Statistics
Study of the collection and analysis of data
Follow Statistics on Notably News to receive short updates to your email — rarely!
| 2022 | A follow-up study examined the practical impact of the competing recommendations for statistical significance thresholds, revealing limited implementation of the proposed p < 0.005 threshold or case-by-case justification approach in scientific research. |
| 2022 | A study suggested that many earlier brain-phenotype studies (brain-wide association studies or BWAS) produced invalid conclusions, noting that replication requires samples from thousands of individuals due to small effect sizes. |
| 2021 | A study by Marta Serra-Garcia and Uri Gneezy investigated the citation patterns of non-replicable research papers in leading journals across general interest, psychology, and economics fields. The research revealed that papers with non-reproducible findings tend to be cited more frequently over time, despite their lack of scientific validity. |
| 2021 | A University of California, San Diego study discovered that papers which cannot be replicated are more likely to be cited, even after replication studies are published. |
| October 2021 | The Reproducibility Project: Cancer Biology examined 53 top cancer research papers published between 2010 and 2012, finding that effect sizes were 85% smaller on average compared to original findings when studies were attempted to be reproduced. |
| 2020 | Regional Census Centers oversaw the operation of 248 Area Census Offices during the decennial census. |
| March 2020 | American FactFinder was decommissioned after 20 years, with data.census.gov becoming the new primary data dissemination platform. |
| 2019 | A study published in Scientific Data analyzed 1,989 articles on water resources and management from 2017, revealing an extremely low reproducibility rate of between 0.6% to 6.8% due to insufficient methodological information. |
| July 2019 | Census Bureau stopped releasing new data via American FactFinder. |
| 2018 | A study published in Nature Human Behaviour replicated 21 social and behavioral science papers from Nature and Science, finding that only about 62% could successfully reproduce original results. |
| 2017 | A large group of scientists and mathematicians signed the paper 'Redefine statistical significance', proposing to change the statistical significance threshold from p < 0.05 to p < 0.005 to improve research reproducibility. |
| 2017 | The Laura and John Arnold Foundation provided an additional $10 million in funding, further supporting efforts to improve scientific research methodology and reproducibility. |
| 2017 | Analysis revealed that most of 1,987 articles in water resources and management were not replicable due to insufficient online information sharing. |
| 2016 | Publication of 'The Problem with p-values', a paper discussing the logical problems of inductive inference in statistical analysis. |
| 2016 | Nature conducted a survey of 1,576 researchers, revealing that only a minority had ever attempted to publish a replication study, and many who had published failed replications were pressured by editors and reviewers to downplay comparisons with original studies. |
| 2016 | A study published in Science journal replicated 18 experimental studies from The American Economic Review and Quarterly Journal of Economics (2011-2014), revealing that approximately 39% of the studies failed to reproduce their original results. |
| 2016 | John Ioannidis published an article critiquing clinical research, arguing that most medical research is not useful and calling for significant reforms, including a shift towards patient-centered research approaches. |
| 2016 | Nature published a survey of 1,576 researchers revealing that over 70% have failed to reproduce another scientist's experiment results, with significant variations across scientific disciplines. |
| July 2016 | The Netherlands Organisation for Scientific Research allocated €3 million in funding specifically for replication studies in social sciences, health research, and healthcare innovation, supporting both reanalysis of existing data and collection of new data. |
| August 2015 | Brian Nosek published the first open empirical study of reproducibility in psychology, The Reproducibility Project: Psychology. The project involved redoing 100 studies from three high-ranking psychology journals, revealing that only 36% of replications yielded significant findings, with effect sizes approximately half the magnitude of original studies. |
| 2014 | Data Colada blog authors coin the term 'p-hacking' and contribute significantly to initiating discussions about the replication crisis in scientific research. |
| 2013 | A special edition of the journal Social Psychology was dedicated to replication studies, highlighting the growing focus on addressing the replication crisis in psychology by encouraging researchers to retest important findings. |
| 2013 | The Laura and John Arnold Foundation provided a $5.25 million grant to launch the Center for Open Science, aiming to address issues in scientific research reproducibility. |
| January 1 2013 | The Census Bureau consolidated its twelve regional offices into six, closing the Boston, Charlotte, Dallas, Detroit, Kansas City, and Seattle offices due to increasing data collection costs and changes in survey management technologies. |
| 2012 | C. Glenn Begley and Lee Ellis published a pivotal paper demonstrating that only 11% of 53 pre-clinical cancer studies could be replicated, exposing critical reliability issues in cancer research. |
| 2012 | An investigation of replication rates in psychology revealed that replication studies involving original authors had higher success rates (91.7%) compared to studies without author overlap (64.6%). |
| 2012 | End of period for cancer research papers examined in the Reproducibility Project: Cancer Biology. |
| 2012 | A survey of psychologists revealed significant methodological issues in research practices: 56% admitted to early stopping of studies, 46% reported only analyses that 'worked', and 38% acknowledged post hoc data exclusion after initial analysis. |
| 2011 | Researchers at Bayer pharmaceutical company conducted an analysis revealing that at most a quarter of their in-house scientific findings could be replicated, highlighting significant challenges in research reproducibility. |
| 2011 | Daryl Bem's experiments on extrasensory perception were critically analyzed, with subsequent direct replications failing to confirm his original findings, revealing methodological flaws in psychological research practices. |
| 2010 | Fanelli conducted a study finding that 91.5% of psychiatry/psychology studies confirmed their expected effects, suggesting significantly higher positive result rates compared to fields like astronomy or geosciences. |
| 2010 | Social priming controversy emerges when the widely-cited 'elderly-walking' study by John Bargh failed to replicate in two direct replications, sparking heated debate in the research community. |
| 2010 | Start of period for cancer research papers examined in the Reproducibility Project: Cancer Biology, which investigated reproducibility of scientific experiments. |
This contents of the box above is based on material from the Wikipedia articles Replication crisis & United States Census Bureau, which are released under the Creative Commons Attribution-ShareAlike 4.0 International License.