An application of cognitive diagnostic models in the evaluation of reading comprehension in Chilean university students

Abstract

Reading Comprehension skills are fundamental to the academic performance of university students. These abilities are not innate; therefore, systematic training is required before entering university. The main objective of this study was to evaluate the reading comprehension skills of Chilean University Students. Method: A sample of 579 students participated in the study, responding to four texts that comprise the Reading Comprehension Strategies for University Students (ELCU). An automated platform was used to apply and correct the test with immediate results and feedback. The tool allowed the use of a broad and representative sample of the various regions of Chile. The ELCU was developed using a cognitive diagnostic assessment (CDA) considering cognitive, metacognitive and linguistic strategies. The analyses were performed using two different approaches: The Linear Logistic Test Model (LLTM) and the deterministic inputs, noisy “and” gate (DINA) model. Results: The rejection of the LLTM implies that the strategies do not differ in cognitive complexity (difficulty) while the adequate goodness of fit to the DINA model allows to characterise the students according to their reading abilities. Conclusion: The ELCU test provides reliable and valid information on the reading comprehension ability and metacognitive awareness of university students.