Using moderated nonlinear factor models to adjust for differential item functioning in the Student-Teacher...

Area of Science:

• Child Development
• Educational Psychology
• Psychometrics

Background:

• Teacher-student relationships are vital for children's development.
• Previous studies on relationship quality lack bias adjustments.
• Measurement bias can distort findings on teacher-student interactions.

Purpose of the Study:

• To test and adjust for measurement bias in the Student-Teacher Relationship Scale (STRS).
• To generate accurate estimates of teacher-student relationship quality trajectories.
• To examine how bias affects the interpretation of predictors like race and socioeconomic status.

Main Methods:

• Utilized longitudinal data from the NICHD Study of Early Child Care and Youth Development (N=1140).
• Applied differential item functioning (DIF) analysis to STRS items across grades K-6.
• Compared growth models with and without bias adjustments for sociodemographic factors.

Main Results:

• Identified differential item functioning in items assessing conflict and closeness.
• Observed bias across child grade, birth sex, race/ethnicity, and maternal education.
• Bias adjustments revealed masked effects of race, education, and maternal sensitivity.

Conclusions:

• Testing and correcting for item bias in the STRS is essential for valid research.
• Unadjusted scores can lead to inaccurate conclusions about teacher-student relationships.
• Accurate measurement is critical for understanding factors influencing child development.