The effect of biological nudges may (also) have been overstated

 There have been a number of recent studies suggesting that the effectiveness of psychological "nudges" may have been substantially overstated. Here's a paper saying something similar about snorting testosterone:

 Dreber, Anna, Magnus Johannesson, Gideon Nave, Coren L. Apicella, Shawn N. Geniole, Taisuke Imai, Erik L. Knight et al. "Investigating the effects of single-dose intranasal testosterone on economic preferences in a large randomized trial of men." Proceedings of the National Academy of Sciences 122, no. 39 (2025): e2508519122. 

Abstract: "There is conflicting evidence on whether testosterone affects economic preferences such  as risk taking, fairness, and altruism, with most evidence coming from correlational  studies or small testosterone administration studies. To credibly test this hypothesis,  we conducted a large-scale, preregistered, double-blind randomized controlled trial with  1,000 male participants—10 to 20 times larger than typical prior studies. Participants  were randomly allocated to receive a single dose of either placebo or intranasal testosterone, and carried out a series of economic tasks capturing social preferences, competitiveness, and risk preferences. We find no evidence of a treatment effect for any of  our nine primary outcome measures, and no strong evidence of an association between  basal salivary testosterone and economic preferences within men. These results fail to  conceptually replicate previous high-impact publications reporting positive findings in  smaller samples, calling into question the idea that short-term testosterone fluctuations  are important drivers of men’s economic preferences. Our results do not rule out the  possibility that different effects might emerge under alternative dosages, administration  protocols, or task timings, or that behavioral effects differ between men and women. The  potential for developmental or long-term effects of testosterone also remains an open  question for future research, though such effects are ethically challenging to investigate  experimentally in humans." 

 

"The number  of participants in the studies reporting any statistically significant effects has ranged from N = 24 ( 18 ) to N = 118 ( 14 ). It is well  known that small, underpowered studies increase the risk that findings reported as statistically significant are false positives..."

...

"We fail to find evidence of a treatment effect of a single-dose of 
intranasal testosterone on any of our eight primary outcome measures or the hypothesized interaction effect for the ninth primary outcome measure. The 99.5% CI can be used to interpret which  effect sizes in the hypothesized direction we find strong evidence  against, see Figs. 1 – 3 . For our eight main effects primary hypotheses, we find strong evidence against effect sizes between about 0.15 Cohen’s d units (Investor Value and Loss Aversion) and 0.26 Cohen’s d units (Risk Aversion), which are considered small effect  sizes. We thus find strong evidence against the hypothesis that  single-dose intranasal testosterone administration has important  effects on economic preferences or behavior in men for all outcome measures in our study.

...

" Our study can be considered a highly powered conceptual replication of several previous results reported in high-impact journals, with a 10 to 20 times larger sample size than most previous  randomized controlled studies. However, our study does not constitute a direct replication of any specific previous study as we did  not base our design on one individual previous study ( 66 ). Our  study fails to conceptually replicate the following previous findings  about treatment effects of testosterone: that testosterone increases  offers in the ultimatum game ( 13 ); that testosterone decreases  offers in the ultimatum game ( 12 ); that testosterone decreases trust  in the trust game ( 16 ); that testosterone increases trustworthiness  (backtransfers) in the trust game ( 16 ), and that testosterone  increases risk taking (17). We furthermore fail to conceptually replicate previous correlational results that testosterone is positively correlated with economic risk-taking ( 26 , 29 , 67 ), that  testosterone is positively correlated with generosity in the dictator  game ( 30 ), and that testosterone is positively correlated with the  rejection of unfair offers in the ultimatum game ( 25 )."