Exercise 10: Compare standardized coefficients with CI's, and more

Use this same data file from last time (don't forget to filter out -1's). As you'll recall, this is a simulated data set in which grades are predicted by homework hours and parent education. Standardize all variables, and run a multiple regression with both predictors (using the standardized variables). Make sure to turn on confidence intervals in the Statistics options. Answer the following questions using complete sentences to explain your reasoning:

(1) Is one predictor more of an influence on grades than another? Is it significantly so? Refer to the CI's. (Remember the usage of b vs. β.)

(2) Imagine you want to "take parent education out of homework." In other words, you want to take the variance in homework that can be accounted for in terms of parent education, and retain what is left over. Use the concept of a residual and build a model between homework and parent education. Describe in a few sentences how you'd save what is left over. Then, do it. (Note: You can do this either on the standardized variables, or on the original variables. It should not matter. Feel free to do it on both. You should be able to talk about what you're doing in these two (non-/standardized) cases... you should be able to interpret b vs. β, and so on.)

(3) Build a model between what was left over in (2) above and grades. Compare the b/β you obtain here, and the original b/β you obtained in the full, two-predictor model. Write a few sentences describing what you see. (Again, go ahead and do it on either the standardized or unstandardized versions. It may be useful to compare those 2 analyses anyway. If you feel confident about what you're doing in this exercise, just do it on one of these versions. In other words: It is not necessary to submit this exercise for both.)

Send answers to psyc7302@gmail.com with subject line "EXERCISE 10."