Audience Diversion Due to Cable Television: An Application of Nonlinear, Nondiagonally Weighted, Generalized Least Squares
A model of television audience shares is estimated and applied to simulate the effect of cable TV carrying distant signals on local stations' audience shares. The model is nonlinear, with a complex error covariance matrix; transformations are used to obtain generalized least squares estimates using an ordinary nonlinear regression package. The conclusion: TV broadcasting will continue to prosper, despite increasing competition from cable. (Author).
What people are saying - Write a review
We haven't found any reviews in the usual places.
121 counties adjacent markets Attractiveness indices estimated audience diversion due audience share equation Below-100 Markets Broadcasting and Cable cable and non-cable cable audience cable observations cable system Cable Television coefficient is constrained complex error covariance CONSTRAINED ESTIMATES DB indicates distant dependent variable distant network stations distant network VHF distant VHF stations due to cable dummy variable effect of cable equa error covariance matrix error terms error variance Estimated coefficients F test first-stage regressions full broadcast day Full Day gression independent UHF stations index for distant indicates distant network indicates local stations INTABC least squares McGowan network signals nonlinear regression package number of diaries o c pVn off-the-air and cable OFF-THE-AIR ATTRACTIVENESS off-the-air observations ordinary nonlinear regression Park percent precisely estimated programs provide nonduplication protection Rand Rand Corporation residuals S9ARE sample sample size smaller markets station share station types Table television audience tions viewing 3F 9F watch local stations