Hacker News Reader

From the article:

> In this study, we tested 77 undergraduates who were randomly assigned to play either a popular video game (Portal 2) or a popular brain training game (Lumosity) for 8 h.

Sorry my comment should have been clearer. I wasn't actually asking for the sample size - i was making a statement that the sample size was ridiculously small as to be irrelevant. It was more like an eye roll. Thank you for responding to my literal comment snd reminding me to be clearer in my messages :)

77 is a large enough sample size to produce statistically significant results in a lot of cases, depending on the problem. For example, using 77 men and women, you could easily see a statistically significant difference in height between genders.

This paper is pretty simple, it's just a straight-forward before-and-after test. They are holding a lot of factors constant and not doing an observational study, so 77 is actually a fine sample size for something like this.

Literally the first line of the abstract.

There's a power analysis in the paper showing that they've got plenty of statistical power. (Estimated N=34 subjects for 90% power, but they used 77 subjects instead).

Central Limit Theorem