Klusowski, J. Prediction Horizons Influence Expectations of Trend Continuation vs. Reversal. Manuscript in preparation.
People commonly use past trends to make future predictions, either through extrapolation or correction. While previous research has largely viewed this process as retrospective, dependent on past trends, the present research demonstrates that it can also be prospective, dependent on future periods or “prediction horizons.” Specifically, people expect greater trend continuation—even for the immediate future (e.g., next month)—when they are presented with a short horizon (e.g., next month) rather than a long horizon (e.g., next three months). Although contradicting some prior literature, these effects manifest robustly across increasing, decreasing, and binary trends and for both skill- and chance-oriented domains, which typically show opposing patterns. These tendencies seem to emerge because, compared to short horizons, long horizons trigger beliefs that imbalances in trends cannot persist, leading to regressive predictions even for the immediate future. Theoretically, these findings extend work on the hot hand vs. gambler’s fallacy by identifying moderators based on prediction characteristics rather than trend characteristics. They also expand research on near- vs. distant-future forecasting, narrow vs. broad bracketing, and short- vs. long-term planning by suggesting distinct predictions. Practically, the results suggest that, when trends themselves are not malleable, prediction horizons can be an alternative means of influencing post-trend expectations across major forecasting domains, including finance and public policy.
Klusowski, J. Small, D.A., &, Goldenberg, J. Even Number Preference in Quantity Selection. Manuscript in preparation.
We find that people choose even numbers more frequently than odd numbers when selecting quantities to acquire or consume (e.g., choosing how many apples to buy for oneself at a store). Drawing on a mix of data sources—including the Kilts-NielsenIQ consumer panel data and online survey data—we find evidence of this tendency not only in shopping decisions but also in a variety of other quantity selection contexts. We propose that this phenomenon occurs because even numbers are more accessible and complete than odd numbers. This extends previous research findings related to round numbers (e.g., typically those ending in 0 or 5) and shows that disproportionate selection of certain numbers occurs at a more granular level. We discuss implications for understanding and influencing quantity selection.
Banker, M., Klusowski, J. & Zauberman, G. Positive Information is Generalized More than Negative Information When Controlling for Prior Beliefs. Manuscript under review at Management Science.
Information is often generalized to similar objects and situations. For example, having a positive (or negative) experience at a restaurant may lead one to predict having positive (or negative) experiences at similar restaurants. Previous research has shown divergent results on whether people are more likely to generalize positive or negative information without a consistent explanation, largely finding stronger negative generalizations if anything. In five preregistered studies and four supplemental studies (N = 8,645), we provide a unifying framework of when and why positive or negative generalizations are stronger. In many previous experimental contexts, people hold positive prior beliefs (e.g., most restaurants are good), which leads to stronger negative generalizations because (a) measurement in these paradigms allow greater shifts in beliefs in the negative direction and magnify negative generalizations and (b) negative information contrasts with their positive prior beliefs more than positive information. For similar reasons, when people hold negative prior beliefs (e.g., most restaurants are bad), positive generalizations are stronger than negative generalizations. To isolate the true effect independent of priors, we employ a novel paradigm to control for prior beliefs. Across several domains, we demonstrate that positive generalizations are stronger than negative generalizations, all else equal. Evidence suggests this effect is not simply driven by impression management concerns, but reflects a belief that positive information is more diagnostic of other objects than negative information.
Klusowski, J. & Lewis, J. Typical Ranges as Scale-Specific Benchmarks: When and Why Percentages Amplify Relative Magnitudes and Their Differences. Forthcoming in Management Science.
Business managers and policymakers must often communicate magnitudes. Yet, conveying large relative magnitudes without desensitizing people to further increases can be challenging due to diminishing sensitivity to large numbers. In this research, we propose that percentages can help convey such magnitudes both compellingly and discriminably. Specifically, large relative magnitudes (e.g., 500%) and large changes thereof (e.g., from 500% to 600%) appear larger in percentage terms than in alternative expressions. We posit an explanation: percentages typically have values between 0% and 100%, so when percentages and percentage-point differences reach 100% or more, they seem unusually large. This hypothesis is supported by data scraped from the New York Times articles and a series of online experiments employing both management-relevant scenarios and incentive-compatible decisions. Existing theories of magnitude perception are either unable to predict all the results of these studies (e.g., numerosity and unitosity) or must be adjusted to reflect the influence of the typical range of values on the relevant numerical scale (e.g., decision-by-sampling and range-frequency theory). We discuss implications for the theory of magnitude and difference perception and the practice of communicating large magnitudes and changes.
The study materials, data, and code are available upon request.
Policy-makers are increasingly turning to behavioural science for insights about how to improve citizens’ decisions and outcomes. Typically, different scientists test different intervention ideas in different samples using different outcomes over different time intervals. The lack of comparability of such individual investigations limits their potential to inform policy. Here, to address this limitation and accelerate the pace of discovery, we introduce the megastudy—a massive field experiment in which the effects of many different interventions are compared in the same population on the same objectively measured outcome for the same duration. In a megastudy targeting physical exercise among 61,293 members of an American fitness chain, 30 scientists from 15 different US universities worked in small independent teams to design a total of 54 different four-week digital programmes (or interventions) encouraging exercise. We show that 45% of these interventions significantly increased weekly gym visits by 9% to 27%; the top-performing intervention offered microrewards for returning to the gym after a missed workout. Only 8% of interventions induced behaviour change that was significant and measurable after the four-week intervention. Conditioning on the 45% of interventions that increased exercise during the intervention, we detected carry-over effects that were proportionally similar to those measured in previous research. Forecasts by impartial judges failed to predict which interventions would be most effective, underscoring the value of testing many ideas at once and, therefore, the potential for megastudies to improve the evidentiary value of behavioural science.
Previous research suggests that choice causes an illusion of control—that it makes people feel more likely to achieve preferable outcomes, even when they are selecting among options that are functionally identical (e.g., lottery tickets with an identical chance of winning). This research has been widely accepted as evidence that choice can have significant welfare effects, even when it confers no actual control. In this article, we report the results of 17 experiments (N = 10,825 online/laboratory participants) examining whether choice truly causes an illusion of control. We find that choice rarely makes people feel more likely to achieve preferable outcomes—unless it makes the preferable outcomes actually more likely—and when it does, it is not because choice causes an illusion, but because choice reflects some participants’ pre-existing (illusory) beliefs that the functionally identical options are not identical.