The math (in the steps 1-6 section) is so bad that it affects the credibility of the rest.

Edit: The way the results vs Times New Roman were blatantly cherry-picked is also disturbing. The description of p-value is probably irrelevant if you're combining the results of 4 different experiments.