Construct validity and expert benchmarking of the haptic virtual reality dental simulator

The aim of this study was to demonstrate construct validation of the haptic virtual reality (VR) dental simulator and to define expert benchmarking criteria for skills assessment. Thirty-four self-selected participants (fourteen novices, fourteen intermediates, and six experts in endodontics) at one dental school performed ten repetitions of three mode tasks of endodontic cavity preparation: easy (mandibular premolar with one canal), medium (maxillary premolar with two canals), and hard (mandibular molar with three canals). The virtual instrument's path length was registered by the simulator. The outcomes were assessed by an expert. The error scores in easy and medium modes accurately distinguished the experts from novices and intermediates at the onset of training, when there was a significant difference between groups (ANOVA, p<0.05). The trend was consistent until trial 5. From trial 6 on, the three groups achieved similar scores. No significant difference was found between groups at the end of training. Error score analysis was not able to distinguish any group at the hard level of training. Instrument path length showed a difference in performance according to groups at the onset of training (ANOVA, p<0.05). This study established construct validity for the haptic VR dental simulator by demonstrating its discriminant capabilities between that of experts and non-experts. The experts' error scores and path length were used to define benchmarking criteria for optimal performance.