This study examines how visitor satisfaction co-varies with augmented reality (AR)-related evaluations in a Chinese museum context, using Suzhou Museum as a case study. We collected on-site survey data from 420 visitors. Because several questionnaire blocks were path-dependent, we conducted exploratory factor analysis and confirmatory factor analysis using the complete 12-item AR-evaluation block (Q24; n = 169), whereas regression analyses included cases with complete Q24 and six-item satisfaction block responses (Q28; n = 135). The Q24 items supported a two-dimensional structure consisting of positive AR experience and technical concerns. In the conventional ordinary least squares model, positive AR experience showed the dominant positive association with core satisfaction (standardized β = 0.747, p < 0.001), whereas technical concerns showed a smaller positive coefficient under raw scoring (standardized β = 0.121, p = 0.032). However, HC3-robust standard errors and response-pattern sensitivity analyses attenuated the technical concerns estimate, so that it was no longer statistically distinguishable from zero. The study is cross-sectional and supports associations rather than causal effects. The main defensible conclusion is therefore that visitor satisfaction co-varies primarily with a coherent positive AR experience dimension, whereas the concern-related dimension remains measurable but is subject to greater inferential uncertainty in this sample.