Michael Chambers
Predictive Power of Film Attributes with Respect to Audience Demographics © 2013
Film attributes were analyzed to investigate if they could be used to predict gender/age demographic proportions of a film’s audience. If relationships exist, film attributes could be used to help automate the process of allocating newly released films to theatre markets. Film attribute data were collected for wide release films seen between 2010 and 2012 by loyalty program card holders. Moviegoer demographic data were also gathered through the same program. The data were aggregated into three regions to investigate if there were any obvious spatial patterns regarding gender/age compositions in these motion picture markets. Multiple linear regression was used to determine if any relationships existed between film attribute and gender/age demographic variables for predictive purposes. Results showed that models using selected film variables had a moderate predictive strength for the parent and older cohorts. Although the predictive strengths were not very strong, the results adequately show that film attributes can be used to predict audience gender/age proportions.