Data analysts should always be skeptical of their data, because it very often goes bad. The Quartz bad data guide is a summary of the most common data flaws, from automatic Excel conversions to inappropriate use of Margins of Error.