Prior research finds that statistically significant results are overrepresented in scientific publications. If significant results are consistently favored in the review process, published results could systematically overstate the magnitude of their findings even under ideal conditions. In this paper, we measure the impact of this publication bias on political science using a new data set of published quantitative results. Although any measurement of publication bias depends on the prior distribution of empirical relationships, we determine that published estimates in political science are on average substantially larger than their true value under a variety of reasonable choices for this prior. We also find that many published estimates have a false positive probability substantially greater than the conventional α = 0.05 threshold for statistical significance if the prior probability of a null relationship exceeds 50%. Finally, although the proportion of published false positives would be reduced if significance tests used a smaller α, this change would not solve the problem of upward bias in the magnitude of published results.

Deterrence theory suggests that extended general deterrent threats are likely to be more effective when a potential challenger views them as capable and credible. When states sign formal defense pacts, they are making explicit extended general deterrent threats. Thus, the population of defense pacts allows us an opportunity to judge the efficacy of extended deterrent threats with different characteristics. We find that defense pacts with more capability and more credibility reduce the probability that a member state will be a target of a militarized dispute. We also find that states can affect the capability and credibility of their extended deterrent threats through alliance design. Members of defense pacts that include higher levels of peacetime military coordination are less likely to be attacked. This analysis provides support for deterrence theory in the context of extended general deterrence. It also provides evidence that should aid policymakers in designing security structures to meet their goals.In explaining international conflict, most scholars place theories of deterrence in a central position. Theories of deterrence consistently lead to the prediction that deterrent threats that are capable and credible are more likely to be successful. While empirical evidence on direct general deterrence has been generally supportive of deterrence theory, evidence on extended deterrence is less straightforward, mainly because most past research on extended deterrence has been conducted at the immediate deterrence stage. We know

<p>Increases in technology have rapidly advanced the capabilities and ubiquity of scientific instrumentation. Coupled with the demand for increased transparency and reproducibility in science, these advances have necessitated new systems of data management and archival practices. Laboratories are working to update their methods of data curation in line with these evolving best-practices, moving data from often disorderly private domains to publicly available, collaborative platforms. At the Hamilton Isotope Laboratory (HIL) of Hamilton College, the isotope ratio mass spectrometer (IRMS) is utilized across STEM disciplines for a combination of student, faculty, and course-related research, including both internal and external users. With over 200 sets of analytical runs processed in the past five years, documenting instrument usage and archiving the data produced is crucial to maintaining a state-of-the-art facility. However, previous to this project, the HIL faced significant barriers to proper data curation, storage, and accessibility including: a) data files were produced with variable format and nomenclature; b) data files were difficult to interpret without explanation from the lab technician; c) key metadata tying results to respective researchers and projects were missing; d) accessibility to data was limited due to storage on an individual computer; and e) data curation was an intellectual responsibility and burden for the lab technician. Additionally, as the HIL is housed within an undergraduate institution, the high rate of turnover for lab groups created additional barriers to the preservation of long-term, institutional knowledge, as students worked with the HIL for a year or less. These factors necessitate the establishment of new data management practices to ensure accessibility and longevity of scientific data and metadata. In this project, 283 Excel files of previously recorded data generated by the HIL IRMS were modified and cleaned to prepare data for submission to EarthChem, a public repository for geochemical data. Existing Excel files were manually manipulated, several original R code scripts were generated and employed, and procedures were established to backtrace projects and collect key metadata. Most critically, a new internal system of data collection was established with standardized nomenclature and framework. For future usage of the IRMS, data will be exported directly into a template compatible with EarthChem, thereby removing barriers for principal investigators (PIs) and research groups to archive their data in the public domain upon completion of their projects and publications.</p>