Charts You Can't Trust (continued)

When done well, graphs and charts tell a story or make a point quickly and coherently. Education Sector has a regular feature we call Charts You Can Trust, where we use a new data source to make a point about the state of education policy. You can read them all here. Despite Kevin's best intentions, we've never actually run one you can't trust. That critique is left for The Quick and the Ed.

Look at the chart below*. It measures the changes in the University of Georgia's undergraduate female enrollment (a pretty hot issue right now).
The chart makes it seem like some big changes were occurring down in Athens. But in reality, if you took the time to calculate the difference between the Fall 2002 and the Fall 2006 female enrollments, you'd find a change of 521 students, or a little less than 4%. To show this reality, the chart should have looked something like this:I suppose we could quibble about whether you should always make 0 the baseline for graphs, but I think we can all agree on the need for honest visual presentations.

*You can find this chart or make your own at the University System of Georgia website here. I made several; some were normal, but quite a few looked like the one above.