Content and effects of alcohol advertising: comment on NTIS pub. no. PB82-123142.

Atkin and Block's study on the content and effects of alcohol advertising is discounted. The substudies involved in the research are described and their methodologies are criticized. Weaknesses are found in the following areas: coding procedures; definitions and interpretations of categories of advertising content; sampling procedures; and measures of exposure to advertising. It is felt that causal interpretations from such weak measures should not be made. The diversity of research methods used is deemed commendable. However, since all of the substudies use the same basic sample, it is of little surprise that convergent evidence results. Also, since all research methods used are felt to be inadequate, such convergence does not necessarily show external and internal validity.