| Summary: | Standardized reading comprehension tests (RCTs) are indispensable instruments in language abilities assessment. Nevertheless, we do not have such an instrument in Serbian. In the present paper, we attempted to verify the basic assumptions of the RCT conception we proposed: the reading material consisting of a relatively large selection of short newspaper articles followed by a single multiplechoice question each, displayed in temporal succession, not simultaneously. We argued that RCT consisting of such item form and content, as well as the administration procedure, should be of equal or even better construct validity than well-known RTCs in foreign languages, also possessing some other practical advantages. These assumptions were implemented in a computer-ran experimental reading comprehension task (ERCT), in a student sample. ERCT psychometric properties analyses showed good discriminability and somewhat less than preferable reliability. The structure and strength of ERCT x other intelligence tests correlations suggest ERCT belonging to the core of verbal abilities tests. Different strategies subjects employed showed no impact on major performance measure number correct answers. Our findings obtained by ERCT implementation support the RCT conception we propose. We therefore suggest possible further research towards the construction of standardized RCT in Serbian.
|
|---|
