Variable- and person-centered approaches to examining construct-relevant multidimensionality in writing self-efficacy

Area of Science:

• Educational Psychology
• Writing Studies
• Psychometrics

Background:

• Self-efficacy is crucial for student motivation and writing success.
• Existing research lacks empirical models for writing self-efficacy's multidimensionality.
• The Self-Efficacy for Writing Scale (SEWS) needed validation for its multidimensional structure.

Purpose of the Study:

• To examine the multidimensionality of writing self-efficacy.
• To provide validity evidence for an adapted SEWS.
• To explore student profiles based on writing self-efficacy.

Main Methods:

• Utilized a bifactor exploratory structural equation model with 1,466 8th-10th graders.
• Applied latent profile analysis to SEWS factor scores.
• Assessed concurrent, divergent, and discriminant validity using predictors and outcomes.

Main Results:

• A bifactor model best represented the SEWS, showing global and specific writing self-efficacy.
• Three distinct student profiles were identified based on global writing self-efficacy.
• Validity evidence was established through analysis of demographics, writing assessments, and grades.

Conclusions:

• The adapted SEWS effectively models the multidimensionality of writing self-efficacy in adolescents.
• Identified student profiles offer insights into varying levels of writing self-efficacy.
• Findings support the scale's utility for theoretical and practical applications in writing instruction.