If your organization runs a Web site, you may think that the statistics you're getting from your log analysis software are presenting a pretty accurate picture of your site's traffic. Well, not so fast. A newspaper research and consulting firm named Belden Associates did a study that shows most site traffic numbers are way off base for a variety of reasons. For example, Belden estimates that half of the daily users of a newspaper website access it from more than one computer, resulting a lot of double counting. I couldn't find the report on the Belden site, but a write-up in the Christian Science Monitor gives some details. Even though the study was done for the newspaper industry, most of what's reported is not specific to newspaper Web sites.