Comparison of Internet-based and paper-based questionnaires in Taiwan using multisample invariance approach

Area of Science:

• Psychometrics
• Mental Health Assessment
• Digital Health

Background:

• Assessing depression often relies on questionnaires like the Center for Epidemiologic Studies Depression Scale (CES-D).
• The shift towards digital health necessitates evaluating the psychometric properties of Internet-based assessments compared to traditional paper-based methods.

Purpose of the Study:

• To determine the psychometric equivalence between an Internet-based and a paper-based version of the Chinese CES-D.
• To compare the factor structure, reliability, and validity of both questionnaire formats.

Main Methods:

• A randomized controlled trial involving 2,400 teachers in Taiwan.
• Application of the multisample invariance approach derived from structural equation modeling (SEM).
• Analysis included testing for equivalent factor structures and latent mean differences across groups.

Main Results:

• The Internet-based and paper-based CES-D demonstrated equivalent factor structures and item functioning.
• Latent means for "depressed mood," "positive affect," and "interpersonal problems" were not significantly different.
• A statistically significant difference was found in "somatic symptoms" latent means, but with a negligible effect size.

Conclusions:

• The Internet-based CES-D is a psychometrically sound and equivalent alternative to the paper-based version.
• Digital administration of the CES-D maintains its validity, reliability, and factor structure.
• The Internet-based questionnaire presents a promising and practical option for depression screening and research.