There are many times when data bites the person using it. Maybe it is unexpectedly incomplete or biased. Maybe it is more dated than for what it is used for. Possibly it is the wrong data entirely for the use at hand. Using data poorly is a short path to a bit of embarrassment.
All data has use though:
- Incomplete data implies a process that doesn’t require completeness. There are likely participants who are either cheating the system or not using it appropriately.
- Biased data implies that there are no controls on quality. People are entering the numbers that make them look best as opposed to real and accurate data. This can indicate the incentives that people have for gaming the system.
- Dated data is related to incomplete data. The system may not be applying system flags for dates entered or updated which prevent you from knowing when it is from.
- Data used inappropriately often reflects a misunderstanding of the core problem or what the data really means. Sometimes this just implies poor statistical application but it could also imply cherry picking of data to support a pre-defined position.
All data points are useful if you know how to use them.