The panel investigation provided longitudinal data with relevant control procedures to allow for greater specificity of causal relationships than the more descriptively oriented national study allowed. The data came from two waves of personal interviews with a probability sample of 517 adult residents of the greater Buffalo, Denver, and Milwaukee metropolitan areas. Interviews were conducted in September 1979 prior to the start of the campaign and again in April-May 1980 when the campaign was several months old. Within the panel, the overall rate of exposure to the 'Detective Dog' advertisements was somewhat lower than in the national sample (18 percent versus 30 percent); however, the overall pattern of those exposed in the panel did not differ sharply from that of the national cohort. The analysis indicated that persons who were initially concerned about crime and in need of crime prevention information were likelier to report subsequent exposure to the advertisement. Persons' initial feelings and behaviors about prevention, however, were largely unrelated to exposure. Those exposed tended to be from lower social status groups and were heavier attenders to crime-related mass media content. The effects of the campaign on those exposed were examined through various pre-to-post measure change score analyses. Exposure to the advertisement appeared to increase concern about both crime and crime prevention among those who initially saw themselves more at risk. Campaign exposure also notably increased respondents' involvement in various crime prevention activities. These were largely persons who initially saw themselves somewhat less at risk, including larger segments of the working and middle class. Over the sample as a whole, campaign exposure was not found to be associated with changes in respondents' sense of personal responsibility for crime prevention, level of confidence in helping prevent crime, perceived knowledge of prevention techniques, perceived effectiveness of prevention techniques, or increased use of household security devices. Tabular data are provided. (Author summary modified)