In contrast to a scanning study which looks for hints about what might be going on by glancing at many technologies, activities, and outcomes, a proof-tracking study usually gathers information in depth on one triad in order to create convincing evidence for taking action.

Limitations of proof-tracking studies: We've never seen a cause-effect study that was definitive - it's rather difficult to be certain about any cause-effect relationship involving human minds - so, no matter what you do, the results will be suggestive, not absolutely conclusive. That same 'uncertainty principle' goes for tracking change over time - our instruments are necessarily crude. It's usually sufficient, however, if a proof-tracking study gives us some valid hints about what's been going on. Studies can't eliminate the risks associated with big decisions, but their findings can reduce those risks.

Rather than continuing to describe proof-tracking studies in the abstract, let's use an example: the Houston case study on this site. The following pages describe one way to gather evidence about whether available technology is helping students learn algebra at the University of Houston Downtown. Please keep in mind that there are almost as many (good and bad) ways to assess the use of technology in an algebra course as there are ways to paint a picture of the scenery outside your window. What follows represent one person's choices during approximately two hours of making a preliminary sketch of a study. Let the reader beware!