Testing for equal and unequal variance: How do I work out the alpha value so that I can use the right f-crit table? I'm a stats noob and I'm following this guide on how to work out whether the variances in 2 samples I intend to use in a 2-sample t-test are equal.
In the video he refers to an f-crit table for Alpha values of 0.05. How do I figure out what Alpha value my sample data has?
 A: Alpha values are typically assigned a default value of $p=0.05$ and if yours in not listed, that is likely what it is. In fact, for some statistical software implementations, alpha cannot even be changed. What that refers to is A.K.A. type I error a type I error (alpha error) is the incorrect rejection of a true null hypothesis (a "false positive"), while a type II error (beta error) is incorrectly retaining a false null hypothesis (a "false negative"). More simply stated, a type I error is detecting an effect that is not present, while a type II error is failing to detect an effect that is present.
In effect, many users habitually find that it is acceptable to falsely accept a result 1 time in 20 in order to assign a calculate a p-value. Those p-values, in turn have the meaning that if they are less than 0.05, the results are "significant." Let's make up a beta error up, say the "answer" from the "alpha p=0.05" test is p=0.02. That means that there is a 2% probability of failing to detect significance and a 5% chance of that being a fictional result.
In some publications, e.g., the probability for finding an exoplanet, one sees 5 sigma tests. That would refer to a much lower alpha values because a lot can go wrong in exoplanet detection, and the astrophysicists want to be sure they didn't mess up and falsely identify a planet. After all, planets disappearing can be a source of embarrassment. BTW, can you find where I put Earth? Hello, Carl to Earth, are you there? My falsely accepting that Earth exists would be a type I or alpha error.
