NEWS MEDIA CHOICE AND AUDIENCE GRATIFICATIONS: AN APPLICATION OF THE EXPECTANCY-VALUE AND LEXICOGRAPHIC MODELS (USES, CHOICE BEHAVIOR, INFORMATION-PROCESSING)

Two data sets labeled 1984 and 1985 studies were collected under the same research objectives and were analyzed by the same procedures. The sample sizes for each study were 351 and 254, respectively. Four news media were examined and three choice-behavior models were employed to investigate the relationship between perceived gratifications and subsequent news media choice. The four daily news media were television, newspapers, cable TV, and radio. Two among the three choice models were compensatory models--the unweighted expectancy model and the weighted expectancy-value model, and one non-compensatory lexicographic model.;It was interpreted to mean that newspaper choosers tended to hold more cognitive capabilities than television choosers in discriminating news attributes and in weighing salient attributes for their subsequent news media choice. The present study, therefore, tied together empirically and theoretically the uses and gratifications approach, choice models, media selections, and demographics as well.;Two need attributes for news, informative-cognitive and stimulative-affective dimensions, were retained from SAS factor analysis for further analyses. Both the 1984 and 1985 studies showed consistent result patterns. The expectancy model performed as well as the expectancy-value model in predicting audience members' tendencies to choose television for later use. Both compensatory models correctly predicted television choosers at the level of about 75%. The lexicographic model performed even better than compensatory models in classifying television choosers' tendencies to select television 90% of the time. For newspaper choosers, the expectancy-value model performed better than the other two models. Demographic analyses showed that newspaper choosers were more educated than television choosers.