Justifying Secondary Data Posted on February 29, 2008 Bruce G. Link (bio) describes a litmus test for deciding to use data other than your own.

When [secondary data is] collected for a different purpose, I think that you're almost always not going to have exactly what you want. But your job is to find as close a match as you can. And then you justify it by saying, "This is good, and I'm willing to test my hypothesis, my cherished hypothesis, that I worked so hard to think through with this data."

If you can say that, then you've justified it. It may not be perfect. It may not test every aspect of what you need to test. But if it tests some part of it well enough that you're willing to say, "I can test this using this data," then that's the litmus test. If you can't, then you don't do it. You look someplace else or you collect your own data.

« Back to Article