Item response theory analysis of the Chinese version compulsive shopping scale

Abstract

Compulsive shopping, a behavioral disorder with significant personal and social repercussions, necessitates reliable assessment tools, especially within different cultural contexts. While several scales exist to measure compulsive buying behavior, there is a lack of validated instruments tailored to the Chinese population. This study aimed to examine the psychometric properties of the Chinese version of the Compulsive Shopping Scale (CSS) using Item Response Theory (IRT). A total of 637 young healthy participants (42.5 % males and 57.5 % females), with a mean age of 21.32 years ( SD = 2.06), both undergraduate and postgraduate students, were recruited from various cities in China. Sixty-four participants were retested after a two-week interval to assess test-retest reliability. The results indicated that a one-dimensional factor structure was appropriate. The reliability analyses, including test-retest reliability, ω, α, and λ 6 tests, demonstrated good internal consistency. The rating scale model analysis showed infit and outfit MNSQ values between 0.6 and 1.4, indicating a good fit. The item information curve and test information curve indicated a normal distribution of scores. Differential item functioning was observed in two items, suggesting potential gender-related differences. These findings indicate that the Chinese version of the CSS is a reliable and valid instrument for measuring compulsive shopping behaviors in Chinese populations. • Faber's compulsive shopping scale's reliability, validity, and model were tested. • The classical test and item response theories were used for the analysis. • The infit MNSQ and outfit MNSQ of the scale were between 0.6 and 1.4. • Two items also had differential item functioning on the compulsive shopping scale. • The scale is effective for measuring compulsive shopping in China.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 5 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.