Abstract / Summary
Atlas covers communicate thematic identity and shape first impressions, yet their design remains largely guided by heuristic experience rather than empirical evidence. Although eye-tracking research has examined map use, packaging design, and visual interfaces, little is known about whether atlas-cover variables that enhance perceived attractiveness also align with gaze-based visual-attention indicators. This study addresses this gap by examining how color harmony, graphic style, and layout configuration influence subjective and eye-tracking responses to atlas covers. Forty valid participants viewed eight atlas-cover stimuli generated from a 2×2×2 within-subject factorial design. Each participant completed two counterbalanced viewing cycles, resulting in 16 trials per participant. The two repeated observations for each participant-cover combination were averaged before linear mixed-effects modeling. The results showed that contrast harmony, concrete landscape imagery, and centralized layouts significantly increased perceived attractiveness. The strongest subjective effect was found for graphic style, with landscape images outperforming map symbols (MD=0.66, pHolm=0.003), followed by color harmony (MD=0.29, pHolm<0.001) and layout configuration (MD=0.14, pHolm=0.031). For eye-tracking indicators, centralized layouts produced longer total fixation duration (MD=0.34 s, pHolm<0.001) and more fixations (MD=1.13, pHolm<0.001). Color harmony showed smaller effects on visual exploration, whereas graphic style mainly affected initial attentional orienting. Mixed-effects association models showed no robust associations between perceived attractiveness and the four gaze-based indicators after Holm-Bonferroni correction (all pHolm≥0.110). These findings suggest a partial dissociation between aesthetic evaluation and gaze-based visual attention, supporting a dual-layer logic in which layout guides visual organization, while color harmony and graphic style enhance affective appeal and thematic expressiveness.