During the campaign season we conducted regular tests on samples of American voters in which individuals were asked to view and rate television ads in a pre-test/post-test web-based format. From May until September, samples of voters viewed nineteen advertisements from the presidential campaigns or 527 groups in a multistaged design. First, voters were asked to complete a survey instrument in which they identified their current preference in the presidential election. Second, they watched a campaign commercial three times, rating the ads continuously (through automated devices) in terms of aspects such as interest and believability. Finally, the voters were asked to complete the survey instrument once again in order to gauge any differences from their original positions.

We begin our analysis by examining reaction to the first Swift Boat Vets advertisement, "Any Questions." Table 1 presents the pre-test/post-test results among voters originally expressing a preference for John Kerry in the presidential race.

TABLE 1

Advertisement - Swift Boat Vets #1 "Any Questions"

Pre-test/Post-test Intent to Vote for John Kerry by Party Affiliation

As the table shows, there were some notable differences in the pre-test and post-test responses, particularly among independents who had expressed a preference to vote for John Kerry before the commercial, with a substantial drop among those saying they were definitely or most likely supporting the Democratic candidate. Conversely, over one-quarter of independents who were at least leaning towards Kerry before seeing "Any Questions" were not sure of their support after viewing it (see Figure 1).

While it appears that the first "Swift Boat" ad had a negative impact on support for John Kerry among independent voters, it does not seem to have any significant impact upon support for George Bush. Figure 2 shows no change between the pre- and post-test results.