University of Twente Student Theses

Login

Using Task Features to Predict Item Difficulty and Item Discrimination in 3F Dutch Reading Comprehension Exams

Groot, Nancy de (2023) Using Task Features to Predict Item Difficulty and Item Discrimination in 3F Dutch Reading Comprehension Exams.

[img] PDF
1MB
Abstract:Reading comprehension is an important target skill in education. However, there has been criticism on the assessment of reading comprehension, mainly on the construct validity. This study attempts to consider some of the criticism, aiming to identify task features that either support or impede construct relevance, reflected by the psychometric quality indicators item difficulty and item discrimination. The present study focused on 3F Dutch language comprehension exam items (N = 182), aimed at the highest reference level for vocational education, created by Cito and administered between 2015 to 2022. The focus was on finding (groups of) task features with predictive value for the item parameters from previous research, followed by analyses on the relation between task features and item parameters in the present item sample, with special attention on the lexical richness of the text, represented by the proportional complexity, based on concept maps. Key findings include text and text content as predictors for item parameters, and the degree to which the key is the best possible response (increased discriminative ability, and easier item). Regarding item difficulty, the number of plausible response options increased difficulty. Other predictors were found in the group of intrinsic task features and access skills.
Item Type:Essay (Master)
Clients:
Cito, Arnhem, Netherlands
Faculty:BMS: Behavioural, Management and Social Sciences
Subject:81 education, teaching
Programme:Educational Science and Technology MSc (60023)
Link to this item:https://purl.utwente.nl/essays/95142
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page