| Popis: |
In the standard, extensively studied misinformation paradigm, one original detail is followed by a maximum of one misleading detail. For example, a car that was a Toyota in the original event is described by a witness as a Volvo. In Blank et al. (2022), we studied, for the first time, what happens when there are two misleading details concerning the same original detail. So, in the example above, one witness could describe an original Toyota as a Volvo, and another as a Skoda. We found that double, compared to single, misinformation can either increase or decrease the misinformation effect, depending on circumstances (see below). We attributed this to two possible mechanisms operating in the double misinformation situation. Firstly, presenting multiple misleading details provides multiple chances to be misled; therefore, double misinformation can increase the misinformation effect. On the other hand, multiple misleading details invite noticing contradictions between them, making it more probable that participants act as if warned, and are therefore less likely to report misinformation. Looking at the data, we concluded (Blank et al., 2022) that the multiple chances to be misled mechanism dominates when misinformation availability is low (clearly below 50% in our analysis), whereas, in conditions of high misinformation availability (clearly above 50%), the second mechanism (noticing contradictions) dominates. To directly test this prediction, and investigate the role of misinformation availability, we propose to conduct another experiment, directly manipulating misinformation availability. Therefore, the present experiment aims to investigate the effects of double misinformation compared to standard (single) misinformation under conditions of low and high misinformation availability. To create conditions of low and high misinformation availability, we will (a) give participants less vs. more time to read and compare two witness statements (which are used to transport the double misinformation), (b) interrupt vs. not interrupt them while doing so, (c) give them instructions to focus on the emotional tone vs. the content details of the witness statements, and (d) present the two witness statements either simultaneously or with a temporal distance of five minutes. We will check the success of these manipulations using (a) participants’ ratings as to whether they had enough time to read the statements, and (b) the percentages of discrepancy detection (between the two pieces of the double misinformation). A successful manipulation with misinformation availability shall produce higher levels of discrepancy detection in conditions of high misinformation availability than in conditions of low misinformation availability. Tests of the following hypotheses will only fully make sense once these assumptions are reasonably met. We might also use the size of the misinformation effect in the standard condition as a proxy of misinformation availability (see Blank et al., 2022, for details). This analysis will also tell us if MA in the two groups was (roughly) symmetrical around 50% (which would be the ideal outcome). |
