An Item Response Theory Analysis of the Teacher and Classmate Support Scale: Calibration and Cross-National Measurement Invariance in Spain and China
Year: 2026
Type: article
Abstract: The purpose was to calibrate the psychometric properties of the Teacher and Classmate Support Scale (TCMS) in Spain and China, through IRT. The sample was composed of 923 adolescents (44.05% males). The results showed two empirically independent dimensions in both samples, confirming the two-factor structure. On the other hand, the observed pattern of declining information and increasing measurement error at higher levels of the trait is compatible with a ceiling effect in the classmate support subscale. However, the instrument showed higher discrimination parameters in the middle and lower regions of the latent continuum, indicating greater conditional information in these ranges for both teacher and classmate support. Finally, the results supported a partial invariance solution across the Chinese and Spanish samples, although several items displayed DIF and cross-national comparisons should therefore be interpreted cautiously. This is the first time that the psychometric properties of TCMS is calibrated through IRT framework.
Source: Educational Assessment
Authors: Igor Esnaola(Conceptualization, Data curation, Methodology, Project administration, Writing – original draft, Writing – review & editing), Víctor B. Arias(Conceptualization, Data curation, Formal analysis, Methodology, Writing – original draft, Writing – review & editing), Iratxe Antonio-Agirre(Conceptualization, Project administration, Writing – original draft, Writing – review & editing), Lorea Azpiazu(Conceptualization, Project administration, Writing – original draft, Writing – review & editing), Yina Wang(Conceptualization, Data curation, Project administration, Writing – original draft)
Institutions: University of the Basque Country, Universidad de Salamanca
Language: English
Field: Decision Sciences
Domain: Social Sciences
SDG: Quality education
Open Access status: hybrid
OpenAlex ID: W7236143320
