PSYCH 210 Lecture 22 Outline of Last Lecture I. Within-Subjects ANOVAa. Calculations for F observedb. Hypothesis testing stepsOutline of Current Lecture I. Finish Within-subjects ANOVAa. Hypothesis testing stepsb. Post-hoc testc. AssumptionsII. Factorial ANOVACurrent LectureI. Continue counterfactual conditioning examplea. Return to steps of hypothesis testingi. We cannot necessarily say our hypothesis is correct even if we reject H01. Calculate the critical value for Tukeya. Tukey critical = qt√MS error/nb. See Tukey handout from last classi. Conclusion1. Do not state significant difference ‘between’IV and DV; rather significant difference in DV‘according to’ IV (state conditions)2. Include results of omnibus F3. Include results of post-hoc test4. Describe pattern of means5. Handouta. Will not have to write paragraph on exam!ii. Assumptions for Within-Ss ANOVAInterval or ratio dataiii. Independent observationsiv. Normal distributionv. Homogeneity of variancevi. Homogeneity of covarianceThese notes represent a detailed interpretation of the professor’s lecture. GradeBuddy is best used as a supplement to your own notes, not as a substitute.1. Fits assumption:a. Individual ppl maintaining rank order across conditionsA B CP1 9 (#1) 10 (#1) 12 (#1)P2 8 (#2) 9 (#2) 10 (#2)P3 6 (#3) 7 (#3) 9 (#3)2. Does NOT fit assumptionA B BP1 9 (#1) 2 (#3) 7 (#3)P2 8 (#2) 6 (#2) 15 (#1)P3 6 (#3) 15 (#1) 10 (#2)II. Compare Between-Ss and Within-Ss designsa. Example on handouti. Not significant for Between-Ss, but YES significant for Within-Ss1. What accounts for the difference in significance for each test in the example?a. Same idea as difference between related and independent-sample t-testsb. Between-Ssi. Fobs = MSB/MSw= 67.6/19.25 = 3.512 = NOT SIGNIFICANTc. Within-Ssi. Fobs = MSB/MSE = 9.197 = SIGNIFICANT!III. Factorial ANOVAa. Why do we need a new test?i. So Far:1. ANOVAa. Ability to test > 2 Tx groupsb. Use different Ps in each group (between-Ss) or the same Psin all groups (within-Ss)ii. What if we want to test more than one IV (or attribute variable) at the same time?1. Example: sociability in dogsa. Adult Dogs: Sociability (to people)i. Factors (IV):1. Prior treatment by people2. Early socialization with other dogs/people3. Genetics4. Current Context5. Breedii. Testing single Independent Variable only looking at effect of one ‘slice of the pie’ on the entire situation1. ‘Empty’ part (of ‘pie’) = unexplained variabilityiii. General definition:1. Two of more Independent Variables (not just levels/conditions of one IV)2. For each IV, can also have 2+
View Full Document