PSYC 316 Lecture Notes - Lecture 11: Null Hypothesis, Factorial Experiment, Libido

• School: Concordia University
• Department: Psychology
• Course: PSYC 316
• Professor: Dean Graham

Document Summary

Y = b0 + b1x1 + b2x2 + e. Largely remains the same: hypotheses (omnibus): i. ii. H0: 1 = 2 = 3 = j. H1: not all the j are equal: translation (anova = glm), sum of squares (ss): i. Sstotal = ssbetween + sswithin = ssmodel + sserror: multiple comparisons, contrasts = comparisons: i. Complex: compare more than 2 means, hypotheses: i. ii. iii. iv. H3: 1 2 3: planned a priori: i. ii. Often based on theory, previous research, hypotheses: post-hoc ( unplanned comparisons ): i. ii. iii. Bonferonni and lsd: anova key questions, how good is the model? i. ii. iii. < 0. 05): dose accounted for 37% of the variation in libido. Single variable data coding: glm, model statement: i. X = 1, 2, 3 ii. iii. iv. b0 = intercept b b#x# = b for that level of variable (don"t need to multiply). Low dose = 5. 00 + (-1. 8) = 3. 2.