Multiple testing problem: Each student tosses a coin 50 times, then performs a test of the hypothesis H0: p = 0.5 versus H1: p ≠ 0.5. If some students reject H0 at the α = 0.05 level, can we conclude their coins were unfair? Or is there a better explanation?

Construct a histogram of the P-values. Is it approximately uniform?
How does the uniform distribution of the P-value explain the relationship between the significance level and the probability of a Type I error?