STAT 2040 Study Guide - Final Guide: Normal Distribution, Confidence Interval, Null Hypothesis

• School: University of Guelph
• Department: Statistics
• Course: STAT 2040
• Professor: Jeremy Balka

Document Summary

Must have simple random sample from the population. Population must be normally distributed; or use the central limit. Theorem (large sample size (n): margin of error = z /2 ( N : has a z-distribution; aka a standard normal distribution, z values for confidence levels: 99% = 2. 576: effects on moe, increases as increases (linear, decreases as n increases (non-linear, increases as the confidence level increases (1- value) (non-linear, estimation of ci when is unknown, moe = t /2 ( s. And then it depends on the degrees of freedom (df: same assumptions as z-distribution, but some rough guidelines: If n between 15 and 40, t will perform reasonably well. If n<15, have to check normality first: if not normally distributed, can: Use distribution free procedures: can check normality with a normal quantile-quantile plot should be quite linear. Probability (type 1 error | h0 true) = : type ii: not rejecting h0 when it is false.