Chapter 8: Hypothesis Testing

/ e, d)

l

what does hypothesis testing do?

uses sample data to test the validity of a hypothesis about a population parameter

steps of hypothesis testing

1. state hypothesis; 2. set criteria for decision; 3. collect data and compute statistics; 4. make decision

null hypothesis

there is no change/difference/effect (no evidence)

alternative hypothesis

there is a change/difference/effect (sufficient evidence)

alpha level

significance level; probability used to define "very unlikely" outcomes; most common alpha level = 0.05

null hypothesis is rejected when

sample statistic is located in the critical region = statistically significant

fail to reject null hypothesis when

sample statistic is not located in the critical region

type I error

rejecting a null hypothesis when the null is actually true; false positive; alpha level = probability of making a Type I error

type II error

fails to reject a null hypothesis that is really false; false negative

Cohen's d

measure of effect size; mean difference/standard deviation

Cohen's d of 0.2

small effect

Cohen's d of 0.5

medium effect

Cohen's d of 0.8

large effect

statistical power

the probability that the test will correctly reject a false null hypothesis; will detect a treatment effect if there is one; 1 - probability of a Type II error

increases in power

increased effect size, large sample size, using a one-tailed test

decreases in power

reducing the alpha level, using two-tailed tests