BMC Public Health (Aug 2024)

Health behavior inventory – short form: a Chinese validation study using classical test theory and item response theory

  • Lai Kun Tong,
  • Mio Leng Au,
  • Mu Rui Zheng,
  • Yong Bing Liu,
  • Guang Lei Fu,
  • Yue Yi Li

DOI
https://doi.org/10.1186/s12889-024-19774-2
Journal volume & issue
Vol. 24, no. 1
pp. 1 – 10

Abstract

Read online

Abstract Background It is imperative to use a validated tool to measure and understand health behavior as it greatly impacts health status. It has been demonstrated that the Health Behavior Inventory - Short Form (HBI-SF) is valid in several countries. The purpose of this study was to translate the HBI-SF into Chinese and test its psychometric properties. Methods This study employed a two-phase methodology. The phase one entailed the cross-cultural adaptation of the HBI-SF, while the subsequent phase evaluated the psychometric properties of the scale. This evaluation encompassed classical test theory (CTT) and item response theory (IRT) tests to assess reliability and validity. It involved 1058 participants, of whom 1036 completed the questionnaire successfully. Out of these, 518 were analyzed for CTT, along with 518 for IRT. Results The S-CVI was found to be 0.935, while the I‐CVI ranged from 0.889 to 1.000. The results of the confirmatory factor analysis suggested the goodness-of-fit indices for the four-factor model was acceptable. Regarding the subscales, the average variance extracted (AVE) and Heterotrait-Monotrait Ratio of Correlations (HTMT) matrix also met the cutoff values (AVE > 0.5 and HTMI < 0.85). Furthermore, the internal consistency and composite reliability indices of all factors were greater than 0.7. Infit and/or outfit values indicated that all items fitted the Rasch model. The Wright map revealed that the average person measures for the participants (mean = − 0.994, SD = 0.430) were comparatively lower than the average item measures (mean = 0.000, SD = 0.257). The person separation reliability values for the four factors ranged from 0.600 to 0.746, indicating an acceptable level of reliability. Two items showed differential item functioning. Conclusions The findings derived from both CTT and IRT analyses demonstrate favorable levels of reliability and validity for the Chinese version of the HBI-SF.

Keywords