LikeLike

]]>LikeLike

]]>LikeLike

]]>LikeLike

]]>LikeLike

]]>But the grade for a course is an average of many individual grades. Assuming that the individual grades are a random sampling of the individual student’s “ability score”, then the average should approach a normal distribution by the Central Limit Theorem. If it does not approach a normal distribution, then we might reasonably conclude that the individual grades were not an accurate measurement, which might indicate a failing of the experimental design. Since we can’t re-run the experiments, it makes sense to try and account for the shortcoming of our design in the subsequent analysis. Whether or not imposing a normal distribution is the correct accounting for such shortcoming is debatable, but it is not totally counter-intuitive.

LikeLike

]]>LikeLike

]]>Thanks a lot for sharing these essential insights!

I have a question regarding the stock return plot. How would you modify the code for a list of ticker symbols and the x-axis?

In my case, I want to plot the 3-year return after the IPO. So how would you deal with the timeline to show a meaningful plot?

Thanks a lot:)

Best,

Victor

LikeLike

]]>I agree with Andi Melengu’s question

There is a mistake here, it should be reviewed:

as.vector(Cl(AAPL)[sig == 1])[-1] – Cl(AAPL)[sig == -1][-table(sig)[[“1”]]]

LikeLike

]]>LikeLike

]]>