Latent Profiles of AI Risk Perception and Their Differential Association with Community Driving Safety Concerns: A Person-Centered Analysis

Abstract

Public attitudes toward artificial intelligence (AI) and driving safety are typically studied in isolation using variable-centered methods that assume population homogeneity, yet risk perception theory predicts that these evaluations covary within individuals as expressions of underlying worldviews. This study identifies latent profiles of AI risk perception among U.S. adults and tests whether these profiles are differentially associated with community driving safety concerns. Latent class analysis was applied to nine AI risk-perception indicators from a nationally representative survey (Pew Research Center American Trends Panel Wave 152, n = 5,255); Bolck-Croon-Hagenaars corrected distal outcome analysis tested class differences on nine driving-safety outcomes, and survey-weighted multinomial logistic regression identified demographic and ideological predictors of class membership. Four classes emerged: Moderate Skeptics (17.5%), Concerned Pragmatists (42.8%), AI Ambivalent (10.6%), and Extreme Alarm (29.1%), with all nine driving-safety outcomes significantly differentiated across classes. Higher AI concern mapped monotonically onto greater perceived driving-hazard severity; the exception, comparative evaluation of AI versus human driving, was driven by trust rather than concern level. The cross-domain covariation provides person-level evidence for the worldview-based risk structuring posited by Cultural Theory of Risk and yields a four-class segmentation framework for AV communication that links AI risk orientation to transportation safety attitudes.

Figures (5)

• Figure 1: Analytic pipeline overview. Six stages proceed from data preparation and variable operationalization through weighted LCA enumeration and model selection (yellow, RQ 1), BCH-corrected distal outcome analysis (orange, RQ 2), and survey-weighted multinomial logistic regression (red, RQ 3), concluding with four robustness checks. The dashed sidebar on the left maps each research question to its corresponding analytic stage.
• Figure 2: Model fit indices for latent class solutions ($K = 1$--$7$). (A) BIC, SABIC, and AIC with the shaded region highlighting the $K = 4$ elbow. (B) Classification entropy with the 0.80 adequacy threshold. (C) Log-likelihood trajectory. (D) Free parameters per model with BLRT and VLMR-LRT significance annotations. The red bar marks the recommended $K = 4$ solution.
• Figure 3: Item-response probability profiles for the four-class solution ($K = 4$). Each panel displays one of the nine LCA indicator items; within each panel, the four classes are shown with response-category probabilities. Class labels and weighted prevalences: C1 = Moderate Skeptics (17.5%), C2 = Concerned Pragmatists (42.8%), C3 = AI Ambivalent (10.6%), C4 = Extreme Alarm (29.1%).
• Figure 4: BCH-corrected class-specific means for nine driving-safety distal outcomes. Error bars denote sandwich-estimated standard errors. For driving-problem items (1--3 scale), lower bars indicate greater perceived severity. For road rage (1--5) and safety trend (1--6), higher bars indicate more pessimistic assessment and greater frequency, respectively.
• Figure 5: Forest plot of odds ratios from the survey-weighted multinomial logistic regression, with Concerned Pragmatists (C2) as the reference class. Points indicate odds ratios; horizontal lines denote 95% Wald confidence intervals. The dashed vertical line at OR $= 1.0$ marks no effect. Filled points indicate statistically significant coefficients ($p < 0.05$).