R Code for “Student Perception Data as Low-Stakes Evidence for Classroom Inquiry: A Multilevel Validity Study of Perceived Instructional Quality in Primary Classrooms”

This dataset accompanies the manuscript entitled “Student Perception Data as Low-Stakes Evidence for Classroom Inquiry: A Multilevel Validity Study of Perceived Instructional Quality in Primary Classrooms.” The study examines whether student perception data can be interpreted as low-stakes classroom assessment evidence for classroom inquiry in fourth-grade public primary classrooms in Türkiye. The data were collected from fourth-grade students nested within public primary school classrooms.

The measures include student-perceived instructional quality, represented by cognitive activation, classroom monitoring, and emotional support, and student motivational beliefs, represented by self-efficacy and intrinsic motivation. The dataset was anonymised before sharing, and no directly identifying student or teacher information is included. The accompanying R script provides the analysis workflow used in the manuscript, including data preparation, scale scoring, confirmatory factor analysis, ordinal sensitivity checks, reliability estimation, intraclass correlation coefficients, design effects, within-class agreement evidence, gender measurement invariance, within- and between-classroom validity-network correlations, and robustness checks.