This preview shows page 1. Sign up to view the full content.
Unformatted text preview: EXECUTIVE SUMMARY We began the project by running examining each variable’s distribution. Most of them seemed approximately normal except for the percentage of men reading the magazine which represented a bimodal distribution. We later looked at the correlation between the dependent and the independent variable and found Percentage of Men as well as the Median Income to have a fairly low correlation with the page cost, implying that they must not be crucial factors in determining the actual price of each page in the magazine. After analyzing each variable individual variable we conducted a multiple regression analysis. The results were puzzling as it showed a moderately high correlation, yet had two variables over the .05 alpha level. In order to further determine if the equation derived from the multiple regression test was a good fit for the date we looked at the residual plot. When examining the residual plot we saw some a “funneling in” effect which suggests that the spread of the error terms is decreasing as the horizontal...
View Full Document
This note was uploaded on 10/12/2008 for the course BUAD 310 taught by Professor Lv during the Spring '07 term at USC.
- Spring '07