Why data management?ΒΆ
- The short answer:
- save time;
- do better science.
By taking some time at the beginning of a project to think carefully about how you will handle your data, simulations and analysis results; or, if you are already in the middle of a project, taking some time to reflect on your data handling and, possibly, reorganise your existing data and results, you can save yourself a huge amount of time and energy in the future, as well as improve the reproducibility of your results.
Some of the problems arising from poor data management:
- taking too much time to find relevant files;
- increased risk of errors and mistakes;
- problems understanding what you did six months ago;
- problems working with collaborators;
- difficulty in explaining things to new students;
- general confusion...
These problems are particularly acute in systems and computational neuroscience, especially when using MEAs, due to the complexity of the experimental protocols, the large number of parallel channels and of correlations between them, and the consequent complexity of the data analysis workflows used.
Probably less important at the moment, but perhaps of increased importance in the future, many funding bodies require data management plans, and in future they, and perhaps journals, may increasingly require depositing your data in a public repository on completion of the project/publication of the paper.