Depending on the nature of the error (random vs. systematic) either discard data or use a systematic ...
If possible, the investigators consider salvaging the data through the use of systematic correction of the data. If the data correction attempt is unsuccessful, the data will be discarded, and a new plan formed.
These are some possible outcomes:
The error in data collection appears to be a systematic one. Thus after implementing an appropriate data correction procedure, the integrity of the collected data is maintained.
The errors in data collection are random and the data is discarded. The investigators decide to use the remaining data for analysis, and retrain the staff to improve the data collection procedure.
After retraining staff, the staff returns to the field and attempt to re-collect the necessary data.