corpus-based insights into modeling a level-specific persian language proficiency test (plpt): development and factor structure of the plpt listening tasks

the factor structure of the listening section of a persian language proficiency test (plpt), developed and used for academic purposes, was examined in this study. a structural equation modeling (sem) was employed using amos (v. 18) to analyze the responses of a number of persian language learners (n=120) who participated in the first piloting phase of the test in 2014. to examine whether the listening factor corresponds to the test hypothesized structure, three models (unitary, correlated and uncorrelated) were postulated on the basis of the literature. the results from model testing suggested that the correlated model (i.e., correlated receptive skills of listening and reading) fitted the obtained data best, supporting the reporting of distinctive listening and reading factors. the results of the current pilot study provide empirical evidence for reporting valid listening scores and interpretations based on separate scores found for the plpt listening skill. implications for persian language teaching, learning, and assessment are discussed.