Turkish Adaptation of the Clean Food Consumerism Scale: A Validity and Reliability Study

Background: Clean food consumerism reflects consumers' preferences for foods perceived as natural, authentic, transparent, and health-promoting. This study aimed to culturally adapt the Clean Food Consumerism Scale (CFCS) into Turkish and evaluate its psychometric properties. Methods: The CFCS was translated, back-translated, reviewed by experts, pilot-tested, and psychometrically evaluated. For the pilot study and factor analyses, the Turkish version was administered to 1637 adults, who were allocated to three mutually exclusive subsamples: n = 180 for the pilot study, n = 500 for EFA, and n = 957 for CFA. An additional, separate sample of 124 participants completed the CFCS twice at a three-week interval for the test-retest analysis. Internal consistency, construct validity, convergent validity, discriminant validity, criterion validity, and test-retest reliability were assessed using Cronbach's alpha, item-total correlations, EFA, CFA, composite reliability (CR), average variance extracted (AVE), the Fornell-Larcker criterion, heterotrait-monotrait (HTMT) ratios, correlations with the Sustainable and Healthy Eating Behaviors Scale (SHEBS), and test-retest correlations. Results: The 18-item CFCS demonstrated excellent internal consistency (Cronbach's α = 0.948). Sampling adequacy was supported by the KMO value (0.927) and significant Bartlett's test (χ² = 7525.229, p < 0.001). EFA supported a five-factor structure comprising Perceived Health Benefits, Production Process Transparency, Authenticity, Perceived Ease of Use, and Familiarity, explaining 80.92% of the total variance. CFA supported the five-factor model. After adding one content-justified residual covariance between two closely related Perceived Health Benefits items, the final model showed acceptable fit (χ² = 641.39, df = 124, p < 0.001; CFI = 0.965, TLI = 0.956, NFI = 0.957, RMSEA = 0.066 [90% CI 0.061-0.071], SRMR = 0.047, GFI = 0.926, AGFI = 0.899). Standardized factor loadings ranged from 0.748 to 0.938, composite reliability from 0.873 to 0.940, and average variance extracted from 0.697 to 0.771. HTMT ratios for all factor pairs remained below 0.685 and the Fornell-Larcker criterion was satisfied for every factor, supporting discriminant validity. Test-retest reliability over three weeks was excellent (r = 0.889, p < 0.001), and CFCS scores were positively correlated with SHEBS scores (r = 0.441, p < 0.001). Conclusions: The Turkish version of the CFCS demonstrated strong validity, reliability, and temporal stability while preserving the original five-factor structure. It provides a psychometrically sound instrument for assessing clean food consumerism among Turkish-speaking adults and may support future research in nutrition, public health, sustainability, and consumer behavior.