Does blatantly contradictory information reduce the misinformation effect? A Registered Report replication of Loftus (1979)

Abstract

AbstractPurposeLoftus (1979, Reactions to blatantly contradictory information) demonstrated that participants who received a piece of blatantly contradictory information were not susceptible to it (the boundary condition effect). In addition, participants who had received the blatant misinformation were also less susceptible to the more subtle pieces of misinformation compared to the participants who did not receive the blatant misinformation (the inoculation effect). Researchers often cited this finding as evidence for the idea that detecting a piece of blatant misinformation is similar to being given a warning. However, to our knowledge, the inoculation effect has not been replicated.MethodsExperiment 1 was a direct replication. Participants either encountered four pieces of nonblatant misinformation (subtle condition) or the same four pieces of nonblatant misinformation and an additional piece of blatant misinformation (blatant condition). Experiment 2 was a conceptual replication in which we also manipulated the presence or absence of initial testing. Furthermore, we presented the slides as a video and presented four neutral and four misleading details in the narrative.ResultsIn two high‐powered experiments, we failed to replicate both the inoculation effect and the boundary condition effect. Using a multi‐verse analysis approach, we failed to observe a significant advantage from the inclusion of a piece of blatant misinformation in most (but not all) analyses.ConclusionWe conclude that encountering a piece of blatant misinformation was not sufficient to inoculate witnesses against other nonblatant pieces of misinformation. We further discuss our results in the context of misinformation, warning, and discrepancy detection.