Adam Altmejd
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Adam Altmejd has a total of 64 co-authors. Here are the top co-authors by number of shared publications:
Anna Dreber
Department of Economics, Stockholm School of Economics, Stockholm, Sweden
6 shared publicationsEskil Forsell
Department of Economics, Stockholm School of Economics, Stockholm, Sweden
5 shared publications- Jürgen Huber
University of Innsbruck
5 shared publications Michael Kirchler
Universität Innsbruck, Innsbruck, Austria
5 shared publications- Teck-Hua Ho
Nanyang Technological University, Singapore
4 shared publications - Thomas Pfeiffer
NZ IAS, Massey University, Auckland, New Zealand
4 shared publications
Latest Publications
- Open Access
- 1 January 2025
- Published by Elsevier in SSRN Electronic Journal
Pipeline vs. Choice: The Global Gender Gap in STEM Applications
Women account for only 35% of global STEM graduates, a share that has remained unchanged for a decade. We use administrative microdata from centralized university admissions in ten systems to deliver the first cross-national decomposition of the STEM gender gap into a pipeline gap (academic preparedness) and a choice gap (first-choice field conditional on eligibility). In deferred-acceptance platforms where eligibility is score-based, we isolate preferences from access. The pipeline gap varies widely, from -19 to +31 percentage points across education systems. By contrast, the choice gap is remarkably stable: high-scoring women are 25 percentage points less likely than men to rank STEM first.
- 1 January 2025
- Published by Elsevier in SSRN Electronic Journal
Open AccessPipeline vs. Choice: The Global Gender Gap in STEM Applications
Women account for only 35% of global STEM graduates, a share that has remained unchanged for a decade. We use administrative microdata from centralized university admissions in ten systems to deliver the first cross-national decomposition of the STEM gender gap into a pipeline gap (academic preparedness) and a choice gap (first-choice field conditional on eligibility). In deferred-acceptance platforms where eligibility is score-based, we isolate preferences from access. The pipeline gap varies widely, from -19 to +31 percentage points across education systems. By contrast, the choice gap is remarkably stable: high-scoring women are 25 percentage points less likely than men to rank STEM first.
- Open Access
- 9 November 2023
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 120 (46)
Inequality and COVID-19 in Sweden: Relative risks of nine bad life events, by four social gradients, in pandemic vs. prepandemic years
The COVID-19 pandemic struck societies directly and indirectly, not just challenging population health but disrupting many aspects of life. Different effects of the spreading virus—and the measures to fight it—are reported and discussed in different scientific fora, with hard-to-compare methods and metrics from different traditions. While the pandemic struck some groups more than others, it is difficult to assess the comprehensive impact on social inequalities. This paper gauges social inequalities using individual-level administrative data for Sweden’s entire population. We describe and analyze the relative risks for different social groups in four dimensions—gender, education, income, and world region of birth—to experience three types of COVID-19 incidence, as well as six additional negative life outcomes that reflect general health, access to medical care, and economic strain. During the pandemic, the overall population faced severe morbidity and mortality from COVID-19 and saw higher all-cause mortality, income losses and unemployment risks, as well as reduced access to medical care. These burdens fell more heavily on individuals with low income or education and on immigrants. Although these vulnerable groups experienced larger absolute risks of suffering the direct and indirect consequences of the pandemic, the relative risks in pandemic years (2020 and 2021) were conspicuously similar to those in prepandemic years (2016 to 2019).
- 9 November 2023
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 120 (46)
Open AccessInequality and COVID-19 in Sweden: Relative risks of nine bad life events, by four social gradients, in pandemic vs. prepandemic years
The COVID-19 pandemic struck societies directly and indirectly, not just challenging population health but disrupting many aspects of life. Different effects of the spreading virus—and the measures to fight it—are reported and discussed in different scientific fora, with hard-to-compare methods and metrics from different traditions. While the pandemic struck some groups more than others, it is difficult to assess the comprehensive impact on social inequalities. This paper gauges social inequalities using individual-level administrative data for Sweden’s entire population. We describe and analyze the relative risks for different social groups in four dimensions—gender, education, income, and world region of birth—to experience three types of COVID-19 incidence, as well as six additional negative life outcomes that reflect general health, access to medical care, and economic strain. During the pandemic, the overall population faced severe morbidity and mortality from COVID-19 and saw higher all-cause mortality, income losses and unemployment risks, as well as reduced access to medical care. These burdens fell more heavily on individuals with low income or education and on immigrants. Although these vulnerable groups experienced larger absolute risks of suffering the direct and indirect consequences of the pandemic, the relative risks in pandemic years (2020 and 2021) were conspicuously similar to those in prepandemic years (2016 to 2019).
- Open Access
- 9 February 2023
- Published by MDPI AG in International Journal of Environmental Research and Public Health
- Vol. 20 (4) , 3040
Nowcasting COVID-19 Statistics Reported with Delay: A Case-Study of Sweden and the UK
The COVID-19 pandemic has demonstrated the importance of unbiased, real-time statistics of trends in disease events in order to achieve an effective response. Because of reporting delays, real-time statistics frequently underestimate the total number of infections, hospitalizations and deaths. When studied by event date, such delays also risk creating an illusion of a downward trend. Here, we describe a statistical methodology for predicting true daily quantities and their uncertainty, estimated using historical reporting delays. The methodology takes into account the observed distribution pattern of the lag. It is derived from the “removal method”—a well-established estimation framework in the field of ecology.
- 9 February 2023
- Published by MDPI AG in International Journal of Environmental Research and Public Health
- Vol. 20 (4) , 3040
Open AccessNowcasting COVID-19 Statistics Reported with Delay: A Case-Study of Sweden and the UK
The COVID-19 pandemic has demonstrated the importance of unbiased, real-time statistics of trends in disease events in order to achieve an effective response. Because of reporting delays, real-time statistics frequently underestimate the total number of infections, hospitalizations and deaths. When studied by event date, such delays also risk creating an illusion of a downward trend. Here, we describe a statistical methodology for predicting true daily quantities and their uncertainty, estimated using historical reporting delays. The methodology takes into account the observed distribution pattern of the lag. It is derived from the “removal method”—a well-established estimation framework in the field of ecology.
- Open Access
- 24 December 2022
- Published by Center for Open Science in SocArXiv
Inequality and COVID-19 in Sweden: Relative risks of nine negative life events, along four social gradients, in pandemic vs. pre-pandemic years
The COVID-19 pandemic struck societies directly and indirectly, challenging not just people’s health but many aspects of life. But pandemic burdens fell more heavily on some groups than others. These different consequences of the spreading virus – and the measures to fight them – are reported and analyzed in different scientific fora, with hard-to-compare methods that largely follow disciplinary boundaries. As a result, it is hard to grasp the overall impact of the pandemic on inequality. This paper relies on individual-level, administrative data for Sweden’s entire population to describe how different social groups fared in terms of nine outcomes: three types of COVID-19 incidence, as well as six other negative life events. During 2020, the population faced severe morbidity and mortality from COVID-19 and saw higher all-cause mortality, income losses and unemployment risks, as well as reduced access to medical care. In terms of relative risks, these burdens fell disproportionately on those with low income or education, and on residents born outside of Sweden. In the pandemic, all-cause mortality, unemployment, substantial income loss, poor mental health, and reduced access to health care went up for all groups in Sweden. But relative risks across social groups were strikingly similar to those in pre-pandemic years.
- 24 December 2022
- Published by Center for Open Science in SocArXiv
Open AccessInequality and COVID-19 in Sweden: Relative risks of nine negative life events, along four social gradients, in pandemic vs. pre-pandemic years
The COVID-19 pandemic struck societies directly and indirectly, challenging not just people’s health but many aspects of life. But pandemic burdens fell more heavily on some groups than others. These different consequences of the spreading virus – and the measures to fight them – are reported and analyzed in different scientific fora, with hard-to-compare methods that largely follow disciplinary boundaries. As a result, it is hard to grasp the overall impact of the pandemic on inequality. This paper relies on individual-level, administrative data for Sweden’s entire population to describe how different social groups fared in terms of nine outcomes: three types of COVID-19 incidence, as well as six other negative life events. During 2020, the population faced severe morbidity and mortality from COVID-19 and saw higher all-cause mortality, income losses and unemployment risks, as well as reduced access to medical care. In terms of relative risks, these burdens fell disproportionately on those with low income or education, and on residents born outside of Sweden. In the pandemic, all-cause mortality, unemployment, substantial income loss, poor mental health, and reduced access to health care went up for all groups in Sweden. But relative risks across social groups were strikingly similar to those in pre-pandemic years.
- Open Access
Nowcasting Covid-19 statistics reported withdelay: a case-study of Sweden
The new corona virus disease -- COVID-2019 -- is rapidly spreading through the world. The availability of unbiased timely statistics of trends in disease events are a key to effective responses. But due to reporting delays, the most recently reported numbers are frequently underestimating of the total number of infections, hospitalizations and deaths creating an illusion of a downward trend. Here we describe a statistical methodology for predicting true daily quantities and their uncertainty, estimated using historical reporting delays. The methodology takes into account the observed distribution pattern of the lag. It is derived from the removal method, a well-established estimation framework in the field of ecology.
Open AccessNowcasting Covid-19 statistics reported withdelay: a case-study of Sweden
The new corona virus disease -- COVID-2019 -- is rapidly spreading through the world. The availability of unbiased timely statistics of trends in disease events are a key to effective responses. But due to reporting delays, the most recently reported numbers are frequently underestimating of the total number of infections, hospitalizations and deaths creating an illusion of a downward trend. Here we describe a statistical methodology for predicting true daily quantities and their uncertainty, estimated using historical reporting delays. The methodology takes into account the observed distribution pattern of the lag. It is derived from the removal method, a well-established estimation framework in the field of ecology.
- Open Access
- 1 January 2020
- Published by Elsevier in SSRN Electronic Journal
O Brother, Where Start Thou? Sibling Spillovers on College and Major Choice in Four Countries
Family and social networks are widely believed to influence important life decisions but identifying their causal effects is notoriously difficult. Using admissions thresholds that directly affect older but not younger siblings' college options, we present evidence from the United States, Chile, Sweden and Croatia that older siblings' college and major choices can significantly influence their younger siblings' college and major choices. On the extensive margin, an older sibling's enrollment in a better college increases a younger sibling's probability of enrolling in college at all, especially for families with low predicted probabilities of enrollment. On the intensive margin, an older sibling's choice of college or major increases the probability that a younger sibling applies to and enrolls in that same college or major. Spillovers in major choice are stronger when older siblings enroll and succeed in more selective and higher-earning majors. The observed spillovers are not well-explained by price, income, proximity or legacy effects, but are most consistent with older siblings transmitting otherwise unavailable information about the college experience and its potential returns. The importance of such personally salient in-formation may partly explain persistent differences in college-going rates by geography, income, and other determinants of social networks.
- 1 January 2020
- Published by Elsevier in SSRN Electronic Journal
Open AccessO Brother, Where Start Thou? Sibling Spillovers on College and Major Choice in Four Countries
Family and social networks are widely believed to influence important life decisions but identifying their causal effects is notoriously difficult. Using admissions thresholds that directly affect older but not younger siblings' college options, we present evidence from the United States, Chile, Sweden and Croatia that older siblings' college and major choices can significantly influence their younger siblings' college and major choices. On the extensive margin, an older sibling's enrollment in a better college increases a younger sibling's probability of enrolling in college at all, especially for families with low predicted probabilities of enrollment. On the intensive margin, an older sibling's choice of college or major increases the probability that a younger sibling applies to and enrolls in that same college or major. Spillovers in major choice are stronger when older siblings enroll and succeed in more selective and higher-earning majors. The observed spillovers are not well-explained by price, income, proximity or legacy effects, but are most consistent with older siblings transmitting otherwise unavailable information about the college experience and its potential returns. The importance of such personally salient in-formation may partly explain persistent differences in college-going rates by geography, income, and other determinants of social networks.
- Open Access
- 5 December 2019
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 14 (12) , e0225826
Predicting the replicability of social science lab experiments
We measure how accurately replication of experimental results can be predicted by black-box statistical models. With data from four large-scale replication projects in experimental psychology and economics, and techniques from machine learning, we train predictive models and study which variables drive predictable replication. The models predicts binary replication with a cross-validated accuracy rate of 70% (AUC of 0.77) and estimates of relative effect sizes with a Spearman ρ of 0.38. The accuracy level is similar to market-aggregated beliefs of peer scientists [1, 2]. The predictive power is validated in a pre-registered out of sample test of the outcome of [3], where 71% (AUC of 0.73) of replications are predicted correctly and effect size correlations amount to ρ = 0.25. Basic features such as the sample and effect sizes in original papers, and whether reported effects are single-variable main effects or two-variable interactions, are predictive of successful replication. The models presented in this paper are simple tools to produce cheap, prognostic replicability metrics. These models could be useful in institutionalizing the process of evaluation of new findings and guiding resources to those direct replications that are likely to be most informative.
- 5 December 2019
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 14 (12) , e0225826
Open AccessPredicting the replicability of social science lab experiments
We measure how accurately replication of experimental results can be predicted by black-box statistical models. With data from four large-scale replication projects in experimental psychology and economics, and techniques from machine learning, we train predictive models and study which variables drive predictable replication. The models predicts binary replication with a cross-validated accuracy rate of 70% (AUC of 0.77) and estimates of relative effect sizes with a Spearman ρ of 0.38. The accuracy level is similar to market-aggregated beliefs of peer scientists [1, 2]. The predictive power is validated in a pre-registered out of sample test of the outcome of [3], where 71% (AUC of 0.73) of replications are predicted correctly and effect size correlations amount to ρ = 0.25. Basic features such as the sample and effect sizes in original papers, and whether reported effects are single-variable main effects or two-variable interactions, are predictive of successful replication. The models presented in this paper are simple tools to produce cheap, prognostic replicability metrics. These models could be useful in institutionalizing the process of evaluation of new findings and guiding resources to those direct replications that are likely to be most informative.
- Open Access
- 16 January 2019
- Published by Center for Open Science in MetaArXiv
Predicting the Replicability of Social Science Lab Experiments
We measure how accurately replication of experimental results can be predicted by a black-box statistical model. With data from four large- scale replication projects in experimental psychology and economics, and techniques from machine learning, we train a predictive model and study which variables drive predictable replication.The model predicts binary replication with a cross validated accuracy rate of 70% (AUC of 0.79) and relative effect size with a Spearman ρ of 0.38. The accuracy level is similar to the market-aggregated beliefs of peer scientists (Camerer et al., 2016; Dreber et al., 2015). The predictive power is validated in a pre-registered out of sample test of the outcome of Camerer et al. (2018b), where 71% (AUC of 0.73) of replications are predicted correctly and effect size correlations amount to ρ = 0.25.Basic features such as the sample and effect sizes in original papers, and whether reported effects are single-variable main effects or two- variable interactions, are predictive of successful replication. The models presented in this paper are simple tools to produce cheap, prognostic replicability metrics. These models could be useful in institutionalizing the process of evaluation of new findings and guiding resources to those direct replications that are likely to be most informative.
- 16 January 2019
- Published by Center for Open Science in MetaArXiv
Open AccessPredicting the Replicability of Social Science Lab Experiments
We measure how accurately replication of experimental results can be predicted by a black-box statistical model. With data from four large- scale replication projects in experimental psychology and economics, and techniques from machine learning, we train a predictive model and study which variables drive predictable replication.The model predicts binary replication with a cross validated accuracy rate of 70% (AUC of 0.79) and relative effect size with a Spearman ρ of 0.38. The accuracy level is similar to the market-aggregated beliefs of peer scientists (Camerer et al., 2016; Dreber et al., 2015). The predictive power is validated in a pre-registered out of sample test of the outcome of Camerer et al. (2018b), where 71% (AUC of 0.73) of replications are predicted correctly and effect size correlations amount to ρ = 0.25.Basic features such as the sample and effect sizes in original papers, and whether reported effects are single-variable main effects or two- variable interactions, are predictive of successful replication. The models presented in this paper are simple tools to produce cheap, prognostic replicability metrics. These models could be useful in institutionalizing the process of evaluation of new findings and guiding resources to those direct replications that are likely to be most informative.
- Top Cited Papers
- 27 August 2018
- Published by Springer Nature in Nature Human Behaviour
- Vol. 2 (9) , 637-644
Evaluating the replicability of social science experiments in Nature and Science between 2010 and 2015
Being able to replicate scientific findings is crucial for scientific progress1,2,3,4,5,6,7,8,9,10,11,12,13,14,15. We replicate 21 systematically selected experimental studies in the social sciences published in Nature and Science between 2010 and 201516,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36. The replications follow analysis plans reviewed by the original authors and pre-registered prior to the replications. The replications are high powered, with sample sizes on average about five times higher than in the original studies. We find a significant effect in the same direction as the original study for 13 (62%) studies, and the effect size of the replications is on average about 50% of the original effect size. Replicability varies between 12 (57%) and 14 (67%) studies for complementary replicability indicators. Consistent with these results, the estimated true-positive rate is 67% in a Bayesian analysis. The relative effect size of true positives is estimated to be 71%, suggesting that both false positives and inflated effect sizes of true positives contribute to imperfect reproducibility. Furthermore, we find that peer beliefs of replicability are strongly related to replicability, suggesting that the research community could predict which results would replicate and that failures to replicate were not the result of chance alone.
- 27 August 2018
- Published by Springer Nature in Nature Human Behaviour
- Vol. 2 (9) , 637-644
Top Cited PapersEvaluating the replicability of social science experiments in Nature and Science between 2010 and 2015
Being able to replicate scientific findings is crucial for scientific progress1,2,3,4,5,6,7,8,9,10,11,12,13,14,15. We replicate 21 systematically selected experimental studies in the social sciences published in Nature and Science between 2010 and 201516,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36. The replications follow analysis plans reviewed by the original authors and pre-registered prior to the replications. The replications are high powered, with sample sizes on average about five times higher than in the original studies. We find a significant effect in the same direction as the original study for 13 (62%) studies, and the effect size of the replications is on average about 50% of the original effect size. Replicability varies between 12 (57%) and 14 (67%) studies for complementary replicability indicators. Consistent with these results, the estimated true-positive rate is 67% in a Bayesian analysis. The relative effect size of true positives is estimated to be 71%, suggesting that both false positives and inflated effect sizes of true positives contribute to imperfect reproducibility. Furthermore, we find that peer beliefs of replicability are strongly related to replicability, suggesting that the research community could predict which results would replicate and that failures to replicate were not the result of chance alone.
- Open Access
- 27 August 2018
- Published by Center for Open Science in SocArXiv
Evaluating the replicability of social science experiments in Nature and Science between 2010 and 2015
Being able to replicate scientific findings is crucial for scientific progress. We replicate 21 systematically selected experimental studies in the social sciences published in Nature and Science between 2010 and 2015. The replications follow analysis plans reviewed by the original authors and pre-registered prior to the replications. The replications are high powered, with sample sizes on average about five times higher than in the original studies. We find a significant effect in the same direction as the original study for 13 (62%) studies, and the effect size of the replications is on average about 50% of the original effect size. Replicability varies between 12 (57%) and 14 (67%) studies for complementary replicability indicators. Consistent with these results, the estimated true positive rate is 67% in a Bayesian analysis. The relative effect size of true positives is estimated to be 71%, suggesting that both false positives and inflated effect sizes of true positives contribute to imperfect reproducibility. Furthermore, we find that peer beliefs of replicability are strongly related to replicability, suggesting that the research community could predict which results would replicate and that failures to replicate were not the result of chance alone.
- 27 August 2018
- Published by Center for Open Science in SocArXiv
Open AccessEvaluating the replicability of social science experiments in Nature and Science between 2010 and 2015
Being able to replicate scientific findings is crucial for scientific progress. We replicate 21 systematically selected experimental studies in the social sciences published in Nature and Science between 2010 and 2015. The replications follow analysis plans reviewed by the original authors and pre-registered prior to the replications. The replications are high powered, with sample sizes on average about five times higher than in the original studies. We find a significant effect in the same direction as the original study for 13 (62%) studies, and the effect size of the replications is on average about 50% of the original effect size. Replicability varies between 12 (57%) and 14 (67%) studies for complementary replicability indicators. Consistent with these results, the estimated true positive rate is 67% in a Bayesian analysis. The relative effect size of true positives is estimated to be 71%, suggesting that both false positives and inflated effect sizes of true positives contribute to imperfect reproducibility. Furthermore, we find that peer beliefs of replicability are strongly related to replicability, suggesting that the research community could predict which results would replicate and that failures to replicate were not the result of chance alone.