Validity and reliability analysis of the Turkish life satisfaction scale developed through artificial intelligence

ATİK S., ERDEMİR N.

BMC Psychology, vol.13, no.1, 2025 (SSCI, Scopus) identifier identifier

  • Publication Type: Article / Article
  • Volume: 13 Issue: 1
  • Publication Date: 2025
  • Doi Number: 10.1186/s40359-025-03246-2
  • Journal Name: BMC Psychology
  • Journal Indexes: Social Sciences Citation Index (SSCI), Scopus, IBZ Online, EMBASE, MEDLINE, Psycinfo, Directory of Open Access Journals
  • Keywords: Artificial intelligence, Life satisfaction, Reliability, Scale development, University students, Validity
  • Inonu University Affiliated: Yes

Abstract

This study evaluates the validity and reliability of a Turkish Life Satisfaction Scale developed using artificial intelligence (ChatGPT) to explore AI’s potential in creating psychometric tools. The scale was tested on three independent samples of Turkish university students: 503 for Exploratory Factor Analysis (EFA), 301 for Confirmatory Factor Analysis (CFA), and 79 for test–retest reliability. EFA identified a unidimensional structure, accounting for 67.50% of the total variance (factor loadings.75–.89). CFA confirmed adequate model-data fit (e.g., χ2/sd = 2.63, RMSEA = 0.07). The scale demonstrated high internal consistency (Cronbach’s α =.88) and temporal stability (test–retest correlation =.95). Criterion validity was supported by strong positive correlations with established Life Satisfaction (r =.74) and General Well-Being (r =.63) scales. These findings indicate that AI can expedite scale development while yielding robust psychometric instruments. This research underscores the innovative potential of AI-supported psychometric tools in the social sciences and offers valuable insights for future scale development.