Do They Rate in the Same Way?

Abstract

Abstract. The internal structure of the Strengths and Difficulties Questionnaire (SDQ) has been widely discussed in the literature. The first purpose of this study was, therefore, to analyze the internal structure of the SDQ as completed by German parents and teachers. Parents and teachers of 1,135 children between the ages of 3 and 6 years in German daycare facilities were asked to complete the SDQ. Using confirmatory factor analysis (CFA), we found that the original five-factor model with added correlations between items with similar content and allowing cross-loads of reverse-worded items to the prosocial behavior factor is the model that fits the data the best. Additionally, we found significant mean differences between parents and teachers in four of the five SDQ subscales. Nevertheless, to compare the scores across both groups, it is necessary that the SDQ measures social-emotional and behavioral difficulties for both parents and teachers in the same way. Hence, the second purpose of this study was to test for measurement invariance of the SDQ across both groups. Using multigroup confirmatory factor analysis (MGCFA), we found that strict measurement invariance is tenable. Therefore, the observed means of the SDQ scores from parents and teachers are readily comparable and interpretable.

Interrater agreement and discrepancy when assessing problem behaviors, social‐emotional skills, and developmental status of kindergarten children

“…Interrater discrepancies were found to be stronger for externalizing than for internalizing problems (e.g., Grietens et al, 2004;Winsler & Wallace, 2002;but cf. In Rogge, Koglin, and Petermann (2018), parents reported more conduct problems and hyperactivity (and prosocial behavior) than did teachers, whereas teachers reported more emotional and peer problems. In Rogge, Koglin, and Petermann (2018), parents reported more conduct problems and hyperactivity (and prosocial behavior) than did teachers, whereas teachers reported more emotional and peer problems.…”

“…Berg-Nielsen et al, 2012). In Rogge, Koglin, and Petermann (2018), parents reported more conduct problems and hyperactivity (and prosocial behavior) than did teachers, whereas teachers reported more emotional and peer problems.…”

“…However, attention has been paid only recently to the issue of measurement invariance when examining interrater agreement and discrepancy (e.g., Narad et al, 2015;Rogge et al, 2018). For example, agreement might be artificially diminished or discrepancy artificially increased because items differ between the groups in their meaning or because the answer scale does not work equally for the groups.…”

“…For example, agreement might be artificially diminished or discrepancy artificially increased because items differ between the groups in their meaning or because the answer scale does not work equally for the groups. However, attention has been paid only recently to the issue of measurement invariance when examining interrater agreement and discrepancy (e.g., Narad et al, 2015;Rogge et al, 2018). Therefore, for most of the studies reviewed above, it remains unclear as to what extent their results reflect methodological artifacts or "real" agreement or discrepancy, respectively.…”

“…Therefore, for most of the studies reviewed above, it remains unclear as to what extent their results reflect methodological artifacts or "real" agreement or discrepancy, respectively. To our knowledge, only one study (Rogge et al, 2018) focusing on preschoolers has tested for measurement invariance. However, this valuable study relied on just one instrument, meaning ratings of the same behavior could not be compared or investigated across different instruments, and it did not include an assessment of the children's developmental status.…”