Fig 5
| Forecasting errors by prediction approach. The estimated means and 95% CIs are based on a restricted information maximum likelihood linear mixed-effects model with model type (data-driven, hybrid or intuition/ theory-based) as a fixed-effects predictor of the log(MASE) scores, domain as a fixed-effects covariate and responses nested in participants. We ran separate models for each tournament (first: N groups = 86, N observations = 359; second: N groups = 120, N observations = 546). Scores below the dotted horizontal line show better performance than a naive in-sample random walk. Scores below the dashed horizontal line show better performance than the median performance in M4 tournaments 7 .
Source publication
How well can social scientists predict societal change, and what processes
underlie their predictions? To answer these questions, we ran two
forecasting tournaments testing the accuracy of predictions of societal
change in domains commonly studied in the social sciences: ideological
preferences, political polarization, life satisfaction, sentiment...
Contexts in source publication
Context 1
... Article https://doi.org/10.1038/s41562-022-01517-1 data modelling (but no consideration of subject matter theories) and (3) hybrid approaches. Roughly half of the teams relied on data-based modelling as a basis for their forecasts, whereas the other half of the teams in each tournament relied only on their intuitions or theoretical considerations (Fig. 5). This pattern was similar across domains ( Supplementary Fig. ...
Context 2
... analyses with approach as a factor, domain type as a control dummy variable and MASE scores nested in forecasting teams as a dependent variable revealed that forecasting approaches significantly differed in accuracy (first tournament: F(2, 149.10) = 5.47, P = 0.005, R 2 = 0.096; second tournament: F(2, 177.93) = 5.00, P = 0.008, R 2 = 0.091) (Fig. 5). Forecasts that considered historical data as part of the forecast model ling were more accurate than models that did not (first tournament: F(1, 56.29) = 20.38, P < 0.001, R 2 = 0.096; second tournament: F(1, 159.11) = 8.12, P = 0.005, R 2 = 0.084). Model comparison effects were qualified by a significant model type × domain ...
Context 3
... apart from a few domains concerning racial and gender-career bias, scientists' original forecasts were typically not much better than naive statistical benchmarks derived from historical averages, linear regressions or random walks. Even when we confined the analysis to the top five forecasts by social scientists per domain, a simple linear regression produced less error roughly half of the time ( Supplementary Figs. 5 and 9). ...
Similar publications
Battery systems are increasingly being used for powering ocean going ships, and the number of fully electric or hybrid ships relying on battery power for propulsion and maneuvering is growing. In order to ensure the safety of such electric ships, it is important to monitor the available energy that can be stored in the batteries, and classification...
Citations
... To our knowledge, there is currently no comprehensive study eliciting knowledge for the crucial topic of risk Who can predict farmers' choices in risky gambles? 3 preferences in European agriculture (Iyer et al. 2020 ) which would provide insights into system-specific expertise and the potential impact of financial incentives on prediction improvement. Such knowledge would enhance our understanding of more specific expertise beyond social scientists' ability to predict social phenomena (Grossmann et al. 2023 ) and help farmers identify reliable sources of advice (Wuepper et al. 2021 ;Rust et al. 2022 ). ...
... Extending this research to investigate the forecasters' ability to predict the outcomes of individual farmers or smaller, more homogeneous groups of farmers would allow us to better understand whether it is task comprehension or the sample that causes low accuracy. Additionally, incorporating quantitative data on past behavior of farmers or farmer groups could enhance our understanding of the differences between intuition-driven and data-driven forecasts (Grossmann et al. 2023 ). Furthermore, providing a brief summary of the research results to all respondents offers an opportunity to steer interest in the results (Höhler et al. 2024 ) or to examine whether forecasters update their beliefs after participating in multiple predictions when receiving feedback (Vivalt and Coville 2023 ). ...
Risk is a pervasive factor in agriculture and a subject of great interest to agricultural economists. However, there is a lack of comprehensive understanding of the knowledge held by farm advisors, students, and economists with regards to farmers' risk preferences. Misconceptions about farmers’ willingness to take risks could lead to misguided advice. This study builds upon a recent multinational endeavor that employed a multiple price list to assess risk preferences among European farmers. We expand this research by gathering predictions for farmers’ risk preferences from 561 farm advisors, students, and economists. Our objectives are threefold: firstly, we explore variations as to how accurately participants can predict risk preferences in different specializations; secondly, we compare the predictive accuracy of different groups of forecasters; and thirdly, we assess whether modifying incentive mechanisms can improve the accuracy of predictions. Whereas our findings reveal substantial variation in individual predictions, the averages closely align with the observed responses of farmers. Notably, the most accurate predictions were provided by a sample of experimental economics researchers. Furthermore, predictions for different production systems exhibit minimal disparities. Introducing incentive schemes, such as a tournament structure, where the best prediction receives a reward, or a high-accuracy system, where randomly selected participants are compensated for the accuracy of their predictions, does not significantly impact accuracy. Further research and exploration are needed to identify the most reliable sources of advice for farmers.
JEL-Codes: Q12, Q16, C91
Cryptocurrencies have ballooned into a billion-dollar business. To inform regulations aimed at protecting consumers vulnerable to suboptimal financial decisions, we investigate crypto investment intentions as a function of consumer gender, financial overconfidence (greater subjective versus objective financial knowledge), and the Big Five personality traits. Study 1 (N = 126) found that people believe each Big Five personality trait as well as consumer gender and financial overconfidence to predict consumers’ crypto investment intentions. Study 2 (N = 1,741) revealed that less than 1 in 10 consumers from a nationally representative sample (Norway) are willing to invest in crypto. However, the proportion of male (vs. female) consumers considering such investments is more than twice as large, with less (vs. more) agreeable, less (vs. more) conscientious, and more (vs. less) open consumers also being increasingly inclined to consider crypto investments. Financial overconfidence, agreeableness, and conscientiousness mediate the link between consumer gender and crypto investment intentions. These results hold after accounting for a theoretically relevant confounding factor (financial self-efficacy). Together, this research offers novel implications for marketing theory and practice that help understand the observed gender differences in consumers’ crypto investments.
Social networks can provide insights into the emotions expressed by a society. However, the dynamic nature of emotions presents a significant challenge for policymakers, politicians, and communication professionals who seek to understand and respond to changes in emotions over time. To address this challenge, this paper investigates the frequency, duration, and transition of 24 distinct emotions over a 2-year period, analyzing more than 5 million tweets. The study shows that emotions with lower valence but higher dominance and/or arousal are more prevalent in online social networks. Emotions with higher valence and arousal tend to last longer, while dominant emotions tend to have shorter durations. Emotions occupying the conversations predominantly inhibit others with similar valence and dominance, and higher arousal. Over a month, emotions with similar valences tend to prevail in online social network conversations.
