S2 lecture 7 - intermediate analysis: nominal data

Question 1

What does Pearson’s Chi-Square test do?

Answer 1

measure statistical association for categorical (nominal) data.

Question 2

What are the assumptions of Pearson’s Chi-square test?

Answer 2

Each participant, item, or entity must only contribute to just one cell of the table. More than 5 frequencies in each cell, but 20% are allowed to be less.

Question 3

How should you lay out your data in an independent sample design?

Answer 3

one row per participant, dependent variable column, grouping variable column.

Question 4

How should you lay out your data in a dependent-measures design?

Answer 4

one row per participant, one column per condition.

Question 5

What are alternatives to Pearson’s Chi-Square test?

Answer 5

Phi, Cramer’s V, Lambda, and Kendall’s statistic.

Question 6

If an assumption for Pearson’s Chi-Square test is not met, what alternative test(s) could we use to measure the strength of association?

Answer 6

Phi or Cramer’s V.

Question 7

What is the effect size guideline for Cramer’s V?

Answer 7

0-0.1 is weak
0.1-0.3 is moderate
.3-1 is strong