In class we used a simulation to determine when the large sample approximation was valid for testing the mean of a normal distribution by comparing the results of tests with large sample approximation to the exact t-test for 10,000 simulated random samples from a known normal distribution. The large sample approximation for testing variance of a normal distribution is based on the z-statistic Z = S − σ σ⁄√2n Conduct a simulation to determine how large the sample size must be for the approximate test to be valid by comparing the approximate test to the exact test for the following situation H: σ = 25 H: σ > 25 with a significance level of 0.05. Simulating random samples drawn from a normal distribution with µ = 10 and σ = 25, perform the simulation for sample sizes n = 3, 10, 30, 60 and 100. Of the 5 sample sizes, which is the smallest sample size you would recommend using the approximate method? Show supporting histograms