The Construct Validity of Writing Skill’s Scoring Rubric in the Persian Proficiency Test of Ferdowsi University of Mashhad

Language assessment is an essential part of each Language teaching syllabus.The effectiveness of language educational systems highly depends on the validity of language assessment methods. This research tries to study the construct validity of the writing scoring rubric in the Ferdowsi University’s Persian Proficiency test. For this purpose the results from one of the official tests performed in the Ferdowsi international center for teaching Persian to non-Persian speakers was analyzed with Rasch and factor analysis models. The results showed that the three constructs which are topic development, quality of Language and organization, all have high factor loads. While organization has the highest loads with 0.98 this amount for the two other constructs are 0.97. In order to score each construct, the writing rubric in this test has a six grades’ scale. The Rasch models showed that all raters used this scale in an appropriate way because the order of the thresholds are well organized and match with order of the scores. Also the Write map indicates that this scale can differentiate well elementary, intermediate and advanced examinees and the grades cover all students with different levels of abilities. The rater reliability for this test is 0.98 which is highly acceptable.