Abstract
Background: Mental imagery influences sensory expectations, craving, liking, and eating behaviour, but existing measures are either modality-specific or use prompts not tailored to food. We developed and validated the Food Imagery Scale (FIS), a self-report measure of trait imagery vividnessacross orthonasal smell, flavour, taste, texture, and appearance.
Methods: A preliminary item-generation study identified foods and drinks with distinct sensory characteristics. Experiment 1 (N = 354) used exploratory factor analysis and parallel analysis to refine the item pool and examine internal consistency. Experiment 2 (N = 83) assessed test–retest reliability over three to four weeks. Experiment 3 (N = 393) used confirmatory factor analysis and correlations with established visual and olfactory imagery questionnaires to assess factor structure and convergent validity. Experiment 4 (N = 781) examined criterion-related associations with trait food craving.
Results: Experiment 1 supported a correlated five-factor solution for the 25-item scale (RMSR = 0.023, RMSEA = 0.065, CFI = 0.962), with high internal consistency. The total score showed good test–retest reliability (ICC = 0.75, 95% CI [0.62, 0.82]). Experiment 3 replicated the five-factor structure (χ²(265) = 503.15, CFI = TLI = 0.998, RMSEA = 0.048, SRMR = 0.040); appearance correlated most strongly with visual imagery vividness (ρ = 0.47), and smell correlated most strongly with olfactory imagery vividness (ρ = 0.51), both p < 0.001. In Experiment 4, smell and flavour imagery showed small positive associations with trait food craving after covariate adjustment (r = 0.110, p = 0.002; r = 0.080, p = 0.026, respectively).
Conclusions: The FIS is a reliable, food-specific measure of self-reported trait imagery vividness across five related sensory dimensions. Further work should establish its incremental predictive validity beyond general imagery measures and its cross-cultural generalisability.