L33 Psych 300 Lecture Notes - Lecture 8: Type I And Type Ii Errors, Null Hypothesis, Statistical Power
Document Summary
When we calculate a z-test, we compare data from our sample to a: Distribution representing the idea that the null hypothesis is true. Identify the populations, distribution, and assumptions; then choose the appropriate hypothesis test. State the null and research hypotheses in both words and symbolic notation. Determine the characteristics of the comparison distribution. Determine the critical values (cutoffs) that indicate the points beyond which we will reject the null hypothesis. Decide whether to reject or fail to reject the null hypothesis. Two populations may have different means, but their distributions could overlap a lot. Effect size: if you have a large enough sample size (n), even a very small real difference in population means will turn out to be statistically significant. Mean differences, alone, can"t tell you enough about effect size. Tells us to what extent we can believe the effect is real. It"s another question to what extent the effect is important for an particular practical purpose.