Week 6 - One-Way Anova

Question 1

One - Way Anova

Answer 1

Can express every ANOVA as a regression

Modelling information from Y by looking at groups (Why are some people low or high on Y)

Question 2

Experiment conditions allow to

Answer 2

Determine the differences between means of groups

Question 3

Null for ANOVA

Answer 3

Groups come from the same population (if different between groups - null is rejected)

Is difference far enough away from 0 to reject the null

Question 4

How to reject Null

Answer 4

Variation within the groups becomes important

SD used to estimate the SE of the difference between the means (groups)

If more than 1.96 = Significant difference

Null suggest that differences due to random individual difference ( if correct would be no significant effect of X (systematic variation)

Question 5

Grand mean

Answer 5

Mean of all the means

Question 6

Group mean

Answer 6

Mean of participants in one group

Question 7

Variation

Answer 7

Total - Data point - Grand Mean
Systematic - Group mean - Grand Mean
Residual - Group mean - data point (Not accounted for by being int he group)

Question 8

Anova

Answer 8

In ANOVA difference between groups is equivalent to systematic variation in regression

Every data point have the same systematic variation as every other data point in that group

Question 9

Line of best fit

Answer 9

Connect the mean of both group means (Grand Mean )

Question 10

Can create variance in the outcome

Answer 10

Experimental manipulation vs control group

Try to make groups vary as much as possible

Question 11

Variance already occurring in an outcome

Answer 11

Non experimental or quasi experimental design

Question 12

Regression VS ANOVA

Answer 12

Regression look for covariance between X and Y (How they vary together) - Systematic

ANOVA- Difference between means show the systematic effect
Variation of Y within different groups of X, index of random variation in Y within groups that are treated the same

ANOVA = In true experiment all other systematic source will stay in error term and will not covary with Y

Regression = Systematic variance not accounted for by X, will covary with Y and effect it (specification so important!)

ANOVA = Factor
Regression = Variable

Question 13

Random Assignment

Answer 13

Effect interpretation of ANOVA

Question 14

GLM and ANOVA

Answer 14

Data = obtained sample variance in Y
Variance = by how much are scores equal to the mean

Question 15

Similarities ANOVA and Regression

Answer 15

Both address same mathematical equation

Follow same GLM procedure

Question 16

i in statistical model

Answer 16

Number people in groups

Question 17

j in statistical model

Answer 17

number of groups

Question 18

Levenes test

Answer 18

Significance mean assumption is violated