The treatment recommendation based on a network meta-analysis (NMA) is usually the single treatment with the highest expected value (EV) on an evaluative function. We explore approaches that recommend multiple treatments and that penalise uncertainty, making them suitable for risk-averse decision-makers. We introduce loss-adjusted EV (LaEV) and compare it to GRADE and three probability-based rankings. We define properties of a valid ranking under uncertainty and other desirable properties of ranking systems. A two-stage process is proposed: the first identifies treatments superior to the reference treatment; the second identifies those that are also within a minimal clinically important difference (MCID) of the best treatment. Decision rules and ranking systems are compared on stylised examples and 10 NMAs used in NICE (National Institute of Health and Care Excellence) guidelines. Only LaEV reliably delivers valid rankings under uncertainty and has all the desirable properties. In 10 NMAs comparing between 5 and 41 treatments, an EV decision maker would recommend 4–14 treatments, and LaEV 0–3 (median 2) fewer. GRADE rules give rise to anomalies, and, like the probability-based rankings, the number of treatments recommended depends on arbitrary probability cutoffs. Among treatments that are superior to the reference, GRADE privileges the more uncertain ones, and in 3/10 cases, GRADE failed to recommend the treatment with the highest EV and LaEV. A two-stage approach based on MCID ensures that EV- and LaEV-based rules recommend a clinically appropriate number of treatments. For a risk-averse decision maker, LaEV is conservative, simple to implement, and has an independent theoretical foundation.

Real-world evidence (RWE) is increasingly used and accepted by health technology assessment (HTA) bodies as supportive evidence to inform the approval of new technologies. However, the criteria driving RWE acceptance are often unclear.

This study aims to improve understanding of the role and value of RWE in HTA decision-making and outline the best practices in building real-world external control arms (ECAs).

A mixed approach of a targeted literature review and HTA expert interviews was applied. The HTA reports of ten selected technologies and the expert interviews from France, Germany, Italy, Spain, and the UK informed the criteria driving the acceptance of RWE. Overall, the UK and Spanish HTA bodies are more receptive to accepting RWE, whereas the French and German are the least accepting. When RWE is used to substantiate efficacy claims, the level of scrutiny from regulators and HTA bodies is considerably higher than when RWE has different intended uses. Representativeness of the data source, overall transparency in the study and robust methodologies are the key criteria driving RWE acceptance across markets.

Using a laboratory experiment, we investigate complexity in decision problems as a cause of failures in contingent reasoning. For this purpose, we introduce three dimensions of complexity to a decision problem: the number of contingencies, the dominance property of choices, and reducible states. Each decision problem is designed to reflect variations in complexity across the three dimensions. Experimental results show that the number of contingencies has the most significant effect on failures in contingent reasoning. The second dimension, the dominance property of choices, also has a statistically significant effect, though the effect size is smaller than in the existing literature. In contrast, the third complexity dimension has no impact; presenting the decision problem in a reduced or reducible form does not change subjects’ performance on contingent reasoning. Additionally, we examine the Power of Certainty and show its existence. This effect is particularly pronounced when the number of contingencies is large.

We measured brain activity using a functional magnetic resonance imaging (fMRI) paradigm and conducted a whole-brain analysis while healthy adult Democrats and Republicans made non-hypothetical food choices. While the food purchase decisions were not significantly different, we found that brain activation during decision-making differs according to the participant’s party affiliation. Models of partisanship based on left insula, ventromedial prefrontal cortex, precuneus, superior frontal gyrus, or premotor/supplementary motor area activations achieve better than expected accuracy. Understanding the differential function of neural systems that lead to indistinguishable choices may provide leverage in explaining the broader mechanisms of partisanship.

Bilinguals use languages strategically and make decisions differently depending on the language context. Here, we explored whether verbal feedback modulates language use and risk-taking in bilinguals engaged in a coin-drawing game that incentivises lying. In the game, participants announced bets in Chinese or English, and feedback on the outcome of the current bet was given in the same language. They selected Chinese over English after receiving positive feedback in Chinese, and no language difference was found when feedback was provided in English. They also tended to take more risks after receiving positive than negative feedback. Furthermore, participants were more likely to switch from one language to the other following negative feedback as compared to positive feedback, and when telling the truth, they were faster after negative than positive feedback. Thus, the language in which bilinguals receive feedback constrains language use, which may have implications for understanding interactions in multilingual communities.

Risk aversion—the preference for certainty over potential gains or losses—is reduced when using a foreign language. We investigated 2 mechanisms for this foreign language effect using incentivized gambles with verbal probability expressions: (1) that people perceive prospects of winning as larger when a decision is made in their foreign language; (2) that people experience reduced negative affect toward risk in a foreign language. In Experiment 1, N = 229 proficient Polish–English bilinguals, using ungridded slider, assigned numerical values to 29 verbal probability expressions in both languages. We found small bidirectional differences in 13 of them, leading us to reject the first mechanism. In Experiment 2, N = 281 participants gambled in incentivized neutral expected value lotteries using a sample of the verbal probability expressions from Experiment 1. Participants gambled in either their native or foreign language, where participants could either accept around 50% of gambles (debiased to risk-neutral) or more than 50% (biased to risk-seeking). Surprisingly, we observed no significant risk aversion in either language condition, with participants’ gambling behavior close to 50% in both cases. The finding that participants showed no risk aversion in native language condition meant we could not test whether foreign language reduces risk aversion. However, this result did show that using a foreign language does not promote excessive risk-taking. Our findings suggest that using verbal probability expressions does not bias participants’ responses, and may inherently reduce risk aversion.

We develop a new design for the experimental beauty-contest game (BCG) that is suitable for children in school age and test it with 114 schoolchildren aged 9–11 years as well as with adults. In addition, we collect a measure for cognitive skills to link these abilities with successful performance in the game. Results demonstrate that children can successfully understand and play a BCG. Choices start at a slightly higher level than those of adults but learning over time and depth of reasoning are largely comparable with the results of studies run with adults. Cognitive skills, measured as fluid IQ, are predictive only of whether children choose weakly dominated strategies but are neither associated with lower choices in the first round nor with successful performance in the BCG. In the implementation of our new design of the BCG with adults we find results largely in line with behavior in the classical BCG. Our new design for the experimental BCG allows to study the development of strategic interaction skills starting already in school age.

In today’s world, smart algorithms—artificial intelligence (AI) and other intelligent systems—are pivotal for promoting the development agenda. They offer novel support for decision-making across policy planning domains, such as analysing poverty alleviation funds and predicting mortality rates. To comprehensively assess their efficacy and implications in policy formulation, this paper conducts a systematic review of 207 publications. The analysis underscores their integration within and across stages of the policy planning cycle: problem diagnosis and goal articulation; resource and constraint identification; design of alternative solutions; outcome projection; and evaluation. However, disparities exist in smart algorithm applications across stages, economic development levels, and Sustainable Development Goals (SDGs). While these algorithms predominantly focus on resource identification (29%) and contribute significantly to designing alternatives—such as long-term national energy policies—and projecting outcomes, including predicting multi-scenario land-use ecological security strategies, their application in evaluation remains limited (10%). Additionally, low-income nations have yet to fully harness AI’s potential, while upper-middle-income countries effectively leverage it. Notably, smart algorithm applications for SDGs also exhibit unevenness, with more emphasis on SDG 11 than on SDG 5 and SDG 17. Our study identifies literature gaps. Firstly, despite theoretical shifts, a disparity persists between physical and socioeconomic/environmental planning applications. Secondly, there is limited attention to policy-making in development initiatives, which is critical for improving lives. Future research should prioritise developing adaptive planning systems using emerging powerful algorithms to address uncertainty and complex environments. Ensuring algorithmic transparency, human-centered approaches, and responsible AI are crucial for AI accountability, trust, and credibility.

Effect modification occurs when a covariate alters the relative effectiveness of treatment compared to control. It is widely understood that, when effect modification is present, treatment recommendations may vary by population and by subgroups within the population. Population-adjustment methods are increasingly used to adjust for differences in effect modifiers between study populations and to produce population-adjusted estimates in a relevant target population for decision-making. It is also widely understood that marginal and conditional estimands for non-collapsible effect measures, such as odds ratios or hazard ratios, do not in general coincide even without effect modification. However, the consequences of both non-collapsibility and effect modification together are little-discussed in the literature.

In this article, we set out the definitions of conditional and marginal estimands, illustrate their properties when effect modification is present, and discuss the implications for decision-making. In particular, we show that effect modification can result in conflicting treatment rankings between conditional and marginal estimates. This is because conditional and marginal estimands correspond to different decision questions that are no longer aligned when effect modification is present. For time-to-event outcomes, the presence of covariates implies that marginal hazard ratios are time-varying, and effect modification can cause marginal hazard curves to cross. We conclude with practical recommendations for decision-making in the presence of effect modification, based on pragmatic comparisons of both conditional and marginal estimates in the decision target population. Currently, multilevel network meta-regression is the only population-adjustment method capable of producing both conditional and marginal estimates, in any decision target population.

This paper derives career lessons for economists by examining the unusual career of Brian J. Loasby in terms of Loasby’s own key ideas. Loasby’s early career ran into difficulties after he chose to become a postgraduate for lifestyle rather than career reasons, with his insufficiently ambitious reference standards resulting in unduly narrow search for both options and new connections. His eventual success came because of a path-dependent process whereby he developed an organizing framework that was conducive to finding problems and making intellectual connections. It was a process in which making personal connections also had a key role, though some intellectual connections that he earlier failed to make resulted from trusting other people. In the market for contributions to knowledge, his way of branding and positioning his work in relation to management decision-making and the growth of knowledge gave it a wider appeal than would have been the case if he had aligned himself with a particular school of thought or ‘heterodox economics’ in general.

Online experiments allow researchers to collect datasets at times not typical of laboratory studies. We recruit 2336 participants from Amazon Mechanical Turk to examine if participant characteristics and behaviors differ depending on whether the experiment is conducted during the day versus night, and on weekdays versus weekends. Participants make incentivized decisions involving prosociality, punishment, and discounting, and complete a demographic and personality survey. We find no time or day differences in behavior, but do find that participants at nights and on weekends are less experienced with online studies; on weekends are less reflective; and at night are less conscientious and more neurotic. These results are largely robust to finer-grained measures of time and day. We also find that those who participated earlier in the course of the study are more experienced, reflective, and agreeable, but less charitable than later participants.