Quality Assurance 101: Reproducibility
Last week I posted about environmental practitioners limiting their opinions to the areas in which they have expertise. From there I went off on a little rant about the importance of consultants providing data to prove their conclusions and predictions. Today’s little lecture is on the importance of scientific results being reproducible. Reproducibility is a bedrock scientific principle.
When scientists talk about reproducibility, they are not talking about the nocturnal habits of our species. Instead they are referring to the ability of two scientists to be able to conduct the same experiment and come up with the same result. The experiment could be as simple as measuring the volume of a liquid or as complex as measuring the width of an atom. Either way, if multiple scientists are able to get the same answer, the data is said to be reproducible and therefor to have scientific merit. On the other hand if they unable to arrive at the same result, then the resulting data is said to be “irreproducible” or “non-reproducible” and the data has no scientific credibility.
Among scientists, saying that data is non-reproducible a very serious indictment. There is a nerdy humor publication called the Journal of Irreprocucible Results, that scientists get a big chuckle over. The fact that irreproducible results are a source of humor, should give you a sense for their value among professionals.
So what does this have to do with environmental engineering? Glad you asked. I can not tell you how many times OTO has been brought in for a second opinion right before a client was getting ready to spend a large amount of money on remediation. In these cases there is generally “data” indicating a horrible environmental problem. Yet when we retest, we are often not able to reproduce the original data; without reproducible data, the basis for implementing the remediation didn’t make sense. Garbage in, money out the door. Except with so much money on the line, it’s not a joke; and yes this really happens all the time.
So remember the importance of environmental data being reproducible before you make big decisions on cleanups. It’s really the same dictum as the old carpenter’s rule: measure twice, cut once.