DT Wk 14: Correlation, Regression & PRE (NO Peer Required) 88 unread replies.88 replies. We come back to a non-crosstab PRE this week - using the SAME MOA "r" as we did for ANOVA. As you should have read, PRE for stats using for I/R variables focus on the mean of the variables; whereas, the PRE for Crosstabs can only tell us how much we are likely to reduce our errors by guessing. Nominal variables don't produce a meaningful mean, so the PRE wording has to change from that of one that does. In your OWN Words: 1. Why/When do we use a correlation or regression procedure instead of a crosstab procedure? 2. Which PRE Statement below goes with Which MOA: a. By knowing one's sex, we reduce our errors by 13.7% in predicting attitudes towards sports --- Which MOA are we using this type of statement with (R, Lambda/Gamma/Phi)?_________________ b. We see 13.7% of the variation in attitudes towards sports is being explained by the variation in one's marital status.--- Which MOA are we using this type of statement with (R, Lambda/Gamma/Phi)?_________________ 3. What do we need to do to "r" in order to use it in a PRE