Predicting Magazine Audiences with a Loglinear Model.
FLORIDA STATE UNIV TALLAHASSEE DEPT OF STATISTICS
Pagination or Media Count:
A loglinear model for predicting magazine exposure distributions is developed and its parameters are estimated by using the maximum likelihood technique. The accuracy of the loglinear and a Dirichlet-multinomial model are compared using 1985 AGB McNair data. The result show that the loglinear model has significantly smaller prediction errors than the Dirichlet-multinomial model. A simple algorithm for optimal media scheduling is given. Keywords Advertising Statistical analysis Efficiency. Author
- Statistics and Probability
- Sociology and Law