Brandon the way you present and explain these concepts, you are truly gifted! Your own clarity of concepts and the effort put in to make these dense concepts clear to the viewer are very obvious. Kudos sir!
I've been using your videos to refresh my basic stats for my final dissertation defense in biostatistics. Thank you for these videos, they are fantastic!
Hi Brandon, so there are multiple ways of calculating Z-scores. If the sample is regular Z = x-xbar / standard deviation if it is a sampling distribution: Z = xbar - E(xbar) / (standard_deviation / sqrt(n)) if it is a sampling proportion: Z = pbar-E(pbar)/sqrt(p(1-p)/n) Can someone please explain when to use which one? what is the difference between sampling distribution and sampling proportion? Also how does one come with the formula for standard deviation as sqrt(p(1-p)/n)
Dear Brandon, many thanks for all your fantastic videos! One question: in minute 11.01, when calculating the standard error, the formula demands to put in a value for "p" (population proportion), but instead you fill in the value 0.567, which was previously calculated as a value for "p bar" (sample proportion). Why is that so? Do you consider that "p" and "p-bar" are the same?
i'm gonna assume they're the same for now since he only takes one sample proportion (one test to find E(p-bar)), whereas in previous videos on sample means, Mr. Brandon takes multiple samples (9 test to find E(x-bar)). add a grain of salt to that since this is only my hypothesis
Hello, I've got the same doubt after rewatching the video I've got clarity. It's because population proportion is the expected value of the sample proportions E(p(bar)) = P. And E(p(bar)) is the average of the sample proportions that we take. So here E(p(bar)) = P = avg(all(p(bar))). there is only one p(bar). so P = p(bar)
Hi Brandon, You can create your sampling distribution with only one sample because the rule of the approximation to normal is fine for this example. Am I right? or Can you create sampling distribution with only one sample in general? Thank you in advance.
Sorry, what are these tests you are running at 11:30 ? Before you said n/N and its relationship to 5%. Now you are saying stuff about np and n(1-p) and their relationship to 5%. Whats going on?
@BrandonFoltz -- I have the same question as the OP above. I assume that I'll find out why you are using 5 to determine if you can use the normal distribution in a prior video, but when I see it here in the video, I have no idea where it comes from. The 5% that Ethan mentions I think is just a test of significance and unrelated to the 5 mentioned at 11:30 (I think?).
Hi Brandon, thanks for this wonderful video but I'm a little confused in 10:19, i thought the formula for z-score is "z = (x-μ)/σ," are there different formulas in calculating z-score for standard error for mean and standard error for proportion ?
It's sampling distribution of sampling proportion standard error of mean is as standard deviation of sampling distribution. Maybe that is the reason of using standard error instead of standard deviation of population or sample