60, however, with a relatively poor two items; therefore, Cronbach’s α equals the Pearson cor-
(.38) loading on Item 29 (Competition). The second-order relation between the items. Cronbach’s αs for the overarch-
factor loadings were all above .70, except for the factor ing dimensions Warmth/Closeness and Conflict were 0.94
loadings of Warmth/Closeness on Similarity (.60) and of and 0.93, respectively. The internal consistency for these
Conflict on Competition (.63). All factor loadings were sig- dimensions was estimated using the correlations between
nificant (p < .05). the factors representing the qualities; that is, we treated the
In the second step, we conducted tests for measurement first-order factors as indicators. The estimates indicate that
invariance to see whether the factorial structure was invari- the dimensions Warmth/Closeness and Conflict and the 10
ant over time (T1 vs. T2). We first tested the most restricted qualities are measured rather reliably with the SRQ.
model (i.e., scalar invariance). The global fit measures for
this model were: NT-χ² = 770 (df = 392), RMSEA = 0.053,
NNFI = 0.98, and CFI = .98. Overall, the estimates indicate
that the indicators measured the constructs quite reliably Discussion
and, therefore, the SRQ is a good instrument to measure
Warmth/Closeness and Conflict across time. In this study, we tested the theoretical model underlying
the SRQ (Furman & Buhrmester, 1985) and its construct
validity and internal reliability. The first aim of this study
Construct Validity and Internal Reliability was to examine the dimensions Warmth/Closeness and
Conflict in a sample of Dutch adolescent siblings. We,
The output of the analyses for the construct validity of the therefore, first conducted a CFA, which indicated that
dimensions Warmth/Closeness and Conflict revealed no most items loaded on the subsequent qualities, and that
misspecifications. The correlations between the constructs the qualities measured the dimensions Warmth/Closeness
are presented in Table 1. All correlations in Table 1 support and Conflict reliably. To make sure we measured the same
our hypothesis, promoting the construct validity of the concept in a similar way over time, we conducted tests for

© 2010 Hogrefe Publishing European Journal of Psychological Assessment 2010; Vol. 26(4):277–283
282 M. M. S. Derkman et al.: Factorial and Construct Validity of the SRQ

measurement invariance. From these tests we can con- References

clude that the factor structure of Warmth/Closeness and
Conflict, as stated by Furman and Buhrmester (1985),
Armsden, G. C., & Greenberg, M. T. (1987). The inventory of par-
holds across time. Therefore, we can confirm the theoret-
ent and peer attachment: Individual differences and their rela-
ical model of the SRQ and conclude that the SRQ mea- tionship to psychological well-being in adolescence. Journal
sures Warmth/Closeness and Conflict in sibling relation- of Youth and Adolescence, 16, 427–454.
ships in the same way over time. We found a significant Barret, P. (2006). Structural equation modeling: Adjudging model
negative relationship between the two dimensions, which fit. Personality and Individual Differences, 42, 815–824.
indicates that a warmer/closer sibling relationship is asso- Blyth, D. A., Hill, J. P., & Smith Thiel, K. (1982). Early adoles-
ciated with less conflict in the relationship. cents’ significant others: Grade and gender differences in per-
Concerning the construct validity of the SRQ, more ceived relationships with familial and nonfamilial adults and
warmth/closeness and less conflict in sibling relation- young people. Journal of Youth and Adolescence, 11, 425–450.
ships are associated with less internalizing and external- Byrne, B. M., & Stewart, S. M. (2006). The MACS approach to
izing behaviors. Consequently, the links between the testing for multigroup invariance of a second-order structure:
SRQ and problem behaviors conform to our expectations. A walk through the process. Structural Equation Modeling,
Also, the results of this study support the congruence hy- 13, 287–321.
pothesis (e.g., Noller, 2005), in that, more warmth/close- Feinberg, M. E., McHale, S. M., Crouter, A. C., & Cumsille, P.
(2003). Sibling differentiation: Sibling and parent relationship
ness and less conflict in sibling relationships are associ-
trajectories in adolescence. Child Development, 74, 1261–1274.
ated with higher quality of relationship with parents. As
Furman, W., & Buhrmester, D. (1985). Children’s perceptions of
a consequence, we can conclude that the construct valid- the qualities of sibling relationships. Child Development, 56,
ity of the SRQ is adequate. 448–461.
The third aim of our study was to examine the reliabil- Jöreskog, K. G., & Sörbom, D. (1996a). LISREL® 8 User’s refer-
ity of the SRQ. For both the 2 dimensions and the 10 ence guide. Chicago, IL: Scientific Software International.
underlying qualities, with the exception of Competition, Jöreskog, K. G., & Sörbom, D. (1996b). PRELIS® 2 User’s refer-
the internal reliability scores were high. The quality ence guide. Chicago, IL: Scientific Software International.
Competition was reliable to a lesser extent; this is in line Kandel, D. & Davies, M. (1982). Epidemiology of adolescent de-
with the results of Furman and Buhrmester (1985) who pressive mood. Archives of General Psychiatry, 39, 1205–1212.
also found that the internal consistency coefficients for Kim, J., McHale, S. M., Crouter, A. C., & Osgood, D. W. (2007).
all qualities were high, except for the quality Competi- Longitudinal linkages between sibling relationships and ad-
tion. Competition was measured by two items, namely justment from middle childhood through adolescence. Devel-
“How much do you and this sibling try to out-do each opmental Psychology, 43, 960–973.
other at things?” and “How much do you and this sibling Kim, J., McHale, S. M., Osgood, D. W., & Crouter, A. C. (2006).
Longitudinal course and family correlates of sibling relation-
compete with each other?” The reason for the low reli-
ships from childhood through adolescence. Child Develop-
ability might be that these items measured different as-
ment, 77, 1746–1761.
pects of competition. Kim, J. E., Hetherington, E. M., & Reis, D. (1999). Associations
among family relationships, antisocial peers, and adolescents’
externalizing behaviors: Gender and family type differences.
Strengths and Limitations Child Development, 70, 1209–1230.
Kuhn, J. T. & Holling, H. (2009). Measurement invariance of di-
Although the construct and factorial validity of the SRQ vergent thinking across gender, age, and school forms. Euro-
have never been investigated in a sample of Dutch adoles- pean Journal of psychological assessment, 25, 1–7.
cents before, further research will be needed to examine Lempers, J. D., & Clark-Lempers, D. S. (1992). Young, middle,
whether our findings hold in other cultures. A limitation of and late adolescents’ comparison of the functional importance
this study may be that the variation of internalizing and of five significant relationships. Journal of Youth and Adoles-
externalizing behaviors was rather limited. Associations cence, 21, 53–96.
between Warmth/Closeness and Conflict in sibling rela- Marsh, H. W., Hau, K. T., & Wen, Z. (2004). In search of golden
tionships, on the one hand, and internalizing and external- rules: Comment on hypothesis testing approaches to setting
cut-off values for fit indices and dangers in overgeneralizing
izing behaviors, on the other, may be stronger in a sample
Hu and Bentler’s (1999) findings. Structural Equation Model-
of adolescents that display more internalizing or external- ling, 11, 320–341.
izing behaviors. Meredith, W. (1993). Measurement invariance, factor analysis,
We would like to stress that in this study we used CFA, and factorial invariance. Psychometrika, 58, 525–543.
which is a more thorough and stringent way of testing a Moser, R. P., & Jacob, T. (2002). Parental and sibling effects in
theoretical model than principal factor analysis. Our find- adolescent outcomes. Psychological Reports, 91, 463–479.
ings underscore the psychometric properties of the SRQ, Noller, P. (2005). Sibling relationships in adolescence: Learning
which appeared to be a reliable and valid measurement and growing together. Personal Relationships, 12, 1–12.
to assess Warmth/Closeness and Conflict in sibling rela- Richmond, M. K., Stocker, C. M., & Rienks, S. L. (2005). Longi-
tionships. tudinal associations between sibling relationship quality, pa-

European Journal of Psychological Assessment 2010; Vol. 26(4):277–283 © 2010 Hogrefe Publishing
 M. M. S. Derkman et al.: Factorial and Construct Validity of the SRQ 283

rental differential treatment, and children’s adjustment. Jour- Stocker, C. M., Burwell, R. A., & Briggs, M. L. (2002). Sibling
nal of Family Psychology, 19, 550–559. conflict in middle childhood predicts children’s adjustment in
Rosenberg, M. (1965). Society and the adolescent self-image. early adolescence. Journal of Family Psychology, 16, 50–57.
Princeton, NJ: Princeton University Press. Van der Vorst, H., Engels, R. C. M. E., Meeus, W., Dekovi2, M.,
Saris, W. E., Satorra, A., & Van der Veld, W. M. (2009). Testing & Van Leeuwe, J. (2007). Similarities and bidirectional influ-
structural equation models, or detection of misspecifications. ences regarding alcohol consumption in adolescent sibling
Structural Equation Modeling: A Multidisciplinary Journal, pairs. Addictive Behaviors, 32, 1814–1825.
16, 561–582.
Satorra, A. (1992). Asymptotic robust inferences in the analysis
of mean and covariance structures. Sociological Methodology,
22, 249–278. Marleen M. S. Derkman
Scharf, M., Shulman, S., & Avigad-Spitz, L. (2005). Sibling re-
Behavioral Science Institute
lationships in emerging adulthood and in adolescence. Journal
Radboud University Nijmegen
of Adolescent Research, 20, 64–90.
PO Box 9104
Scholte, R. H. J, Engels, R. C. M. E., de Kemp, R. A. T., Harakeh,
6500 HE Nijmegen
Z., & Overbeek, G. (2007). Differential parental treatment, sib-
The Netherlands
ling relationships, and delinquency in adolescence. Journal of
Tel. +31 24 361-2980
Youth and Adolescence, 36, 661–671.
Fax +31 24 361-2776
Stocker, C. M. (1994). Children’s perceptions of relationships
E-mail m.derkman@pwo.ru.nl
with siblings, friends, and mothers: Compensatory processes
and links with adjustment. Journal of Child Psychology, 35,
1447–1459.

© 2010 Hogrefe Publishing European Journal of Psychological Assessment 2010; Vol. 26(4):277–283

View publication stats