When you have a bunch of independent variables to pick from, a good start is to check their correlation with the dependent variable. Youtube here. Look for an 'r' value (Pearson Product Moment) of greater than + 0.7 or smaller than - 0.7. Look at this example, where the dependent variable is Duration.
Length looks hopeful, so try that to start with. You can add others are you go, taking out those that don't work. Watch for multicollinearity, where two or more independent variables are explaining the same thing. Speed and height seem to be collinear (see their high 'r' values) so probably they wouldn't work. Thought experiment: wouldn't you expect that being dropped from a great height would add to your speed?
Youtube on multiple regression here.
No comments:
Post a Comment