A second researcher wishes to determine if Hendrix students eat more M&Ms on average than all college students. This person establishes a significance level of α = 0.05 and creates a sample of size...

A second researcher wishes to determine if Hendrix students eat more M&Ms on average than all college students. This person establishes a significance level of α = 0.05 and creates a sample of size 28. They determine a p value of 0.08, which is ever so slightly above α. Disappointed, they then choose a new sample, of size 25, and determine p = 0.23. They do this five more times (for a total of 7 samples), with various sizes between 25 and 30. Their p values are 0.15, 0.32, 0.09, 0.16, and 0.02. Their final sample gave a p value smaller than the significance level and since this should happen by chance only once in 20 samples, they reason that they have found a real difference, since it only took them 7 tries, and therefore claim that Hendrix students do in fact eat more M&Ms than the average college student does. What can you say about their result and methodology?