Integration of AHP and Regression Analysis in Forecasting Attendance in a Movie Theater

Abstract

Forecasting the number of attendees in a motion picture has often been considered a wild guess since there are many factors or variables involved in forecasting the numbers.  There are many contributing factors to consider when attempting forecasting theatre attendance. Among many possible variables the ones with the most significant impact that are considered in this study are; the day that movie is playing, the time of day that movie plays, the ranking of the movie, the genre of movie, the length of time that the movie has been released and finally whether school is in or out. For the most part, these are quantifiable measures and thus should be able to derive an accurate forecasting module. At first, the aim of this study is to compare all the variables by the decision maker based on the Analytical Hierarchy Process (AHP) and to rank the variables based on the importance according to the AHP process. Once the variables are set, the regression analysis is applied to forecast attendance in a movie theater.  Multiple regression analysis is used in this study based on a sample of 711 observations. Using SPSS statistical software, a model is developed to forecast the number of attendees and the model provides R² = 0.760, which is a strong predictor. Finally, the hypothesis test is conducted to verify the accuracy of the regression model with the actual data and even with a = 0.10 the null hypothesis could not be rejected.