National EMSC Data Analysis Resource Center
Sometimes you’ll have data where it is too late to make any improvements in the way the data are collected. Other times, however, you can make quality control changes that will allow for more accurate results in the future.
There are two major areas where quality problems can be corrected:
Both of these methods will help you ensure better data quality for future projects.
Sometimes, however, you will just need to accept certain levels of bad data. In these cases, it is your responsibility to report these errors.
For example, if you are reporting gender percentages but had a number of them that were unknown, you will need to place an “unknown” category below “male” and “female.”
If you always follow this practice, you allow others to have the information they need to properly use and interpret the data.