2014 IJRSLL – Volume 3 Issue 6
Hamedi, Seyyedeh Mina*
Ferdowsi University of Mashhad, Iran (firstname.lastname@example.org)
Ferdowsi University of Mashhad, Iran (email@example.com)
Ferdowsi University of Mashhad, Iran (firstname.lastname@example.org)
The present study examined the construct validity of Self-Efficacy Scale for English Language Learners’ Textbooks (SES-ELLT) using Rasch measurement model. To this end, a sample of 150 language learners from various language institutes participated in the study for the investigation of psychometric properties of SES-ELLT, particularly the reliability and construct validity of the scale. The participants were asked to rate the importance of 36 self-efficacy items for the English language textbooks. Through the Rasch analysis scale reliability, item fit, unidimensionality, and category functioning were investigated. Consequently, the findings demonstrated a fairly high reliability both for the person separation (0.81), and the item separation (0.87). Moreover, three items were detected as misfitting and were discarded, accordingly. In all, the scale exhibited a good unidimensionality and the items sufficiently covered the range of person abilities; however, a 4-category rating scale seemed to be a better alternative as the third and fourth thresholds did not have the required advance between step calibrations in a 5-category rating scale.
Keywords: self-efficacy; textbooks; Rasch measurement model; reliability; construct validity