Open Science Practices
Advanced Statistical Methods
Making Research Transparent and Reproducible
Open science practices β open data, open code, preregistration, and FAIR principles β increase transparency, enable verification, and accelerate scientific progress through collaboration and reuse.
- Research credibility β Open data allows independent verification of analytical conclusions
- Collaboration β Shared code and datasets enable other researchers to build on published work
- Funding compliance β Increasingly, grant agencies mandate data sharing and open access publication
Open science is not just good ethics β it produces better, more trustworthy science.
The replication crisis in psychology, medicine, and social sciences has catalyzed a paradigm shift toward open science β a set of principles and practices that increase the transparency, accessibility, and reproducibility of research.
The Four Pillars of Open Science
The Reproducibility Project (Open Science Collaboration, 2015) attempted to replicate 100 published psychology studies and found that only 36% yielded statistically significant results in replications, compared to 97% in the originals. This alarming discrepancy underscored the need for structural reforms.
FAIR Data Principles
The FAIR principles (Wilkinson et al., 2016) provide a framework for data management:
Data Sharing and Repositories
Major data repositories include:
| Repository | Domain | DOI Support | Access |
|---|---|---|---|
| Dryad | General | Yes | Free (with publication) |
| Zenodo | General | Yes | Free |
| ICPSR | Social sciences | Yes | Restricted access |
| OpenNeuro | Neuroimaging | Yes | Free |
| Figshare | General | Yes | Free (limited) |
| OSF | Multi-disciplinary | Yes | Free |
Code Sharing and Reproducibility
Python Implementation: Reproducibility Workflow
Preregistration and Transparency
The key distinction is between:
Registered Reports
Registered Reports (RR) are a publishing format where journals peer-review and accept studies before results are known, based on the importance of the research question and the quality of the methodology.
Measuring Reproducibility
The Open Science Collaboration (2015) found R β 0.36 for psychology. Subsequent large-scale replication projects have found:
| Field | Replication Rate | Source |
|---|---|---|
| Psychology | 36% | OSC (2015) |
| Economics | 61% | Camerer et al. (2016) |
| Social Science | 62% | Camerer et al. (2018) |
| Cancer Biology | 46% | Begley & Ellis (2012) |