MATH 140 Lecture Notes - Lecture 13: Sampling Distribution, Null Hypothesis, Alternative Hypothesis
Document Summary
There are two different hypotheses: h0: (null hypothesis) the observed results are entirely due to chance, ha: (alternative hypothesis) the observed results are not due to chance. Using data from a sample you can either (cid:862)reject(cid:863) or (cid:862)do (cid:374)ot reject(cid:863) the (cid:374)ull hypothesis: never say you accept the null hypothesis. Data fro(cid:373) a sa(cid:373)ple that goes agai(cid:374)st the (cid:374)ull hypothesis is called (cid:862)statistically significant evidence(cid:863) Null hypothesis always looks like this: p = _____ Alternative hypothesis can have 3 different forms depending how the question is worded: p _______ / p < ______ / p > _______ Definition: p is the proportion of all times the coin has landed on heads when flipped. The test statistic for this kind of testing is the z-score, which tells you how many standard deviations p is from the mean (*which is always p*)