Thanks for the feedback Jean-Luc. I don't have any suggestions regarding review questions online. (I work only with my own materials, and I don't look too closely at what others are doing.) Cheers.
Can I understand that when calculating standard error in a hypothesis testing, both pooled standard error and the standard error calculated by each sample individual's proportion is eligible?
It depends what you mean by "eligible". A Z statistic based on an unpooled standard error would still have (roughly) the standard normal distribution under H_0. Using that Z statistic in hypothesis testing for large sample sizes would work reasonably well, but we typically prefer the pooled approach that was used in this video.
why use z-stat instead of t-stat? Since the standard error is computed from data with p hat, should it follow a t-distribution? thx
Exactly!! I think we can not use Z test here since we use sample SE
Love the pace. Do you have any suggestions for some good review questions online?
Thanks for the feedback Jean-Luc. I don't have any suggestions regarding review questions online. (I work only with my own materials, and I don't look too closely at what others are doing.) Cheers.
Can I understand that when calculating standard error in a hypothesis testing, both pooled standard error and the standard error calculated by each sample individual's proportion is eligible?
It depends what you mean by "eligible". A Z statistic based on an unpooled standard error would still have (roughly) the standard normal distribution under H_0. Using that Z statistic in hypothesis testing for large sample sizes would work reasonably well, but we typically prefer the pooled approach that was used in this video.