Increasingly, top journals are suggesting or even requiring that authors share the data underlying their articles. Data sharing has become a key component to rigor initiatives in many disciplines, and supports the validity of the findings if other researchers bring an inquiry to the editorial board. There have been cases of journal retractions for lack of supporting data more than 10 years after publication.
Many journals now encourage or require a Data Sharing Statement when you submit your article for review. The data sharing statement should tell the editor when you will share your data, through what platform, and under what terms. An editorial assistant will usually contact you to follow up with requests for links to the dataset when your data release date arrives.
Things to consider when planning your data sharing statement:
- Data sharing should always be done with full respect given to the protection of participants (through IRB guidance and thorough anonymization) as well as any intellectual property concerns.
- You can request a delay (sometimes called an "embargo") on data release in order to preserve your own research priorities.
- Include any data used to generate the tables and figures in the article. You should also plan to share the data underlying any statistical results that were presented in the article's Results section.
- If custom algorithms or scripts were used to generate your results, those should be included with the data if possible.
- Remember to define your data! Include the full names of variables, your variable definitions, units of measurement, codes used for missing data, and important contextual information like the date and place of data collection.