When given a report with an incomplete specification and/or garbage data, and there is not enough time to verify everything, the trick is to document every gosh-darned assumption that you have to make. You can send the report attached to an email that lists in plain language all 20 of your assumptions (Feb 29 handling, using the max expense for a date, etc.). If the consumer of the report disagrees with any of your assumptions, then you can always redo the report but at least they know what they are getting and you can say that you responded quickly.

Having said all that, thanks for the article. It really helps to see how someone else's thought process works. Then you can pick out an approach here or there that perhaps you should try yourself when the opportunity arises.

Curt Coker