# Psychometric properties of the short mood and feelings scale among Chinese adolescents using item response theory

## TL;DR

This study evaluates and improves the Short Mood and Feelings Questionnaire for measuring depression in Chinese adolescents using advanced statistical methods.

## Contribution

The study introduces a revised 9-item version of the SMFQ with validated psychometric properties for Chinese adolescents.

## Key findings

- The SMFQ-9 showed high reliability with a Cronbach’s alpha of 0.86.
- The SMFQ-9 correlated strongly with the full SMFQ (0.975) and other depression measures.
- An optimal cutoff score of 4.5 was identified for detecting depression symptoms.

## Abstract

Depression is a common mental health condition that can manifest at various stages of life, including the early stages such as childhood and adolescence. In particular, adolescence is a critical period where depression can present with numerous significant and severe symptoms, such as persistent sadness, behavioral changes, and difficulties in academic performance and social interactions. These symptoms, if left untreated, can have long-lasting effects and may recur in adulthood. Early identification and monitoring of depression are therefore essential to ensure timely intervention.

The Short Mood and Feelings Questionnaire (SMFQ) is a widely used tool for measuring depression symptoms in adolescents. This study aimed to assess the SMFQ using Item Response Theory (IRT) in adolescents and determine optimal cutoff points for a revised version.

Using IRT and the Graded Response Model (GRM), we evaluated the SMFQ in 906 Chinese adolescents (average age 15 years). Items 1, 3, 4, and 6 were removed, resulting in the SMFQ-9. Reliability and validity were assessed using Cronbach’s alpha, and Receiver Operating Characteristic (ROC) analysis was conducted to determine cutoff points.

We validated the reliability and validity of the SMFQ-9, with the structure showing a Cronbach’s alpha as high as 0.86. It achieved significant correlations with three criterion questionnaires, and the correlation between SMFQ-9 and full version SMFQ reached 0.975. ROC analysis established an optimal cutoff value of 4.5, with an AUC of 0.985.

The SMFQ-9 retains the robustness of the original SMFQ, improves efficiency, reduces respondent burden, and is a reliable tool for assessing mood in adolescents in clinical and research settings.