Cognitive Structure of Text Comprehension: A Psychometric Study of PIRLS-Format Items Among Primary School Students in Uzbekistan

Background. The cognitive-process model used in the Progress in International Reading Literacy Study (PIRLS) has rarely been tested in multilingual, non-Anglophone education systems. This study examined how the four PIRLS reading processes relate to one another among fourth-grade students in Tashkent, Uzbekistan, whether students perform differently across these processes, and whether these differences vary by language of instruction. Methods. In a cross-sectional design, students completed two literary texts on a digital assessment platform; after exclusion of incomplete protocols, 1,835 entered the analysis of the first text and 2,231 the second. Each text carried 14 dichotomously and polytomously scored PIRLS-format items covering retrieval, straightforward inference, interpretation and evaluation. Analyses comprised descriptive and reliability statistics, exploratory factor analysis, repeated measures analysis of variance on a common proportion-of-maximum metric, and comparisons between Uzbek-medium and Russian-medium students. Results. Internal consistency was acceptable for the full 14-item scale of each text (α = .733 and .740) but low for the four process subscales (α = .298–.597), so process-level findings are provisional. Three factors emerged in each text, with a basic component combining retrieval and inference items, but the remaining factors differed between texts. Cognitive process exerted a large within-subject effect (η² p = .678 and .758): performance fell significantly at every step from retrieval to evaluation in the first text, whereas in the second retrieval and inference did not differ and evaluation exceeded interpretation (d z = 0.937), so the hierarchy is not monotone with cognitive demand. Russian-medium students outperformed Uzbek-medium students on every measure, most markedly for interpretation in the first text (d = 1.144); a significant Language × Process interaction (η² p = .129 and .078) showed the gap to be smallest for basic processes and largest for higher-order ones. Conclusions. The findings suggest a broad distinction between basic and higher-order comprehension rather than a consistent four-step hierarchy. The variation in process structure across texts and the larger language-group differences for higher-order processes highlight the importance of examining cognitive-process measures within specific linguistic contexts. Because process-level reliability was limited and language-group comparisons were based on observed scores without measurement-invariance testing, the findings should be interpreted as evidence of differences in assessment performance rather than established differences in underlying reading ability.

Authors

Publication Details

Journal
Research Square
Published
2026-11-19
DOI
https://doi.org/10.21203/rs.3.rs-11565185/v1
Primary Topic
Reading and Literacy Development
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Cognitive Structure of Text Comprehension: A Psychometric Study of PIRLS-Format Items Among Primary School Students in Uzbekistan

Zubayda Murotalieva, Abduvali Ismailov, Nilufar Mambetova, Kunduzkhon Ruzieva et al.
Research Square
Reading and Literacy Development
preprint

Cognitive Structure of Text Comprehension: A Psychometric Study of PIRLS-Format Items Among Primary School Students in Uzbekistan

Zubayda Murotalieva, Abduvali Ismailov, Nilufar Mambetova, Kunduzkhon Ruzieva, Sardor Palvannazirov
preprint en

Abstract

Background. The cognitive-process model used in the Progress in International Reading Literacy Study (PIRLS) has rarely been tested in multilingual, non-Anglophone education systems. This study examined how the four PIRLS reading processes relate to one another among fourth-grade students in Tashkent, Uzbekistan, whether students perform differently across these processes, and whether these differences vary by language of instruction. Methods. In a cross-sectional design, students completed two literary texts on a digital assessment platform; after exclusion of incomplete protocols, 1,835 entered the analysis of the first text and 2,231 the second. Each text carried 14 dichotomously and polytomously scored PIRLS-format items covering retrieval, straightforward inference, interpretation and evaluation. Analyses comprised descriptive and reliability statistics, exploratory factor analysis, repeated measures analysis of variance on a common proportion-of-maximum metric, and comparisons between Uzbek-medium and Russian-medium students. Results. Internal consistency was acceptable for the full 14-item scale of each text (α = .733 and .740) but low for the four process subscales (α = .298–.597), so process-level findings are provisional. Three factors emerged in each text, with a basic component combining retrieval and inference items, but the remaining factors differed between texts. Cognitive process exerted a large within-subject effect (η² p = .678 and .758): performance fell significantly at every step from retrieval to evaluation in the first text, whereas in the second retrieval and inference did not differ and evaluation exceeded interpretation (d z = 0.937), so the hierarchy is not monotone with cognitive demand. Russian-medium students outperformed Uzbek-medium students on every measure, most markedly for interpretation in the first text (d = 1.144); a significant Language × Process interaction (η² p = .129 and .078) showed the gap to be smallest for basic processes and largest for higher-order ones. Conclusions. The findings suggest a broad distinction between basic and higher-order comprehension rather than a consistent four-step hierarchy. The variation in process structure across texts and the larger language-group differences for higher-order processes highlight the importance of examining cognitive-process measures within specific linguistic contexts. Because process-level reliability was limited and language-group comparisons were based on observed scores without measurement-invariance testing, the findings should be interpreted as evidence of differences in assessment performance rather than established differences in underlying reading ability.

Research Square
Reading and Literacy Development
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon