L7: Parametric Tests Part 2

Question 1

What test do we use to compare data among >2 independent groups with continuous, normally distributed data?

Answer 1

One-way ANOVA

Question 2

What are the assumptions for one way ANOVA?

Answer 2

• underlying populations are normally distributed

- underlying populations have equal variances

Question 3

Why do we need to use one way ANOVA to compare > 2 groups rather than conducting all possible independent samples t tests?

Answer 3

To control the overall probability of making a type 1 error on the predetermined significance level (alpha) of 0.05.
If we use multiple independent samples t test, the type 1 error is inflated.

Question 4

What is the null and alternative hypothesis for one way ANOVA?

Answer 4

H0: All the means of the underlying populations are the same.
H1: Not all the means of the underlying populations are the same / The means of at least two of the underlying populations are different.

Question 5

Why do we need post-hoc tests in one way ANOVA?

Answer 5

• once we know that there is a statistically significant difference between the groups, post hoc tests can help us identify the difference(s) while controlling the overall probability of making a type 1 error on the predetermined significance level (alpha)

Question 6

What does a conservative post hoc test mean?

Answer 6

• a larger difference between means is required for significance when a more conservative post hoc test is used
• more conservative -> reduces statistical power -> greater chance for type 2 error, but chance of type 1 error can be controlled

Question 7

What is the use, strength and limitations of using bonferroni adjustment?

Answer 7

• very simple, widely applicable
• very conservative -> statistical power much reduced -> greater chance of type 2 error, but control for type 1 error better
• used for both comparing to one control group and all pairwise comparisons

Question 8

What is the least significant difference (LSD) test?

Answer 8

• no correction to control overall significance level
• least conservative -> greater chance for type 1 error
• greater statistical power

Question 9

What is the Tukey’s test?

Answer 9

• more conservative than LSD test

Question 10

What is the Scheffe procedure?

Answer 10

• most conservative-> least statistical power

Question 11

What is the Dunnett’s test?

Answer 11

Used to compare groups against one single control group

Question 12

What are the tests used for comparing all possible pairwise comparisons?

Answer 12

• LSD test (avoid)
• Tukey’s test
• Scheffe procedure

Question 13

What is the equation for bonferroni adjustment?

Answer 13

alpha divide by m (the number of pairwise comparisons)

Question 14

When reading statistical output after bonferroni adjustment?

Answer 14

Just compare against 0.05

Question 15

When is repeated measures ANOVA used?

Answer 15

Used when there are changes in a particular measure on the same group of subjects over different conditions eg time (when the groups are not independent, but related)