Randy Yang Gao

Accurately measuring consumer valuation of products and services is crucial for social and management science researchers. However, commonly used value elicitation methods do not always produce the same estimate of consumer valuation (willingness-to-pay). Across ten preregistered, incentive-compatible studies (total N = 6,830), we document a robust difference between two commonly used incentive-compatible value elicitation methods that are theoretically equivalent: the Multiple Price List method (MPL) and the Becker-DeGroot-Marschak method (BDM). WTP elicited under the MPL method is substantially lower than WTP under the BDM method. This difference is robust to several design features of the MPL method, different product categories and price points, as well as participant populations. Because the MPL method significantly lowers people's WTP for products, using this method to assess consumer valuation may systematically underestimate market demand. The use of the MPL method can also lower valuation to a larger extent in some other real-world effects on consumer valuation, increasing the risk of false negatives for these effects in some research contexts.

Societies routinely underprepare for low-probability, high-consequence threats despite widespread risk aversion. We identify a hedging paradox: risk aversion can paradoxically discourage hedging when policies are judged by their impact (incremental lives saved) rather than by final states (total survivors). Because a hedge often has no observable impact if the threat does not materialize, it can appear “risky” when evaluated through an impact lens, steering risk-averse decision-makers away from preparedness. Across five preregistered studies (N = 5,077)—including U.S. elected officials, legal professionals, and incentivized decision-makers—we demonstrate that framing policies in terms of their potential impact systematically reduces support for existential risk mitigation compared to framing them in terms of final societal outcomes. This effect persists across threat domains and operates independently of gain/loss framing. We also identify institutional drivers perpetuating this impact focus: legal professionals and potential voters preferentially evaluate policies based on realized impact, creating accountability pressures that discourage prudent hedges. Testing potential remedies for this impact focus, we found that a brief prompt to consider final states rather than personal impact increased support for hedging among sophisticated populations but not the general public. Together, these findings show how cognitive and institutional forces undermine preparedness and suggest a practical remedy: present hedging policies in terms of the final states they secure, not the incremental impact they might (or might not) realize.

Western-centered (primarily American-centered) samples have long faced criticism, yet limited progress has been made to address this issue. The present research validates data from an Eastern culture—China—offering a promising solution. We conducted 16 preregistered studies with 24,932 Chinese and American online participants, examining 21 well-established phenomena across various social science disciplines. The findings provide strong evidence that Chinese online platforms offer geographically and demographically diverse samples from across China. These participants are highly attentive and produce valid data. Additionally, we find tentative evidence that language is a surmountable barrier for Western researchers, as translation variations had little impact on Chinese participants’ responses in our tests. Finally, we share practical insights on using Chinese online platforms for data collection. We hope this paper encourages researchers to embrace diverse samples by trying them in their own research and appreciating it in the research of others.