Abstract
This investigation explored three issues relevant to inoculation theory’s function. First, the stamina of inoculation-generated resistance was examined at 2-, 4-, and 6-week time intervals between treatment and persuasive attack. Second, inoculation and attack messages as booster treatments were compared. Finally, this study tested the ability of inoculation-generated resistance to withstand three persuasive attacks. Results from a 230-participant five-phase experiment indicate erosion of generated resistance with a longer time delay. However, booster treatments in the form of a second inoculation message demonstrate potential to lengthen the inoculation effect. In addition, generated resistance remained after three persuasive attacks.
Notes
1. Some of the conditions were not relevant to the present study.
2. Due to significant sample disproportion, in each analysis separate tests were performed looking for possible sex differences (main and interaction). None were discovered.
3. An independent sample t-test found no significant effect of using an inoculation booster message, as compared to not using one at all, as inoculated individuals who received a booster in the form of a repeated inoculation message (M = .14, SD = 2.62, n = 31) did not display significantly stronger attitudes toward the issue compared to inoculated individuals who did not (M = −.17, SD = 3.01, n = 29), t(58) = .43, p = .67, despite the fact that pattern of means favored the use of booster.
4. An independent sample t-test indicated that inoculated individuals who received two boosters in the form of a repeated inoculation messages (M = .71, SD = 2.75, n = 26) displayed significantly stronger attitudes toward the issue after the attack compared inoculated individuals who did not receive any (M = −1.57, SD = 2.58, n = 32), t(56) = 3.24, p < .01, η2 = .16.
5. Within-subjects ANOVA effects were computed with the Greenhouse-Geisser adjustment, which corrects sphericity problems. The omnibus results of the repeated subjects ANOVAs, with Greenhouse-Geisser corrections, did not reveal significant attitudinal differences following each attack, F(1.79, 51.82) = .56, p = .56.