BIOL2300 Lecture Notes - Lecture 11: Coin Flipping, Simple Random Sample, Rofecoxib
Document Summary
H1: patients taking vioxx: merck have more heart attacks and strokes than those not taking vioxx. Ho: null hypothesis, equal number of heart attacks and strokes in control and treatment groups. Requirements to check before approx population proportion. Fixed number trials, trials independent, 2 outcomes: success of failure. P q are estimated by p hat, resp q hat, proportion of success and failures in samples. Point estimate: single value used as approx for population parameter. P hat-e, p hat+ e, p hat +/- e. 1- alpha confident: p lies in interval. P hat: x/n, n is sample size. (p hat - p)/standard deviation: z score. P hat: number (estimating a mean), p: proportion. Less than 5% in the tail end. Coin flip: does or doesnt have heart attack within a year. Take average (proportion of people who have heart attack in year) 1 is yes, 0 if you dont. Mistake (dont do this): e[n times x]