Experimental Design
In the first part, half of the subjects are randomized into the mindfulness treatment group, while the other half are placed in the control group. In both groups, subjects will listen to an 8-minute audio recording. The content of the recording varies between the two groups: subjects in the mindfulness group will follow a guided exercise where they are instructed to observe their breath, whereas those in the control group will be instructed to let their minds wander.
In the second part, participants are again randomized into two conditions; 50% are assigned to the high-motivation priming condition, and 50% to the low-motivation priming condition. In each condition the participants will complete a Word Search Task (WST) designed to prime them for either high or low motivation. Participants will be presented with 15 sequences of five words, which they must rearrange into grammatically correct four-word sentences. The WST for the low-motivation priming condition includes words such as "lacked," "gave up," "failed," "unmotivated," and "lazy," while the high-motivation priming condition encounters opposite words like "energetic," "high effort," "interested," "motivated," and "hard-working." Five of the WST sequences are neutral and common to both conditions.
The third part involves a judgment task, where participants are presented with a half-page of text that is framed as an answer to an exam question. The participants are asked to mark the text based on three distinct categories. Each category should be marked on a scale from 0 to 10. An answer key for each category is provided to facilitate the marking. We vary the order of the provided answer keys in the following way: In the high-evaluation anchoring condition, half of the participants are randomly assigned to view first an answer key for a category that suggests a high-evaluation. In the low-evaluation anchoring condition, the other half sees first an answer key for a category that suggests a low-evaluation. The remaining answer key for a category that suggests a mediocre-evaluation is presented to all participants at the second place. See Figure 1 in the supporting materials and documents section for a graphical illustration of the randomization process into the various conditions in the mindfulness treatment and the control group.
The participants are informed that the variable payment for this part of the study is calculated based on a benchmark set by six academics with marking experience. These academics have previously marked the same text according to the answer keys and the benchmark for each category is the mean of the marks (rounded to integers) from the six academics. For each category where the participant's score matches the benchmark, a bonus of 1.50 GBP will be awarded. This means that participants can earn up to 4.50 GBP in bonuses. In addition to this variable payment, all participants will receive a fixed payment of 4.50 GBP for their participation. Payments will be processed via Prolific.
We will empirically test the following hypotheses: 1) a short MBI will on average reduce the impact of anchoring in a judgment task (across two opposing priming conditions); and 2) a short MBI will on average reduce the impact of priming in a judgment task (across two opposing anchoring conditions).
In addition, we aim to identify the role of potential interaction effects by testing the following hypotheses: 3) a short MBI will on average reduce the impact of anchoring in a judgment task in a high-motivation priming condition; 4) a short MBI will on average reduce the impact of anchoring in a judgment task in a low-motivation priming condition; 5) a short MBI will on average reduce the impact of priming in a judgment task in a high-evaluation anchoring condition; 6) a short MBI will on average reduce the impact of priming in a judgment task in a low-evaluation anchoring condition and 7) a short MBI will reduce the compound impact of priming and anchoring in a judgment task.
The hypotheses will be tested by employing various difference-in-difference comparisons. Additionally, we plant to employ bootstrapping methods.
Excluded from analysis are data coming from participants who have not fully completed all tasks, failed to answer the control questions correctly or have unreasonably quick response times.