In empirical studies of the TIC, it has been pointed out that there is a lack of robust scales for psychometric assessment and that the target population is biased towards healthcare professionals. This study aimed to examine the reliability and validity of the ARTIC-35 and its shortened version, the ARTIC-10, and to clarify issues with their use in the target population. As a result of examining the structural validity, the five-factor model of the ARTIC-35 met the fit indices' criteria, except for certain indicators (χ²=1823.63, df=550, p<.001, CFI=.80, RMSEA=.06, SRMR=.07). In this study, exploratory factor analysis extracted a four-factor model, ARTIC-19, which demonstrated better model fit compared to the five-factor model (χ²=391.75, df=146, p<.001, CFI=.92, RMSEA=.05, SRMR=.05). Conversely, the one-factor structure of the ARTIC-10 did not meet the acceptable fit criteria (χ²=285.51, df=35, p<.001, CFI=.77, RMSEA=.10, SRMR=.07). Reliability coefficients for the ARTIC-35, ARTIC-19, and ARTIC-10 were within an acceptable range. A correlation analysis was conducted between the expectedly related variables of KAP-TIP, compassion satisfaction, compassion fatigue, burnout, the presence of self-care strategies, and intention to leave. The results tentatively suggest a general tendency toward supporting convergent validity. However, issues previously noted in prior research, including instability of the factor structure and challenges with item phrasing, were also identified in the present study. Future challenges include examining the reliability and validity of the instrument across participants from diverse professional domains and practice areas, conducting qualitative analyses of item expressions in collaboration with multiple experts, and exploring potential areas for refinement and improvement.
View full abstract