- When an investigator is suspected of fraud, can you use the data he/she has generated, and should he be paid?
The issue of proof was discussed at length as being the key indicator in such cases. If fraud is proven, the data should be excluded from analysis. However it was considered that it is possible to use statistical techniques in order to compare analyses which include and exclude the suspect data. This may be a useful tool in order to avoid merely excluding a data sub-set, and may be a technique employed in cases where the fraud is merely suspected. In either case, a full explanation should be included in the study report.
If fraud is proven, payment should be withheld. However if fraud is only suspected, payment may be withheld on a pro rata basis. Historically, it has not been possible to recoup payment already made to a fraudulent investigator, as this requires civil law action, often resulting in costs higher than the initial payments.
|
