Exercise 9: Run multiple regression with some new data

Use this data file, containing some simulated data again involving grades, homework, and parent education (similar to Keith's setup, with the 10-20 scale). In all answers, please report the statistics using t's and F's in the conventional manner (reporting df's, etc.). (NB: SPSS gives you df1 and df2 for the F test. Just get into the habit of reporting degrees of freedom with your F's and t's. We will go into detail about them later, so don't worry about being confused if this is unfamiliar or you are rusty).

(1) Notice anything funny with the grades? Students who did not take the test have a grade of -1. Filter these out of the data using your skills in SPSS.

(2) Run the three possible models with these variables (grades on education, grades on homework, then grades regressed on both) and just compare the R^2, b, beta values, etc. Write a few sentences comparing these values in each model. This can be brief and casual, don't worry about how you're formatting it. (Make sure to report actual numbers when you refer to these values.)

(3) Report the significance of the two predictors and the overall model's significance for the three-variable model (grades regressed on homework and parents education).

Send answers to psyc7302@gmail.com with subject line "EXERCISE 9."