International Journal of Language Testing

International Journal of Language Testing

Item Response Theory Analysis of the Progress in International Reading Literacy Study (PIRLS) 2021 in Kazakhstan

Document Type : Original Research Article

Authors
Maya Rohimah Atmawinata 1
Wiwin Herwina 2
Sri Bulan 3
Wikanengsih Wikanengsih 4
Nani Darheni 5
Gulnoza Tashtemirova 6
1 Universitas Teknologi Digital (Digitech University)
2 Universitas Siliwangi Indonesia
3 UIN Sunan Ampel Surabaya, Indonesia
4 Institut Keguruan dan Ilmu Pendidikan Siliwangi - Cimahi- Bandung- Indonesia
5 Badan Riset dan Inovasi Nasional (BRIN), Nasional Research and Innovation Agency, Indonesia
6 Department of Primary Education Pedagogy, Tashkent State Pedagogical University, Tashkent, Uzbekistan 27 Bunyodkor street, Chilonzor district, Tashkent
10.22034/ijlt.2024.456241.1343
Abstract
Progress in International Reading Literacy Study (PIRLS) is an international comparative study of reading comprehension that is administered by the International Association for the Evaluation of Educational Achievement (IEA) to measure the reading competency of fourth grade students in the participating countries. The test is a very comprehensive measure of reading literacy which covers a wide variety of reading subskills and texts. The purpose of the current study is to evaluate the psychometric qualities of PIRLS 2021 using the national data of Kazakhstan. The 2-parameter item response theory was used to evaluate the test. Item difficulty, discrimination, infit, outfit, and the root mean square deviance (RMSD) fit statistics were examined. The multiple-group IRT was used to examine differential item functioning. Findings sowed that PIRLS 2021 is a robust measure of reading comprehension and does not exhibit differential item functioning (DIF). However, based on the 2PL discrimination and infit and outfit values, several items are problematic. The implications of this finding are discussed.
Keywords
PIRLS
Reading literacy
RMSD
2PL
Multiple-group IRT