When deciding which variable to use in my model I did a correlation analysis of all the variables. The correlation for Month and Day were extremely low at less than 0.1. I used dummy variables for the months in quarters and the days for weekend vs. weekday. The resulting numbers still had a correlation below 0.1. Therefore I discarded these factors
I also disregarded the share as it is just a different method of measuring the success of a movie. The rating is not dependant on the share.
I then did a Regression analysis with the remaining factors. I ran several regressions and determined each other factor was statistically insignificant as they equaled zero in my model.
After deciding on the two factors Fact and Star and separating each network, I ran a regression for each. There were a few data points in each model that did not fit. I then disregarded those models and deleted the outlying points.
E-pasta adrese, uz kuru nosūtīt darba saiti:
Saite uz darbu: