Epsilon and the 10% Rule are our FIRST line of defense for quickly determining both does there appear to be a pattern/association and the approximate strength, but they are NOT the final “summary”.

DT Wk 10S: Measures of Association (MOA) & PRE: OPT’L E.C.
No unread replies. No replies.

Epsilon and the 10% Rule are our FIRST line of defense for quickly determining both does there appear to be a pattern/association and the approximate strength, but they are NOT the final “summary”.

In your OWN WORDS:

1. MOA: Overall, what do the different Measures of Association do for us, why use them instead of Epsilon and 10% Rule?

1a. Give an example as to when it would be appropriate to use Lambda and Gamma (include table size, level of measurement, and any exceptions shown on Pp).

1b. When would we use Phi and when would we use Cramer’s V

2. PRE: You were introduced to PRE with ANOVA – but the PRE wording was for I/R variables (as the DV is I/R). How is the PRE wording different than that for non-crosstab procedures (i.e. what we used for ANOVA))?

2a. Why is a PRE helpful in understanding our data?

2b. What PRE “issues” do Phi and Cramer’s V have?

Last Completed Projects