Indianapolis database expert Mark Nichols said it best in a column of tips for his Data Central readers: “All data is dirty.” Raw data often contains errors, and it also has limitations. “The dataset may be a snapshot in time and may not reflect what’s happening right now. In other cases, it may only be a sampling of people or incidents, and may not reflect the full picture. In all cases, it’s just one source of information, not a ‘be-all-end-all’” source,” he wrote.

His advice to readers was to find all they can about a dataset so they can draw the best possible conclusions from it. Our advice to database managers is to give readers a solid briefing on the strengths and weaknesses of each database offered. Here are two we found in our survey.