Generate two samples of sizes m = 30 and m = 40 from U(O.l). Square the observations in the second sample. What is the theoretical distribution of the squared uniforms? Next, "forget" that you squared...

Generate two samples of sizes m = 30 and m = 40 from U(O.l). Square the observations in the second sample. What is the theoretical distribution of the squared uniforms? Next, "forget" that you squared the second sample and test by Smirnov test equality of the distributions. Repeat this testing procedure (with new samples, of course) 1000 times. What proportion of p-values exceeded 5%?