Anova K>2

Question 1

What are levels/factors of i.v.

Answer 1

→ The diff levels a single variable can have in an exp
→ for e.g. Drug dosage of diff kinds

Question 2

What are the advantages of using k> 2 blwgroup design?

Answer 2

→ controls for type I error as we are sitting the p-alpha for entire exp.
→ we pool variability for the entire exp rather than just 2 groups
→ better interpretation of IV. On. D.V.

Question 3

What is the disadvantage for K72 design?

Answer 3

De don’t know where the diff lies or blw. which groups.

Question 4

Define F statistic.

Answer 4

Conceptually F =
error v. + treatment v. /error or unexplained. v.
→ a ratio of 2 independent - estimates of the same population variance
Mathematically =s sq groups* s2 error =
MS group/ MS Error.

Question 5

What are the assumptions for i-way anova. B/w group designs.

Answer 5

→ random sampling
→ normal distribution of DV’s
→ homogeneity of sample size - the should not be a big diff btw the N.
→ homogeneity of variance
→ atleast 7 scores in each group.

Question 6

Characteristics of f-distribution

Answer 6

→ asymmetric ar positively skewed
→ central tendeay, median =1 if n’s are =
→ starts at 0
→ no upper limit
→ sig. Scores at the right end of the tail.
→ F scares distributed based on ratio of variances

Question 7

What does no treatment effect imply?

Answer 7

→ no differences among the means
→ means do not vary from the grand mean.

Question 8

Distinguish b/w cohen’s f& d

Answer 8

Both are estimates of the diff b/w means of populations
Only diff is the design type
→ if k=2 then D - can compute delta as a function of d&n
→ it k greater than 2 then F.