Search

This registration includes a preregistration and survey materials for a study on differences in existential belief systems.

This registration includes a PDF of the preregistration as well as the survey materials for a study on morally motivated reasoning.

This registration includes a preregistration and the analysis code that will be used on the data.

Study Name: Rationalization and Outcome Awareness (Study 4) 1) Have any data been collected for this study yet? No data have been collected yet. 2) Hypothesis: What’s the main question being asked or hypothesis being tested? When we rationalize a negative state of affairs in the world, how aware are we that we have done this? That is, when we change our beliefs in order to make an unwanted state of affairs in the world more palatable and consistent with our other beliefs and desires, are we aware that we have engaged in this cognitive process, or do we think our new beliefs are the ones we have held all along? We predict that people have little awareness of having engaged in rationalization. 3) Give a brief overview of the methods. American Prolific Academic workers will begin by reporting their attitudes towards three social issues—one critical issue (the wealth of political leaders) and two filler issues (space exploration and social media). For each issue, all participants will respond to three items measuring attitudes (i.e., 9 items total). After this, they will read a news article informing them that the typical congressperson is much wealthier than the typical American, that this has been true for at least the past century, and that this state of affairs is unlikely to change. Following the articles, participants will again report their attitudes about the target issue and the two filler issues, using the same items from the beginning of the study. Participants will then try to recall their original attitudes on the target issue and on the two filler issues, by reporting how they had responded to each of the items they rated first, prior to the articles. To ensure they try hard to report their most accurate memory, they will learn that good performance on this memory task can earn them extra money. Participants will then report using explicit items whether and in what direction they changed their attitudes toward each of the topics in response to the news articles. Finally, participants will complete the System Justification Scale (Kay & Jost, 2003). 4) Variables: We have not divided these into dependent and independent variables, since this study is correlational. Attitude Variables, including Rationalization: Rationalization will be operationalized as the degree to which attitudes toward political leaders being wealthy become more positive after reading the news articles. Specifically, for each participant, we will calculate their pre-article attitude (mean of items rated prior to reading the news articles), and their post-article attitude (mean of items rated after reading the news articles) toward the target issue. All items will be scored such that agreement with the item indicates more positive attitudes towards wealthy political leaders. That is, all items that represent negative attitudes toward wealthy political leaders will be reverse scored. To capture rationalization, we will compute the difference score scored such that more positive scores mean more rationalization. Recall Errors: Following the same strategy, we will first calculate participants recalled attitude (mean of the items rated during the memory task). We will then compute the difference between their recalled attitude and their pre-article attitude. On this index, scores greater than 0 mean that participants mis-remembered their initial attitude as being more positive than it actually was; scores below 0 mean that participants mis-remembered their initial attitude as being more negative than their initial attitude. Perceived Attitude Change: To capture participants’ self-reports of having changed their attitudes for each issue, they will report their level of (dis)agreement with an item that explicitly asked them if reading the news article changed the way they thought about wealthy political leaders, and in what direction it changed their mind if it did. Individual Differences in System Justification: Individual differences in system justification will be captured by averaging scores on the System Justification Scale. 5) Mediator variables: Describe any variables you expect to mediate the relationship between your IV’s and DV. Specify how they will be measured. N/A 6) Moderator variables: Describe any variables you expect to moderate the relationship between your IV’s and DV. Specify how they will be measured. N/A 7) Analyses: Describe what analyses (e.g., t-test, repeated-measures ANOVA) you will use to test your main hypotheses. We will use three tests to determine if people are aware of how much they rationalized. First, we will test whether people can accurately self-report the degree to which they have changed their attitudes. To do so, we will test whether participants’ Perceived Attitude Change scores are correlated with their Rationalization scores. A positive correlation would indicate that to some degree, people who rated wealthy political leaders as more positive after reading about their prevalence (as captured by the Rationalization variable) also explicitly reported that they changed their mind to view wealthy political leaders more positively (as captured by the Perceived Attitude Change variable). Second, we will test whether people’s memory for their initial attitudes are biased by their new, post-article attitudes. To do so, we will conduct a Chi-square test based on two categorical variables: First, based on their Rationalization score, we will label participants as positive changers, negative changers, or non-changers. Second, based on their Recall Error score, we will label participants as positively biased, negatively biased, or unbiased (i.e., perfectly accurate). The chi square will test whether these variables are independent, or contingent on each other such that people who changed their minds in a particular direction tend to be biased in that direction when recalling their prior attitude. (Note: We could also test this relationship by merely examining the correlation between rationalization and recall error; however this test seems too liberal, given that rationalization and recall error will both by necessity be highly negatively correlated with the initial attitude variable) Third, we will test how much people’s memory for their initial attitudes is linked with their actual initial attitudes, and compare that with how well it is linked with their new, post-article attitudes. To do so, we will use regression to predict recalled attitudes using initial attitudes and post-article attitudes, simultaneously. In this model, accuracy is indexed by the coefficient for initial attitudes, and bias is indexed by the coefficient for post-article attitudes; we will compare their magnitudes. 8) More analyses. Are there any secondary analyses you plan to conduct? (e.g., order or gender effects). To ensure that our measure of rationalization is indeed motivated by a desire to justify the system, we will correlate attitude change on the target issue with scores on the System Justification Scale. 9) Sample. Where and from whom will data be collected? How will you decide when to stop collecting data (e.g., target sample size based on power analysis, set amount of time)? If you plan to look at the data using sequential analysis, describe that here. Participants will be recruited from Prolific Academic. All participants will live in America, have reported to Prolific that they are fluent in English, and have a 95% approval rate for at least 100 jobs on Prolific. Data collection will end when we reach the target sample size of 300 participants after the exclusion criteria below. 10) Exclusion criteria: Who will be excluded (e.g., outliers, participants who fail manipulation check, demographic exclusions)? Participants will be excluded if they fail to select the correct responses (bolded) for the following questions… English Language Comprehension Questions: Specifically, if participants fail any one of the following two comprehension questions, they will be automatically taken to the end of the survey, and will not take part in the study. 1. Please select the response choice below that CORRECTLY fills the blank. What time ____ for the ceremony? a. will she be leaving b. she will be leaving c. she going to be leaving d. she leaving going to be 2. Which word can CORRECTLY go in the space? Mrs. Thompson always works _____. a. wellness b. fastish c. carefully d. upsetly Other Exclusion Questions: Participants who fail any of these following questions will still complete the full survey. 1. 4. How does the net worth of the average congressperson compare to that of the average American citizen of the same age group? a. The net worth of the average congressperson is 3 times more than that of the average American their same age. b. The net worth of the average congressperson is 4 times more than that of the average American their same age. c. The net worth of the average congressperson is twice as much as that of the average American their same age. d. The net worth of the average congressperson is the same as that of the average American their same age. 2. Research in decision making shows that people, when making decisions and answering questions, prefer not to pay attention and minimize their effort as much as possible. Some studies show that over 50% of people don't carefully read questions. If you are reading this question and have read all the other questions, please select the box marked 'other' and type 'awareness'. Do not select "social issues". Thank you for participating and for taking the time to read through the questions carefully! What was this study about? a. Social Issues b. News c. Memory d. Other (please specify) ____________ (Correct = awareness) 3. During the study, did you believe that the news articles you read were real? It is extremely important for our study that you be both accurate and honest. a. Yes b. No Pre-registration written by (initials): WJ Pre-registration reviewed by (initials): KL

This is the preregistration for the third study.

This registration describes the methods, hypotheses, and analyses for a study on motivated judgements of what is and is not evidence.

Why do people choose their beliefs? Research on this question has been dominated by a Traditional Monist Perspective, assuming that people think reasoning must always be in service of producing unbiased, evidence-based beliefs, embodying Epistemic Value. But recent research hints at the possibility that this may be an unwarranted assumption. People knowingly hold incorrect beliefs (Walco & Risen, 2017), prescribe morally motivated reasoning to others (Cusimano & Lombrozo, 2020), and report not caring that much about Epistemic Value when directly asked (Ståhl, Zaal, & Skitka, 2016; Pennycook, Cheyne, Koehler, & Fugelsang, 2019). Extending this work, I propose a New Pluralist Perspective, arguing that people find it worthwhile to believe in service of non-epistemic goals, embodying other values. Based on a review of the motivated reasoning literature, I propose a non-exhaustive list of three non-epistemic values about believing that people could explicitly endorse: (1) Emotional Value (that beliefs can be valuable by supporting positive emotions), (2) Moral Value (that beliefs can be valuable by supporting a moral agenda), and (3) Affiliative Value (that beliefs can be valuable by supporting meaningful affiliations). In Study 1 (n=456), I develop a self-report scale, the Values about Belief Scale (VBS), to measure endorsement of these values. In Study 2 (n=207), I assess the convergent validity of the Emotional Value subscale, and its relationship with emotionally motivated beliefs. In Study 3 (n=449), I explore how Emotional Value predicts palliative beliefs about the COVID-19 pandemic. Finally, in Study 4 (n=200) I explore how the non-epistemic values predict a classic case of motivate reasoning in action: system justification. Results generally support the New Pluralist Perspective over the Traditional Monist Perspective. I discuss the implications of the New Pluralist Perspective for the study of belief regulation.

The present research explores free will belief as a causal antecedent to self-handicapping. The overarching goal of this work is to increase our understanding of why some individuals faced with a threatening assessment in an important ability domain will increase effort to ensure success, whereas others will engage in self-handicapping. Previous work has established that uncertainty aroused by non-contingent success feedback encourages the use of self-handicapping. We theorize that when faced with uncertainty and potential failure, those with weaker free will beliefs will be more likely to resort to self-handicapping. In other words, that self-handicapping necessitates that an individual feels a lack of free will and control over future performance and outcomes. Thus, we predicted that stronger belief in free will would be negatively associated with self-handicapping. Study 1 (N = 137) used a correlational design and measured trait self-handicapping, belief in free will, and perceptions of control, among other relevant dispositional variables, in an MTurk sample. The results indicated that belief in free will was negatively correlated with trait self-handicapping. Study 2 (N = 132) replicated this same pattern using a correlational design with a college student sample. Study 3 (N = 98) used an experimental design and manipulated belief in free will using an autobiographical memory task in which participants were asked to recall a time in which they experienced control and freedom of choice, or a lack thereof. We predicted that experimentally inducing a weaker free will belief would increase claimed self-handicapping. Results showed that the memory task did not produce differences in belief in free will, making us unable to test our causal hypothesis. However, we did replicate the correlations found in Studies 1 and 2. The free will manipulation and the implications of the correlational data are discussed.