Dear fellow researchers

I would like to know what I need to take into consideration in my analysis of measurement invariance, given that the groups I want to compare do not all exceed the generally recommend minimum sample size of n=200 per group (Koh & Zumbo 2008, S. 475-476; Meade & Lautenschlager, 2004, S. 66–68).

Here I have listed the groups I would like to test for scalar invariance using Mplus with the number belonging to each group in parentheses:

Gender

male (253)

female (176)

Distribution of questionnaire

PAPI (319)

Online (116)

Subjects

Civic Education (127)

History (215)

Geography (89)

Current state of education

Bachelor of Arts (187)

Master of Education (187)

In-service teacher (53)

I was thinking perhaps stricter cut-off values of the RMSEA und CFI than the often cited

More Marcel Grieger's questions See All
Similar questions and discussions