STA 100 Study Guide - Midterm Guide: Bonferroni Correction, Random Variable, Test Statistic
Document Summary
A (1- )100% ci for (m1-m2) is: ( 1- 2) +/- t /2 (s1 (s2. *if both bounds are positive, it suggests that m1 is. *if both bounds are negative, it suggests that m1 is true smaller than m2. *if the bounds cover 0, it suggests no significant difference between m1 and m2. *power = 1 = probability we reject h0 when ha is. *test statistic: ts = ( 1- 2) 0 / (s1. *the number of se"s our data is from h0. Hypothesis testing & ci of two dependent samples: *trial where subjects are measured before & after. *assume differences are randomly selected/ independent; normally distributed. *p-value: the probability of observing our data (or more extreme) if h0 is true. *or: if the h0 were true, how likely it is to sample out particular data (or more extreme) *one-sided ts: pr{t > ts} or pr{t < -ts} *h0: md = / > / < 0.