Measuring Persuasion Without Measuring a Prior Belief: A New Application of Planned Missing Data Techniques
Abstract
Research on advice taking frequently utilizes a Judge Advisor System paradigm, in which a judge reports a prior belief, receives advice, and then revises their initial estimate. However, recent studies have shown that the cognitive process of advice taking depends on whether one explicitly reports their prior belief or not, indicating that the mere act of measuring a judge’s prior belief functions as a treatment effect on how they utilize advice. This phenomenon is known as measurement reactivity. This research shows that by treating judges’ prior beliefs as missing data, it is possible to use statistical imputation models to estimate how their beliefs change without requiring them to directly report their prior judgment. In two simulation studies and two experimental studies, the feasibility and effectiveness of this planned missing data imputation method are demonstrated. Two variants of the planned missing data research design are discussed: the single variable design, where the imputation model is based on the same type of variable as the behavioral target; and the separate variable design, in which a separate class of auxiliary variables are elicited to fit the imputation model. This research design and associated modeling techniques have potential applications in social science beyond advice taking research. Novel theoretical insights regarding how people take advice when their priors are not explicitly accessed are also described. In a study in which judges were asked to estimate the number of calories in different foods, there were clear differences in how they took advice in the absence of measurement reactivity: they both updated their beliefs more frequently and weighed advice more heavily. In a probability judgment task involving predictions of upcoming baseball games, there were again clear effects of measurement reactivity. However, an existing conceptual model of advice utilization was a poor representation of the results, indicating we may need to rethink notions of advice taking for probability judgment in the absence of measurement reactivity.
Subject Area
Quantitative psychology|Psychology|Cognitive psychology|Behavioral psychology
Recommended Citation
Himmelstein, Mark, "Measuring Persuasion Without Measuring a Prior Belief: A New Application of Planned Missing Data Techniques" (2023). ETD Collection for Fordham University. AAI30820163.
https://research.library.fordham.edu/dissertations/AAI30820163
