part 6 slide

Question 1

looking at multiple factors simultaneously allows us too…

Answer 1

1. study the factors in one experiment instead of multiple experiments
2. study how conditions interact

Question 2

When are two factors crossed?

Answer 2

If every possible combination of factor levels occur in the design.

Question 3

What is a factorial structure, or factorial design?

Answer 3

When the factors are crossed i.e. every possible combination of factor levels occur in the design

Question 4

How many degrees of freedom are found in the interaction?

Answer 4

I = number of levels in factor A

J = number of levels in factof B

(I-1)(J-1)

Question 5

How many degrees of freedom are found in the residual?

Answer 5

N = totall number of observations in the residual error

I = number of factor levels for factor A

J = number of factor levels for factor B

N-(I*J)

Question 6

Why is replication important?

Answer 6

Replication gives more precision to our estimates of model parameters

Question 7

True/False

Replication reveals no information about the errors

Answer 7

False

Replication reveals informaiton about the errors, which then allows us to make inference about model parameters

Question 8

What does type 1 in SAS give us?

Answer 8

Allows the first term to explain as much of the variability away as possible.

Question 9

What does type III in SAS give us?

Answer 9

It allows us the terms to explain variability from last to first.

Question 10

What is the equation for the non centrality parameter?

Answer 10

N = number of subjects in each of the groups we are comparing

Alpha^2i - find the mean of all observations, subtract each observation from this mean, square each of the reamining integers and add them up.

MSE = standard error

N(Alpha^2i)/MSE