Character assessment reimagined: an IRT-based graded response model for honesty and discipline

Abstract

Purpose: to develop and validate an IRT-based graded response model (GRM) instrument for assessing honesty and discipline among junior secondary school students.

Method: This study employed a research and development (R&D) approach involving 226 junior secondary school students. A 30-item likert instrument was evaluated using Aiken’s V, confirmatory factor analysis (CFA), and the graded response model (GRM) to examine its validity, item characteristics, model fit, and measurement precision.

Findings: the developed instrument demonstrated strong content and construct validity, high internal consistency, and good item functioning. The GRM analysis indicated that the items had strong discrimination and satisfactory model fit, while the Test Information Function showed adequate measurement precision, particularly across the lower-to-middle levels of honesty and discipline. Overall, the instrument is suitable for diagnostic and early assessment of students’ honesty and discipline.

Implications: the study provides a psychometrically sound and evidence-based instrument for assessing students’ honesty and discipline, supporting more objective character assessment and data-informed character education practices in junior secondary schools.

Originality: lies in the integration of character-based indicators of honesty and discipline with an Item response theory–graded response model (IRT-GRM) to develop a more precise and objective affective assessment instrument for junior secondary school students.