| Summary: | (1) Background: In self-concept research, Likert scales are still relied upon despite the fact that they pose methodological difficulties for research in culturally diverse societies. This calls the validity of the data into question. In the present study, we develop a mixed methods design for culture-sensitive academic self-concept research. We test it in a study about chemistry self-concept with secondary school students; (2) Methods: Interview (<i>N</i> = 43) and questionnaire (<i>N</i> = 116) data were collected; (3) Results: The mixed methods approach allowed connecting self-concept with culturally shaped narratives: in the quantitative data, we found the well-documented gender gap in favor of boys. However, among the students with a Turkish migration background, the girls showed stronger chemistry self-concepts. The interviews suggested that girls with Turkish migration background find it easier to connect their chemistry learning to their personal life than the boys with Turkish migration background; (4) Conclusion: Based on further literature, we hypothesize that these differences might be due to a less masculine conception of science in the Turkish society. The mixed methods approach allows detecting measurement bias, which increases the validity of science self-concept data in culturally diverse contexts.
|
|---|
