The National Association for Research in Science Teaching (NARST), Maryland, United States Of America, 30 March - 02 April 2008, pp.265
The purpose of this study was to validate an instrument, Mentoring for Effective Primary Science Teaching
(MEPST), for a Turkish sample. Preservice elementary teachers (n=411) participated in the validation of the
MEPST from two universities in Turkey. MEPST measures preservice teachers’ perceptions about their mentors
based on their mentoring practices. The instrument was developed by considering constructivist theory. MEPST
included five dimensions, which are personal attributes, system requirements, pedagogical knowledge, modeling,
and feedback. Originally MEPTS included 34 items. In this study, we obtained not good but acceptable model
with 25 items. Five factors that describe effective mentoring practices in elementary science and mathematics
teaching were supported by confirmatory factor analysis. Cronbach alpha coefficients of internal consistency reliability of each dimension were found as .61 for personal attributes, .77 for system requirements, .93 for pedagogical knowledge, .80 for modeling, and .84 for feedback. Final model fit indexes were found as 2 = 1736.30, df =
265, CMIDF = 6.55, IFI = .86, CFI = .86, RMR = .046, RMSEA = .12. Relationship between preservice elementary teachers’ perceptions of their mentoring practices and self-efficacy beliefs was also investigated.