PSY-PC 2110 Study Guide - Final Guide: Complement Factor B, Linear Combination, Sampling Distribution

• School: Vanderbilt University
• Department: Psychology and Human Development
• Course: PSY-PC 2110
• Professor: Professor Maria Osina

Document Summary

We use anova because if you would use 2 independent samples t-test, there would be an increased likelihood of type 1 error (each test has 5%) X = grand mean x . j = group mean. There are 4 independent groups with different means because of treatment effect and sampling variability (error), which are captured by types of deviation. Treatment effect & error: x . j - x . Total variation without group separation: xij - x . Nj = number of participants in group. F-ratio x )^2 nj (xij - x . j)^2 (x . j - Among group variability ss reflects error and treatment effects. Within ss reflects only error because group differences are not taken into account. Total ss reflects total variability in the dataset; treatment effect + double error. If you have unequal n, you will have new means, and you would need to take weighted average of the group means. X = (cid:4666)(cid:3015)(cid:3037) (cid:3051). (cid:3037)(cid:4667) (cid:3041)(cid:3037)