Evaluating an Anti-Drinking and Driving Advertising Campaign with a Sample Survey and Time Series Intervention Analysis
Murry, John P., Jr., Stam, Antonie, Lastovicka, John L., Journal of the American Statistical Association
Each year, drinking and driving behavior are responsible for at least 24,000 traffic fatalities and cause more than one-half million injuries. Unfortunately, a disproportionate number of these accident victims are young people between age 15 and 24. No other comparable age group has such a record of death and injury on the highway (Koop 1988). Recent marketing research, however, suggests that this social problem might be alleviated with mass media programs based on an understanding of the consumer psychology of young male drinking and driving behavior (Lastovicka, Murry, and Joachimsthaler 1990; Lastovicka, Murry, Joachimsthaler, Bhalla, and Scheurich 1987). Thus this research reports an anti-drinking and driving mass media campaign designed to decrease youthful male drinking and driving behavior.
For this research it is useful to think about two types of mass media campaigns: a donated media campaign and a paid advertising campaign. Most public health mass media campaigns have not been successful; however, they have relied on donated media using so-called public service announcements, or PSA's (Salmon 1989). In general the mass media sell their best time or space to paying advertisers. As a result, PSA's are frequently aired in donated TV time during undesirable time slots that were not sold. This decreases the potential effectiveness of such campaigns, because reliance on donated media makes it impossible to control when, where, how often, and even whether or not targeted populations are exposed to drinking and driving public service announcements. In contrast, a paid advertising campaign provides control over when, where, and how often a targeted population is exposed to the messages.
Furthermore, the limited marketing research and production budgets generally afforded to donated media campaigns decreases the likelihood that effective advertisements will be produced. Again, this is in contrast with what occurs in many paid advertisting campaigns; private sector advertisers routinely invest millions of dollars into developing and executing each advertising campaign, because they consider such investments necessary to develop campaigns to change marketplace behaviors.
Because of the limitations of donated media campaigns, the role that mass media campaigns might play in reducing drinking and driving remains unknown, as a paid advertising campaign to combat drinking and driving has yet to be rigorously evaluated. Therefore, this research examines if a well-funded, anti-drinking and driving advertising campaign targeted at young males can be effective. Although this research briefly reports on the development of the advertising campaign, its focus is on the multicity field experiment used to evaluate its effectiveness.
1. CAMPAIGN DEVELOPMENT
The marketing research reported in Lastovicka et al. (1987) guided the development of the campaign's advertisements. In addition, before final production of the advertisements, ideas for different ads were pretested under laboratory conditions; for example, inexpensive cartoon-like versions of different ideas for television commercials were pretested in so-called "mall intercept" samples of 18- to 24-year-old males (Murry, Lastovicka, and Bhalla 1989). The production and creative quality of the campaign's final-version advertisements was demonstrated by the OMNI advertising award it received.
The advertisements were run in a 6-month paid media schedule using television, radio, newspapers, and billboards. If this paid media schedule at the experimental site were scaled up to a national level, it would be equivalent to a $25 million national campaign.
2. EVALUATION DESIGN
A multiple time series design with both an experimental and a control site was used. In the experimental site (Wichita, Kansas), an anti-drinking and driving advertising campaign was aired March through August 1986. A control site (Omaha, Nebraska) received no advertising campaign treatment during the same period. …