Commentary: Searching for media effects.

The central conclusion reached by Hanewinkel et al. is that exposure to incidental portrayals of alcohol use in US movies has contributed to the early onset of alcohol use by a group of German adolescents. As is always the case when relying on cross-sectional data, the authors do acknowledge limitations in their ability to assume causal order as well as the likelihood that other confounding factors may have contributed to the relationship found. This study opens the door to an important discussion regarding issues that constantly vex researchers who seek to understand the impact of media messages on our beliefs and behaviours. Several important points are worth noting. First, is the possibility that frequent movie viewing—or any other form of excessive media use for that matter—may actually represent a ‘socio-cultural marker’ indicating the existence of a youth subculture for which alcohol use is already a behavioural norm. Those adolescents who watch the most movies also might be the least supervised, have the most time in their hands, and are, in general, low academic performers (and, therefore, might be inclined to associate with others who are similar to themselves in thinking and behaviour). Hanewinkel et al. confirm this: age, gender, year in school, parental drinking, peer drinking, school performance, etc., all predict alcohol use. As children grow older, associate with peers who drink, spend more time with the media, etc., they are more likely to drink. Second, the ubiquitous nature of mass media in our lives makes it difficult to partial out or detect some of its subtle effects. In many ways, the media is like ‘cultural river’ whose current moves us down stream. We are often not conscious of the degree, or speed, to which it moves us. As previously suggested, children and teenagers who watch a lot of movies are also likely to watch a lot of television and are equally likely to use other forms of media that also might include incidental portrayals of alcohol use (e.g. magazines, Internet). Because of the cumulative effect of this multitude of consonant messages permeating our lives and simultaneously shaping our culture, it would be difficult to identify how a single message, or messages from a single medium or genre, independently and uniquely cultivate normative beliefs, outcome expectancies, and behavioural intentions—especially in small, subtle increments. This is both an epistemological and methodological issue. As researchers we constantly try to overcome this, but, in the end, we are always forced to acknowledge this key limitation of our work. What is agreed upon by most researchers interested in potential media effects, however, is that the accumulation of all these messages creates a symbolic environment that can become very real—and believable—to us. Finally, and perhaps most vexing, is the issue of attention. We must differentiate between exposure to a media message and attention to it. Attention involves the cognitive analysis, or encoding, of information into our memories. The fact is our world has become so saturated with mediated messages that it is not possible to attend to all of them. As a consequence, we selectively attend to only a small portion of the messages to which we are actually exposed. Simply stated, exposure does not guarantee attention. In other words, we can ‘see’ without actually ‘seeing’. This filtering process not only influences what we actually see but also the way in which we process, store, and remember the message. Because the filtering process is a function of an individual’s previous experience and knowledge, we retroactively fit new messages to our existing schema. As a result, no two individuals experience a media message in quite the same way. This would have to be particularly true of children and teenagers, who are far less cognitively complex. The participants in the Hanewinkel et al. study may have reported seeing many of the movies listed in the survey instrument, thus making it possible to estimate some level of exposure to alcohol use. What we don’t know, however, is whether they actually attended to the specific incidents of alcohol use or, even more importantly, to what degree they comprehended and internalized the drinking depictions. So the question of what the participants in this study might have taken away from these movies, particularly in regard to normative beliefs and expectancies for alcohol use, and to what degree their behaviour was specifically influenced, remains uncertain. Hanewinkel, Tanksi, and Sargent have certainly presented us with a provocative piece of research—perhaps not so much because of what they have found, but because of the perplexing questions they raise for us.