Biology 2244A/B Chapter Notes - Chapter 14: Sampling Distribution, Confidence Interval, Normal Distribution
Document Summary
Tests of significance: asses evidence provided by data about a claim concerning the population. Sample mean close to population mean: within central bulk of sampling distribution. Indicates sample mean is likely if parameter is true. Sample mean far from population mean: unlikely if parameter is true. Hypotheses: always refers to population, not particular outcome. One sided: only says that the parameter is larger or smaller than the null value. Two sided: says the parameter is different from the null value. Choose 2 sided if unsure which direction alternate may vary. P value: probability that measures strength of the evidence against a null hypothesis. Say the observed result unlikely if h0 were true. Alternate hypothesis: sets direction that counts as evidence against h0. Significance levels:we can use fixed values as standards for evidence against. 0. 05: if p< or = 0. 05, no more than 1/20 chance a sample gives evidence this strong if h0 is true.