Making mistakes

The work we do is complicated, and mistakes happen. When you notice a mistake (a bug, misinterpretation, mislabeling, or any other error), it is critical that you report the mistake immediately. Whereas mistakes are unavoidable in science, negative impacts can be minimized by fostering a workplace where reporting mistakes is celebrated and accepted as part of the natural course of getting things done. Mistakes are opportunities to learn and grow, and identifying or noticing mistakes should be celebrated as part of our growth as researchers. However, real harm can come from failing to report mistakes soon enough. There is a proverb that says the best time to plant a tree was 20 years ago; the second best time is now. Analogously, the best time to identify and correct a mistake may have been in the past– but the second best time is right now!

Example scenarios (not an exhaustive list):

  1. You’ve shared a figure, statistic, or other result, and you’ve realized there’s a bug in your code.
  2. You tried to collect some data and the experiment crashed or yielded corrupted data.
  3. You’re re-reading a paper that you shared, and you notice a mistake or typo.
  4. You made a plan with your project team and you realized it’s flawed in some way, or that there’s potentially a better solution or approach.
  5. You released a software package and you’ve found a bug or error.

Appropriate actions for each of the above scenarios (this should happen immediately after you notice the mistake):

  1. Verify: Double-check to confirm the mistake. Review your code, re-run experiments if necessary, or revisit reference materials.
  2. Report: Create a GitHub issue detailing the problem. Include information on how to replicate the issue, expected versus observed behavior, and any relevant system details like operating environment and software versions.
  3. Collaborate: Use Slack or other communication channels to discuss the issue with your project team. Collaborate to understand the root cause and formulate an effective plan of action.
  4. Seek guidance: If unsure, reach out to colleagues for assistance. Utilize resources like Google or Stack Exchange to explore potential solutions.

Remember, consulting with others when in doubt is always a wise move. Every mistake is an opportunity for growth and improvement, contributing to our collective journey as researchers. Let’s continue fostering an environment where learning from mistakes is celebrated and supported, ensuring that our work maintains its highest standards of quality and reliability.