Adaptation and validation of Richmond Compulsive Buying Scale in Chinese population

Abstract

Background and aimsCompulsive buying (CB) is a behavioral addiction that is conceptualized as an obsessive–compulsive and impulsive–control disorder. The Richmond Compulsive Buying Scale (RCBS), a six-item self-reporting instrument that has been validated worldwide, was developed based on this theoretical background. This study aimed to adapt RCBS to the Chinese population (RCBS-TC) to guide future national and international prevalence studies.MethodsThis methodological study was conducted in two phases. Phase 1 involved the forward and backward translation of RCBS, the content and face validation of the RCBS, and the evaluation of its translation adequacy. Phase 2 involved the psychometric testing of RCBS-TC for its internal consistency, stability, and construct validity using confirmatory factor analysis (CFA).ResultsIn Phase 1, RCBS-TC obtained satisfactory item-level (I-CVI = 83.3%–100%) and scale-level content validity index (CVI/AVE = 97.2%), comprehensibility (100%), and translation adequacy [intraclass correlation coefficient (ICC) = 0.858]. In Phase 2, based on data collected from 821 adults, RCBS-TC demonstrated a satisfactory internal consistency (Cronbach’s α = .88; corrected item-total correlation coefficients = 0.61–0.78) 2-week test–retest reliability (ICC = 0.82 based on 61 university students). For construct validation, the CFA results indicated that the corrected first-order two-factor models were acceptable with the same goodness-of-fit indices (χ2/df = 8.56, CFI = 0.99, NFI = 0.98, IFI = 0.99, and RMSEA = 0.09). The 2-week test–retest reliability of RCBS-TC (n = 61) was also satisfactory (ICC = 0.82).Discussion and conclusionsThis methodological study adopted appropriate and stringent procedures to ensure that the translation and validation of RCBS-TC was of quality. The results indicate that this scale has a satisfactory reliability and validity for the Chinese population.

Adaptation and validation of the health belief model scale for colorectal cancer screening

“…To the best of our knowledge, this is the first attempt to develop a culturally and contextually sensitive measure for beliefs about FOBT among Koreans in an HBM-driven context. This study used several sequential methodological processes, which are common methods in the literature, to adapt and validate the HBM scale for Koreans (Champion et al, 2005;Lam, Chan, Chong, Wong, & Ye, 2018;Lee, Lee, & Aranda, 2018). The HBM subscales for barriers and selfefficacy were modified according to Korean culture and context.…”

Adaptation and psychometric testing of the hoarding rating scale (HRS): a self-administered screening scale for epidemiological study in Chinese population

“…The content validity index (CVI), "which indicates the proportion of responses that agree with the relevancy of the scale, was computed based on the mean ratings given by experts that gave ratings of 3 or 4." ( [26], p. 761). The scale-(S-CVI) and item-level CVIs (I-CVI) were of 0.80 or above considered satisfactory [27,34].…”

“…In order to provide independent evaluation of psychometric properties and avoid psychological carry-over effects, samples used in phase 1 were excluded. The participants were newly recruited from the general public in three major districts in Hong Kong-Hong Kong Island, New Territories, and Kowloon-which included a good mix of participants with different socio-demographics [26]. A research nurse invited pedestrians to respond self-administrated questionnaires, including demographics and the CHRS by the use of paper-and-pencil method.…”

“…The internal consistency of the CHRS was computed using conventional "Cronbach's α statistics (where α ≥ 0.70 indicates a satisfactory result) and the corrected item-total correlation coefficient (where r ≥ 0.30 indicates a homogenous item)" ( [26], p.763). Stability was examined by test-retest reliability over a 2-week period.…”

“…Because of the principle of anonymity, the samples recruited from the general public did not contain any contact information, which made impossible for retest. A convenience sample of 62 undergraduate students was selected to answer the questionnaire twice (T1 as well as T2, 2 weeks later) [26]. The anonymous T1 and T2 responses collected from each student were matched using self-generated codes (i.e., combinations of mobile and student identity numbers) as described in a previous study [28].…”