Conclusion: The current data imply that the performance index as defined herein is a valid measure of the performance
of a trainee using the virtual reality phacoemulsification simulator. Further, the performance index increase linearly
with measurement cycles for less than five measurement cycles. To fully use the learning potential of the simulator
more than four measurement cycles are required. Materials and methods: Altogether, 10 trainees were introduced to the
simulator by an instructor and then performed a training program including four measurement cycles with three iterated
measurements of the simulation at the end of each cycle. The simulation characteristics was standardized and defined in
14 parameters. The simulation was measured separately for the sculpting phase in 21 variables, and for the evacuation
phase in 22 variables. A performance index based on all measured variables was estimated for the sculpting phase and
the evacuation phase, respectively, for each measurement and the three measurements for each cycle were averaged.
Finally, the performance as a function of measurement cycle was estimated for each trainee with regression, assuming a
straight line. The estimated intercept and inclination coefficients, respectively, were finally averaged for all trainees.
Results: The performance increased linearly with the number of measurement cycles both for the sculpting and for the