Selection Bias

Derek Lowe has a great post today on how selection of starting materials in synthesis is often driven by neither random nor rational reasons, but by availability of and/or the past experience of the chemist with the starting materials. While this deals with inorganics, he references other examples. While not directly applicable to analysis, one can find the same mistake often in the characterization of materials. We stick with what we are familiar with.

The “we’ve always looked at it by this method” can be a powerful driver for the analyst too. It’s not always the best choice. Sometimes it is, and sometimes it’s part of the answer, but other times it’s not. Take sorting recycled polymers. FTIR will tell you if its PE or PP, but if you want to know the grade of PE, well, there should be a DSC in the testing plan somewhere.

Now a lot of the time, the tests are required by contract, or someone actually did the work to make sure the best approach is being used. However, this isn’t always true and sometimes it’s more like “well, we had an X in house, so we used it.” That’s why we always like to talk to you about your project and not just run the samples. Sometimes, there is a better way.