| Abstract: | Abstract This paper tests and evaluates the best available methods for estimating the duplicated audience of pairs of television programs, pairs of magazines and television-magazine pairs. Previous research in audience duplication estimation is surveyed, and the best existing methods for some situations are found to be inadequate or nonexistent. Improved estimation methods are provided for these cases. In particular, new models are developed for estimating the duplication of audiences of different episodes of a television program, and for estimating duplication of audience within particular target markets. The various models are empirically tested using diary data from a large, national sample. |
