Given the sheer quantity of line items and the time it would take to conduct cost realism on all line items, is it acceptable to conduct the cost realism using a sampling methodology, and extrapolate from there? Are you aware of anyone doing this in the past?
Open full Question Details
I am not aware of anyone doing this in the past, however, this does not mean it has not been done in the past, I just did not see it in any of the GAO reports I read. If in your solicitation you said you were going to do stratified sampling of the cost clins and then random sample X cost clins and perform cost realism analysis on those X clins and came up with a decrement that you applied to the whole stratum, I think you would be okay. But the method you are going to use to do your cost realism needs to be in the solicitation. I could not find anywhere in the FAR where it said you could not take this approach.
I recommend you seek legal’s opinion on this approach.