STA 101 Study Guide - Midterm Guide: Confidence Interval, Summary Statistics, Point Estimation

• School: Duke University
• Department: Statistical Science
• Course: STA 101
• Professor: Jesse Windle

Document Summary

Increase se if t2 is worse: assumptions, random sampling generalizability, random assignment / experiment causality. 2 sets of means: paired means, paired two sets of data that have special correspondence/are not independent. Identify the point estimate: = xx i. ii: hypotheses for paired means, h0: diff = 0 a b = 0, ha: diff 0 a b 0, conditions. If sample w/o replacement ndiff < 10% of population: sample size / skew, ndiff 30 v. Ndiff: xxdiff ~ n(mean=0, se, calculate z-score, z = xdiff diff. Sediff: sketch curve and shade for +/- x value ix. If sample w/o replacement ndiff < 10% of population: sample size / skew, ndiff 30 iii. Identify other variables of interest / summary statistics: xxa, xxb, xxdiff, sdiff, ndiff, se xxdiff = s diff. Identify z-score associated with indicated confidence level: 90% = 1. 65, 95% = 1. 96, 98%=2. 33, 99%=2. 58, plug into xxdiff +/- z* se xxdiff, solve vii.