I'm reaching out with a question that, at first glance, might seem elementary, but given some perplexing advice from my thesis supervisor, I find myself in need of your expertise.
The scenario is straightforward: we have two groups of shooters, each with 100 participants. Every shooter takes 10 shots, and each shot is either a success or a failure.
I aim to determine if there's a significant difference in shooting accuracy between the two groups. My initial approach was to calculate the individual success rates and then compare these rates using a t-test. However, my mentor has indicated that this method is not suitable for analyzing proportions. On the other hand, if I use a proportion test (like prop.test or Fisher's exact test), I must consider each shot as an independent event, which doesn't seem correct since each shooter takes multiple shots.
I'm currently feeling uncertain and wonder if I might have overlooked something.
Thank you for your time and input.