NEWS BLOG (WSAU) How do we determine the difference between data manipulation and honest mistakes? There’s a clue. Look who benefits from the so-called error. If  a “mistake” is to someone's advantage, it’s more likely to be intentional manipulation.

Supermarkets have explained bar-code errors as accidents and computer glitches as mistakes. But researchers have found that almost all mistakes are in the store’s favor. The most-common mistake: sales or price reductions advertised on the shelves that aren’t in the computers that scan the bar-codes yet. But when those sales expire, the price increases that follow take effect immediately. This seems more like manipulation than an honest error. After all, there are almost no cases where consumers are under-charged.

Suppose someone runs a cash-based business. At the end of the day they need to total their receipts and reconcile it with their cash box. If the cash on-hand is ‘over’ some days and ‘under’ on others, honest mistakes are being made. If the only mistake is that money is missing, the employee looks like a thief – and probably is.

Now consider the mis-coding of violent crime in Milwaukee. The police department claims this is an honest mistake involving poorly trained data entry clerks and a misunderstanding of what a simple assault versus an aggravated assault. These mistakes have overwhelmingly been in Milwaukee’s favor, making it appear the violent crime rate is down when it isn’t. More than 5,000 cases were more serious than reported, only 1,200 were less serious.

It’s possible this is a mistake. It would be more convincing if the mistakes closer to 50-50 instead of 5-to-1 in the city’s favor. The city will now spend $25,000 on an outside auditor to review and correct the data. Some mistakes are more expensive than others. And, of course, some mistakes aren’t mistakes at all.

Chris Conley
7/18/12