Rekkefølgeeffekter i spørreundersøkelser blant ungdom

Abstract

depressive symptoms is 3.3 percentage points higher if the bothered options are left-aligned, than if they are right-aligned. The analyses also show that the order effect is independent of gender, age and the amount of time spent to complete the questionnaire. On the other hand, this effect clearly varies according to the youth's academic performance. The effect is strongest among those with low school grades, while there is no effect at all among those with the highest grades.

Time trends in self-reported depressive symptoms, prescription of antidepressants, sedatives and hypnotics and the emergence of social media among Norwegian adolescents

“…In the surveys conducted in 1996 and 2006, the response options of the items assessing symptoms of depression and loneliness were in the reverse order from that presented above. This also applied to the survey in 2012, where half of the participants were asked questions with a reverse response order [ 28 ]. For statistical reasons, the response order in the surveys in 1996, 2006, and 2012 was re-coded in this study.…”

“…For statistical reasons, the response order in the surveys in 1996, 2006, and 2012 was re-coded in this study. Previous research has shown that the effect of the order is moderate, indicating that it is primarily other factors that affect how young people respond [ 28 ].…”

Surveyeksperimentet: Et underutnyttet forskningsdesign for sosiologisk kausalanalyse

“…For det andre har vi surveyeksperimenter der formålet er å utbedre selve måleinstrumentet (survey method experiments); jf. Bakken et al (2019). 5.…”