In 2009, three Grinnell students (Allie Greenberg, Hannah Lytle, and Phillip Brogdon) conducted an analysis to estimate the probability of winning the Academy Award for Best Picture. The Academy Awards, or “Oscars,” are given annually to honor high achievement in the film industry. The Academy consists of over 6000 members who nominate their colleagues and vote to decide on the winners of this prestigious award. In their analysis, these students included all films nominated for Best Picture from 1979 to 2008. Winning Best Picture was considered the response (1 represents a win), and the explanatory variables included whether or not the picture won any of 17 other awards that were given out during that year’s ceremony.
a. Create a logistic regression model using all 17 explanatory variables. Which variables appear to be most significant?
b. Create and compare multiple logistic regression models. Submit the model with the fewest number of terms that best estimates the probability of winning the Best Picture award.
c. Academy Award for Best Picture in 2009 went to Hurt Locker. Use your final model in Part B to predict the likelihood that Hurt Locker would win the Best Picture award. Avatar and The Blind Side were also nominated. Use your final model to estimate the probability that each of these movies would win Best Picture.