On the Equivalence of Constructed-Response an​d Multiple-Choice: Stem-Equivalent, Stem non-Equivalent but Content Equivalent, an​d Stem an​d Content non-equivalent Items in Reading Comprehension Using Multifaceted Rasch Model

دانشگاه ولي عصر

The present study investigated construct equivalence of multiple choice (MC) an​d constructed response (CR) item types across stem an​d content equivalent MC an​d CR items (item type ‘a’), non-stem-equivalent but content equivalent MC an​d CR items (item type ‘b’), an​d non-stem an​d non-content equivalent MC an​d CR items (item type ‘c’). One hundred seventy English-major undergraduates chosen from three Iranian universities completed MC an​d CR forms of a reading comprehension test. The effect of different item formats on test takers’ performance was measured using Multifaceted Rasch Measurement. The results for MC items showed that item type ‘a’ was the easiest followed by equally difficult item types ‘b’ an​d ‘c’. Regarding CR items, item types ‘a’ an​d ‘c’ were equally easier than item type ‘b’. Finally, the results of Facets analysis an​d t-test showed that MC an​d CR item types ‘a’ an​d ‘b’ measured similar constructs while different constructs were measured by MC an​d CR item type ‘c’. The conclusion was that cognitive demands of MC items were higher than CR items. The findings of this study showed important pedagogical implications for test constructors as to the fact that the testing method can make a difference in tests of reading comprehension. Care should be taken in constructing MC an​d CR non-stem an​d non-content equivalent items (item type ‘c’), otherwise it might lead to construct under-representation if a test consists only of MC o​r only of CR item type ‘c’ an​d this is considered as a threat to construct validity of the test.

49040

دانشگاه وليعصر رفسنجان

اكبر عباسي

آموزش زبان انگليسي

حمدالله راوند