1. In the last question, I asked if p-hat = 0.5 only when the p-hat is not given by the question. I’m still confused as in E311-(e), determining sample size n from ME and C, the 풀이 uses 0.237 (given by question) not 0.5. Can you clarify more on this?
============================
if it is given, use it! if not, use 0.5! :)
============================
2. there’s no specific mention on the standard deviation of continuous random variable unlike discrete RV.
Is it not impt for the AP?
============================
it is so important that the entire part IV is needed to explain that, among other things :)
but usually it is given in questions, so no worries.
============================
3. wuts the diff btw yi and y-hat?
residual usually is y - yhat.
but why is the std dev of residual sigma of yi - y?
============================
yi is the ith element, (observed) y hat is the predicted value
standard deviation of residuals is:
so if you saw y instead of yhat somewhere it must be a typo :)
============================
4. in the formula sheet, attached below, whats the 1st, 3rd, 4th, and the last one??
============================
cannot find attached sheet
============================
5. If a distribution of test scores is not symmetric, is it impossible to estimate the z-score of third quartile? What exactly is the "z-score of the third quartile"? I though third quartile only emerges in the boxplot section.
Also, is z-score based only on the normality? is that the reason for the question above?
============================
any distribution has medians and quartiles
finding z-score for nonnormal distributions is pointless; though possible, certainly.
but it could be used as a rough estimate in slightly nonnormal distributions.
find the third quartile point, subtract mean and divide by st. dev as you would any z score.
============================
