What is the best current paradigm for insuring accurate data has been entered. I remember in the old days of high volume keypunch, two keypunchers would key the same data which was then compared at the end of a job for discrepancies. Once the discrepancies were found and corrected, the job was considered validated.

Is this still the best strategy in an environment of high volume entries using FM?

Thanks.

Dave

Blind or double entry has its purpose although it isn't used as much any more because fewer transactions come as hardcopy to computer; much of it comes computer to computer. But those entering checks (such as UPS COD checks) still use these methods (usually two different people enter the same information but I've also seen same person enter it once, then field clears then they enter it again). Much depends upon the importance of the data. In the 'ole days' few people used computers. Now everyone does. So, since everyone does, those who make mistakes need to correct them. If someone has a stack of checks, the sum of the computer entry can be compared to an adding machine total (for instance). When you have something such as payments depending upon the number results then the payee may request blind or double entry validation.

Business rules need to be established before a developer could make suggestions on implementation ...