[Do different survey settings influence the prevalence of symptoms? A methodological comparison using the Youth Self-Report].

OBJECTIVES

The self-reported emotional and behavioural disorders among adolescents were assessed by the Youth Self-Report (YSR). The YSR was administered either in households or in classrooms. The goal of the study was to prove whether these different settings affect the prevalence rates of symptoms reported in the YSR.

METHODS

Mean scores and standard deviations of problem scales of two classroom samples and one household sample that was generally used as a reference were compared. The data were also compared with two classroom samples from Sweden and Greece. Statistical analyses were performed by means of t-test (unpaired), the evaluation of the magnitude of the effects by means of Cohen's criteria.

RESULTS

Classroom samples detected a significantly higher prevalence of symptoms than did household samples. This is the case for almost all of the problem scales in the YSR. The result of our study supports the finding that the setting of surveys that use self-administered questionnaires in classrooms themselves affect the prevalence of self-reported symptoms assessed by the YSR.

CONCLUSIONS

The results of surveys may be influenced, to a much greater degree than previously thought, by the settings in which they are administered. Further research is needed to identify the specific influences that differ for surveys administered at home, respectively at school.