Validating aspects of a reading test


Tezin Türü: Doktora

Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Eğitim Fakültesi, Yabancı Diller Eğitimi Bölümü, Türkiye

Tezin Onay Tarihi: 2018

Tezin Dili: İngilizce

Öğrenci: ZEYNEP AKŞİT

Danışman: Çiler Hatipoğlu

Özet:

This study investigated three aspects of validity (i.e., context, cognitive and scoring) of a reading test: First, the reading test construct was defined based on the cognitive processing model and its criterial features were presented in the test specifications. Secondly, to establish cognitive validity, the cognitive processes that were activated during test taking were investigated through retrospective and introspective verbal data. The results revealed whether the test elicited behavior reflective of reading activities beyond the testing situation. Finally, the scoring validity of the test was examined through item analysis. The findings revealed that the performances were generalizable: the tasks elicited behavior similar to those in real-life reading, and that the criterial parameters of the test were at an acceptable level for the intended population. Moreover, the majority of the item parameters were within the expected ranges. This study and its findings have important implications. At the organizational level, the results have implications for improvement in testing practice: the sociocognitive framework provides a systematic approach and encourages the generation of evidence for validity at all stages of test development. Moreover, test scores that reliably reflect test takers’ ability on relevant aspects of reading help improve reading instruction. Instruction that is grounded in theory and supported with established needs can help students be equipped with skills required for successful academic reading. Implications at the theoretical level are: The cognitive processing reading model, which was originally based on L1 reading, successfully defines L2 academic reading within the context of this study