Development and psychometric validation of the affective evaluation towards pubertal changes scale for late childhood (9-12 years old) Indonesian children.

INTRODUCTION

Puberty represents a major developmental transition involving several aspects. In Indonesia, limited open discussion about puberty, and the widespread use of non- standardized instruments have constrained efforts to assess children's psychological readiness for pubertal changes. This study aimed to develop and validate a developmentally appropriate, culturally relevant Affective Evaluation toward Pubertal changes Scale for boys and girls aged 9-12 years.

METHODS

A quantitative instrument development design was employed, including item generation, expert-based content validation, pilot testing, Exploratory Factor Analysis (EFA) using a polychoric matrix with unweighted least squares extraction and oblique rotation, and that the number of factors determined by parallel analysis. Confirmatory Factor Analysis (CFA) was conducted and the WLSMV estimator used for ordinal data. A bifactor model also tested to verify the presence of a general attitude factor. Calculating omega hierarchical and explained common variance. Discriminant validity assessed via HTMT and invariance between boys and girls tested through multigroup CFA. Data were collected from 514 elementary school children and randomly divided into independent samples for EFA (n = 264) and CFA (n = 250).

RESULTS

This study evaluated the factorial structure, construct validity, reliability, and gender- based measurement invariance of a newly developed instrument. EFA supported a three-factor structure. CFA demonstrated good model fit across multiple indices. Initial assessment indicated marginal convergent validity for one factor; however, model respecification improved Average Variance Extracted values. Discriminant validity was generally acceptable. Bifactor indices suggested a predominantly unidimensional structure with meaningful specific factors. Reliability estimates indicated high internal consistency for each factor and the overall scale. Multi-Group CFA employed across gender, although a relatively high HTMT value between two factors was observed in the male subgroup.

DISCUSSION: The findings provide strong preliminary support for the factorial validity, reliability, and cross-gender applicability of the instrument. The three-factor structure was empirically supported through both exploratory and confirmatory analyses, and dimensionality assessment indicated the presence of a dominant general factor alongside distinct subdimensions. Measurement invariance across gender suggests that the instrument operates equivalently for male and female respondents. However, the elevated discriminant validity estimate between two factors in the male group suggests further investigation.