Alfa Cronbacha – co daje dobre wyniki? Kilka uwag dotyczących budowania kwestionariuszy psychologicznych
Cronbach's alpha - what gives good results? Some notes on building psychological questionnaires

Summary/Abstract: Whatever Cronbach's alpha is measuring - it is not internal consistency, commonly mistakenly understood in psychology as the average strength of relationships between questionnaire items. questionnaire. In this article, we explore the reasons why understanding alpha as internal consistency is mis- internal consistency is erroneous, and we focus on the effect of inflation (overestimation) of the coefficient of- coefficient alpha in practice. Based on computer simulations, we determined the exact (joint) effects on the alpha value of: the number of respondents, the range of measurement scales (Likert), the number of questions in the questionnaire (items) and the average correlation between items. The results po- claim the occurrence of alpha level inflation due to the number of questions: alpha reaches pro- commendable values even with minimal internal consistency if there are a lot of items in the questionnaire. questionnaire has a lot of items. We suggest that in the case of poor measurement tools reliability may be overestimated due to the curvilinear alpha increase. The number of subjects and the range of scales did not affect the alpha value.

• Details
• Contents