Determined to adhere to the letter of good social science, Comstock et al. take no truisms for granted in their exhaustive review of the literature on the impact of television on our cultural and individual psyches. The going is tedious and marked by repeated recapitulations as the team conscientiously examines familiar issues: who watches what and how much; the influence of the medium on children; the effects of commercial and political advertising; the built-in potential for arousing aggression; the presentation of news. Better, indeed best here, are the analyses of the corresponding research-studies: Comstock's group is alert to specious premises, misguided experimental methods, erroneous interpretation of data, and logically false conclusions. ""Our long-range goal is the development of policy,"" the authors attest, and to that end they construct a theoretical base--a formula expressing cause-and-effect relationships between television and behavior--whose jargon and algebraic idiom momentarily mask its ingenuousness. A typical conclusion: ""The evidence at present favors the hypothesis that exposure to television violence increases the likelihood of subsequent aggressiveness."" Sponsored by the Rand Corporation, Comstock's whole project tacitly addresses the movers and shakers charged with assigning policy-studies of the sort he recommends, but it seems unlikely that they or even specialists in the field will find the book's zero-based approach as necessary as it is sufficient. Like an extensively annotated bibliography, too scholastic to read but valuable for serious reference.