The sample data table Decathlon.jmp (Perkiömäki, 1995) contains decathlon scores from several winners in various competitions. Although there are 10 events, they fall into groups of running, jumping, and throwing.
(a) Suppose you were going to predict scores on the 100m running event. Which of the other nine events do you think would be the best indicators of the 100m results?
(b) Run a Stepwise regression (from the Fit Model Platform) with 100m as the Y and the other nine events as X. Use the default value of 0.25 probability to enter and the Forward method. Which events are included in the resulting model?
(c) Complete a similar analysis on Pole Vault. Examine the events included in the model. Are you surprised?