Do the protective effects last? The effectiveness of alternative + forewarning/inoculation techniques in reducing misinformation reliance and reliance regression in the continued influence effect procedure

Abstract

The continued influence effect (CIE) refers to continued reliance on misinformation, even after it has been retracted. There are several techniques to counter it, such as forewarnings or presenting alternative explanations that can replace misinformation in knowledge or mental models of events. However, the existing research shows that they generally do not eliminate CIE, and their protective effects do not appear to be durable over time. In two experiments (N = 441), we aimed to investigate the effectiveness of the alternative explanation technique and a combination of an alternative explanation and a forewarning (Experiment 1) or inoculation (Experiment 2) in both reducing CIE and the effect of increasing misinformation reliance over time, which is called belief regression. We found that an alternative reduced CIE while combining it with a forewarning or inoculation boosted this protective effect in the pretest. Nevertheless, the protective effect of the alternative + forewarning and inoculation techniques was not sustained, as shown by the fact that misinformation reliance increased for over 7 days, despite continued memory of the correction. A similar pattern, albeit with mixed evidence from Null Hypothesis Significance Testing (NHST) versus Bayesian analyses, was found for the alternative + inoculation technique. In the ‘Discussion’ section, we address issues such as the potential cognitive mechanisms of this effect. Despite all the similarities, given the difference in both methodology and results, we proposed that increased misinformation reliance over time in inferential reasoning should be attributed not to belief regression but to a phenomenon we refer to as reliance regression.