I am designing an experiment to test drone pilot reaction time to obstacles introduced to the drone's flight path in a controlled setting. There is one IV with 3 levels, and 3 DV. IV is visual acuity technique used; line of sight, first person view-LCD screen, or first person view-goggles. DVs will be total time to complete a course of flight on a specific path, pilot reaction time to an obstacle introduced to the flight path, and the composite score of the NASA task load index questionnaire completed by each participant after each flight while using one of the three levels of IV.
Each participant will fly the experiment flight course using all three levels. With there being 3 levels, there are 6 orders of application of the levels possible. I originally planned to use 6 groups of participants with each group conducting the experiment in each of the possible orders in order to counterbalance the experiment to address any learning effect of the participants.
However, my instructor has indicated that if I conduct a within-subjects design vice this between subjects design, it would reduce the number of participants needed. Sample size was determined using G*Power; 72 participants needed. Could someone explain how conducting this experiment in a repeated measures fashion would reduce the required sample size? Thanks for your help.