Back to History

Fields Changed

Registration

Field Before After
Trial Title Gender Bias in Investing Loss Sensitivity, Perceived Inequality, and Financial Participation
Trial Status in_development on_going
Abstract This study delves into the behavioral biases in risk aversion measurements, specifically focusing on differences by gender. Using online platforms like MTurk and Prolific, we aim to gather insights from participants across the United States, Denmark, and India. By presenting them with a series of investment choices and lottery options, we seek to understand the underlying factors that might influence their decisions. An added layer of the experiment involves priming participants with thoughts of authority figures, and assessing if this alters their risk preferences. The overarching goal is to investigate if there is a systematic mismeasurement in risk preferences based on gender and if societal norms across different geographies accentuate this potential discrepancy. The outcomes of this study can provide valuable insights into gender gaps in financial decision-making and inform future policies to address these disparities. This study examines how individual differences in loss sensitivity relate to perceptions of wealth and behavior in incentivized investment decisions. The project uses a two-wave online experiment with US adults recruited through Prolific. We measure individual gain and loss utility parameters using a preference elicitation method following Abdellaoui et al. (2016), along with incentivized investment choices. We then re-invite a subset of participants and use an allocation task with randomized framing, alongside survey modules on perceptions of wealth inequality, perceived financial vulnerability, attitudes toward investing, financial socialization, and financial delegation. The two-wave structure enables within-person analysis linking individual-level preference parameters to subsequent perceptions and behavior. This experimental work is intended to complement descriptive evidence from cross-country surveys on financial inclusion.
Trial End Date December 31, 2023 December 31, 2026
JEL Code(s) D81, G11, J16 D81, G41, G51, J16
Last Published September 04, 2023 06:02 AM May 19, 2026 01:31 AM
Intervention (Public) Participants are presented with a series of investment choices and lottery options. A subset of participants is primed with thoughts of authority figures to assess potential influences on their risk preferences. Participants complete incentivized investment and allocation tasks alongside survey measures of financial perceptions and attitudes.
Intervention End Date December 31, 2023 June 30, 2026
Primary Outcomes (End Points) Change in investment choices after priming. Differences in lottery choices by gender. Measured risk preferences after considering authority figure advice. Individual-level utility parameters for gains and losses, including a loss-aversion measure. Allocation behavior in incentivized investment and allocation tasks. Within-person associations between Wave 1 preference parameters and Wave 2 measures of perceptions, attitudes, and allocation behavior.
Primary Outcomes (Explanation) The primary outcomes are straightforward metrics derived from the participants' choices in the survey. For instance, "change in investment choices" will be quantified by comparing the number of risk-averse choices made before and after the priming intervention. Wave 1 primary outcomes are constructed from individual-level utility elicitation (gain and loss utility curves; ratio-based loss aversion measure) and from choices in the incentivized financial choice list. Wave 2 primary outcomes include allocation to the risky lottery in the framed allocation task and survey-based measures of perceived inequality, perceived vulnerability, attitudes, and socialization. Within-person analyses examine associations between Wave 1 preference parameters and Wave 2 measures.
Experimental Design (Public) This is a randomized controlled trial where participants are exposed to investment choices and lotteries. A random subset undergoes a priming intervention to assess its influence on decision-making A two-wave online experiment. Each wave includes randomized assignment to pre-task or task-frame conditions, combined with individual-level measurement of preferences, perceptions, and behavior.
Planned Number of Observations 700-1000 Wave 1: 500 (completed). Wave 2: approximately 100 (target, drawn from Wave 1 participants via Prolific re-contact).
Sample size (or number of clusters) by treatment arms Observations automatically divided into equal treatment arms Wave 1: approximately 167 per arm across three arms (completed). Wave 2: approximately 50 per arm across two arms (gain frame, loss frame).
Power calculation: Minimum Detectable Effect Size for Main Outcomes Wave 2 sample size is set by the pool of Wave 1 participants available for re-contact rather than by ex ante power calculations. Wave 2 analyses are accordingly best interpreted as exploratory with respect to within-person associations between Wave 1 preference parameters and Wave 2 outcomes.
Additional Keyword(s) Behavioral economics, Gender biases, Risk aversion, Investment choices, Cultural norms, Financial decision-making Behavioral economics, Loss aversion, Perceptions of inequality, Financial inclusion, Prospect theory, Investment behavior, Financial socialization
Intervention (Hidden) We utilize two main interventions: 1. Participants are offered a series of investment choices, informed by historical returns and volatility. They are also given options between different lotteries to gauge their risk preferences, particularly in terms of loss aversion. 2. A random group of participants is primed at the beginning of the survey. They are asked to recall and consider the advice of an authority figure in their life, such as a parent, spouse, or manager, before making their decisions. The intention behind this priming is to uncover potential external influences on financial decision-making behaviors. The study uses a two-wave design implemented in Qualtrics on Prolific. Wave 1: 500 US adults were randomly assigned to one of three pre-elicitation conditions: (a) an information condition in which participants read a passage about stocks versus bonds and answered related questions; (b) a priming condition in which participants were asked to recall an authority figure who advises them on financial matters; (c) a placebo condition in which participants read an unrelated passage about FDIC insurance. All participants then completed an individual-level utility elicitation over gains and losses following Abdellaoui et al. (2016) and an incentivized choice list between a safe asset (1-year US Treasury) and risky financial assets at multiple horizons. Wave 2: A subset of Wave 1 participants is re-invited and randomly assigned to either a gain framing or a loss framing of a Gneezy-Potters allocation task with a five-dollar endowment, fair-coin lottery, and 2.5x multiplier. Participants then complete survey modules on perceptions of wealth inequality, perceived financial vulnerability, attitudes toward investing, financial socialization, financial delegation, and demographics. One in ten Wave 2 participants is randomly selected to receive their bonus payment to preserve incentive compatibility while controlling cost.
Secondary Outcomes (End Points) Differences in risk preferences by country. Influence of provided investment advice on choices. Heterogeneity in primary outcomes across participant subgroups defined by demographics and self-reported financial background.
Secondary Outcomes (Explanation) The secondary outcomes will compare the risk preferences across the three countries (United States, Denmark, India) to understand the cultural impact on decision-making. Additionally, the influence of the provided investment advice will be analyzed to determine its efficacy in guiding participants' choices. Secondary analyses will examine whether relationships among primary outcomes vary systematically across participant characteristics. These analyses are exploratory.
Back to top