Virtual simulation and practice time across increasing tooth complexity in dental wax pattern training: an exploratory secondary analysis of a crossover study

Abstract Objective To exploratorily evaluate whether virtual simulation training is associated with different practice-time trajectories during dental wax pattern fabrication compared with traditional physical model training. Methods This exploratory secondary analysis used data from a randomized crossover study. Fourteen dental students were randomly assigned to a physical-first group (n = 7) or a virtual-first group (n = 7). Only first-stage data were analyzed to minimize carryover effects. Practice time across four tooth positions with progressively increasing anatomical complexity was recorded. Linear mixed-effects models were initially specified with random intercepts and random slopes. Because of the small number of participants per group, the random-slope model showed a singular fit; therefore, random-effects variance estimates and asymptotic Wald p-values were not used for definitive inference. Fixed-effect estimates are reported as exploratory findings. Results The virtual-first group showed a smaller increase in practice time across complexity levels than the physical-first group. The estimated slope was 6.40 minutes per complexity level in the physical-first group and 2.32 minutes per level in the virtual-first group. The group-by-complexity interaction estimate was − 4.08. However, because of the small sample size and singular random-slope fit, this estimate should be interpreted as exploratory and hypothesis-generating. No reliable