Construct Validity to Examine the Latent Traits of Teacher Self-Efficacy Instrument

Abstract

Teacher efficacy is an important debatable subject concerning its definition and measure. The purpose of this study was to evaluate the construct and convergent validity of Tschannen-Moran and Hoy (2001)’s ‘Teacher Self-Efficacy Scale (TSES)’. The study used the short version (12-item) TSES scale and Teacher Efficacy Scale (TSE) developed by Gibson and Dembo (1984) as an alternative measure. Total 412 in-service public school teachers from four provinces of Pakistan were selected including 242 (58.74%) female and 170 (41.26%) male teachers. The results show that Principal Axis Factoring yielded three factors, accounting for 64% of the total variance. The factor loading ranged from .50 to .89. Commonalities were in the range of 0.54 to 0.77 with an average of 0.64. There was significant a correlation between two efficacy measures, r (410) = .58, p < .001. Parallel analysis using 1000 sets of simulated data also suggested three factors of the TSES measure for in-service teachers. Female teachers tend to have a higher sense of efficacy than male teachers. The findings of this study provide useful information about to psychometric properties of teacher efficacy scale which further encourages researchers to revise teacher efficacy scale for more reliable and valid measurement. Teacher efficacy is an important debatable subject concerning its definition and measure. The purpose of this study was to evaluate the construct and convergent validity of Tschannen-Moran and Hoy (2001)’s ‘Teacher Self-Efficacy Scale (TSES)’. The study used the short version (12-item) TSES scale and Teacher Efficacy Scale (TSE) developed by Gibson and Dembo (1984) as an alternative measure. Total 412 in-service public school teachers from four provinces of Pakistan were selected including 242 (58.74%) female and 170 (41.26%) male teachers. The results show that Principal Axis Factoring yielded three factors, accounting for 64% of the total variance. The factor loading ranged from .50 to .89. Communalities were in the range of 0.54 to 0.77 with an average of 0.64. There was significant a correlation between two efficacy measures, r (410) = .58, p < .001. Parallel analysis using 1000 sets of simulated data also suggested three factors of the TSES measure for in-service teachers. Female teachers tend to have higher sense of efficacy than male teachers. The findings of this study provide useful information about to psychometric properties of teacher efficacy scale which further encourages researchers to revise teacher efficacy scale for more reliable and valid measurement.