'Uncongeniality meets Metacognition - Part 1' (AsPredicted #59,948)
Author(s) This pre-registration is currently anonymous to enable blind peer-review. It has 2 authors.
Pre-registered on 2021/03/04 01:55 (PT)
1) Have any data been collected for this study already? No, no data have been collected for this study yet.
2) What's the main question being asked or hypothesis being tested in this study? The study investigates the influence of prior knowledge and metacognitive assessments of one’s own knowledge on a) the tendency to reply to online forum discussion posts varying in congeniality, and b) subsequent belief polarization. Hypotheses:
1) Participants indicate a stronger tendency to reply to an uncongenial discussion post than to a congenial discussion post
2) Confidence affects the general tendency to reply in a way that high confidence increases the tendency to reply
3) Confidence moderates the effect of belief stance on the tendency to reply in a way that higher confidence leads to a more pronounced uncongeniality
4) Prior knowledge affects beliefs in a way that low prior knowledge is associated with stronger polarization (a shift away from the neutral scale mid-point)
3) Describe the key dependent variable(s) specifying how they will be measured. The tendency to reply for each discussion post will be measured by a participant’s response to the following question (translated): “How much would you like to reply to this post?” (6-point Likert scale).
Beliefs are measured before and after exposure to posts by combining responses to five questions with 17-point Likert scales (from -8 to +8).
Polarization is defined as a belief change of post-measures that (in relation to pre-measures) move away from the scale mid-point.
4) How many and which conditions will participants be assigned to? Participants are confronted with 12 discussion posts (4 arguing for human-caused climate change, 4 neutral, and 4 arguing against human-caused climate change). Depending on the belief pre-measures of participants, the belief stance of each post can be categorized as congenial, b) neutral, or c) uncongenial with respect to the participant.
Knowledge about climate change will be measured by participants’ individual true/false responses with regard to eight statements about climate change.
Confidence in a person’s knowledge will be measured by participants’ individual responses to the following question: How certain are you that your response [pertaining to the true/false statement] is correct? Responses are captured on a 6-point Likert scale ranging from 50% = guessed to 100% = absolutely certain. From the responses, d’ (hit rate minus false alarm rate) will be computed as unbiased measure of knowledge, and meta_d’ (Maniscalo & Lau, 2014) will be computed as unbiased measure of confidence.
Analyses of response tendency are based on a 3-factorial mixed design, with belief stance (manipulated) as within-subjects variable (congenial vs. uncongenial), knowledge (non-manipulated) as between-subjects variable, and confidence (non-manipulated) as between-subjects variable.
Analyses of beliefs and polarization) are based on a 3-factorial mixed design, with time point (before vs. after exposure to discussion posts) as within-subjects variable, knowledge (non-manipulated) and confidence (non-manipulated) as between-subjects variables.
5) Specify exactly which analyses you will conduct to examine the main question/hypothesis. Analyses will be conducted by applying:
(a) Linear mixed-effect model with DV: response to post, IV: conflict, confidence, knowledge.
(b) Linear regression with DV: polarization, IV: knowledge, confidence.
Neutral posts will be excluded for the main analysis.
6) Describe exactly how outliers will be defined and handled, and your precise rule(s) for excluding observations. Participants will be excluded from the analysis if one of the following statements apply:
1) A standard deviation of zero in the responses to the belief questionnaire.
2) A combined belief score of zero in the first belief questionnaire
7) How many observations will be collected or what will determine sample size? No need to justify decision, but be precise about exactly how the number will be determined. We are aiming to sample N = 200 valid cases. To account for data exclusions based on the below described criteria, we will test N = 210 participants (i.e., oversampling of 5%).
8) Anything else you would like to pre-register? (e.g., secondary analyses, variables collected for exploratory purposes, unusual analyses planned?) Further analyses:
1) The main hypothesis will also be tested by applying a Bayesian Linear Mixed Models approach
2) For each post, participants have to indicate their "like" or "dislike". It will be examined in how far the hypothesized relation between conflict and response tendency is mediated by disliking a particular post.
3) Additional analyses will assess whether the uncongeniality bias also occurs when participants with a strong belief are exposed to neutral posts, and when participants with a moderate belief are exposed to extreme posts in either direction.
4) Calculation of metacognitive efficiency Mratio = meta-d’/d’ and exploratory analysis to assess whether there is a relationship with response to post and polarization.
