Dataset for: Augmenting Self-Reports: Using Eye-Tracking and Questionnaire Data to elucidate the Role of Rating Valence in AI Skin Cancer App Acceptance

Background: AI-based skin cancer screening apps (AISCSAs) offer diagnostic potential but face limited adoption. App-store cues such as ratings may influence acceptance, yet little is known about how users cognitively process app-store information in high-stakes health contexts. To address this gap, eye-tracking was used to measure visual attention while participants evaluated an AISCSA app-store listing.

Objective: This study aimed to test whether one negative rating capture visual attention and whether an extended Technology Acceptance Model (TAM) was able to predict behavioral intention to use an AISCSAs.

Method: Participants (N= 76) evaluated a mock app-store listing for an AISCSA under positive (n=42) or negative (n=34) rating conditions while their eye movements were recorded. Analyses combined fixation durations in defined Areas of Interest (AOIs) with self-reported measures of perceived usefulness (PU), perceived ease of use (PEOU), trust, behavioral intention (BI), willingness to pay (WTP), and the self-rated importance of app attributes.

Results: Normalized fixation durations (s/px²) revealed the highest attention to description (0.166 s/px²), followed by reviews (0.11 s/px²) and ratings (0.04 s/px²), with price and data protection receiving the least attention. Of the five self-rated importance of app attributes only reviews were positively correlated with fixation duration on review-AOI (r = .28, p = .014). Rating valence had no significant effect on gaze patterns, PU, PEOU, trust, BI, or WTP (all ps > .05). However, PEOU (p < .01), PU (p < .001), and trust (p < .001) correlated significantly with BI.  

Conclusion: Although the expected attentional capture effect of negative ratings was not observed, the weak or non-existent associations between fixation duration on AOI and self-rated importance of app attributes suggest that eye-tracking captures aspects of information processing that are not directly reflected in self-reported evaluations. These findings indicate that eye-tracking provides a more direct approximation of actual user behavior by revealing implicit attentional processes beyond self-reported evaluations. TAM constructs and trust predicted BI, but rating valence alone did not affect acceptance or gaze behavior. In high-stakes health contexts, textual information may outweigh rating valence in driving adoption. Future research should explore conditions under which rating valence matters, including more extreme rating contrasts and variations in accompanying review texts, as well as the influence of individual differences such as pre-existing attitudes toward AI and levels of AI literacy.