I recently came across an intriguing paper by a pair of academics from Indiana State University's College of Business. Writing in the Journal of Statistics Education (Vol. 17, no. 1), Constance H. McLaren and Concetta A. DePaolo explore the use movie box office data to teach undergraduates the basic principles of business statistics (in a paper simply entitled Movie Data).
Using a dataset comprising weekend and daily box office data, plus total US domestic gross figures, for 49 titles, the authors explain how such a time series can provide practical examples of graphical analysis, forecasting, curve-fitting, rate of change analysis etc. The paper includes a general overview of the approach to using movie data in this way, alongside exercises and assignment questions for classroom use.
The significance of the movie dataset is twofold. First, a shed load of rich data is readily available from free sources online (the authors got their data from The Numbers, but as the list of box office sources in the left hand sidebar indicates, there are plenty of alternatives).
Second, and perhaps more importantly, movie data help to connect the abstract realm of statistics to the tangible, real world experiences and enthusiasms of students. As the authors note,
So it’s a win-win from a pedagogical point of view. Instructors can easily muster up-to-date and detailed datasets for classroom use, and students get to play around with numbers within a frame of reference they know and appreciate:
In a funny old way it’s another fine example of the power of the moving image, and a reminder, if such was needed, that the film business is above all else a numbers game.







Anyone interested in applying statistics to analyzing box-office data might be interested in a piece I wrote a few months ago on the relationship between genre, gender, and the amount of gross taken by a film during its three-day weekend in the UK. In short, films with female characters (e.g. romance films) take a lower proportion of their gross at the weekend than films with male characters (e.g. action, crime). You can read the full study at http://nickredfern.wordpress.com/2009/03/26/gender-genre-and-the-uk-box-office/.
Posted by: Nick Redfern | 06 July 2009 at 01:23 PM
Thanks for your comment, Nick. I look forward to reading your article- sounds fascinating. I've added your blog to the blogroll.
Posted by: Jim Barratt | 06 July 2009 at 01:47 PM