Measurement System Analysis – Is it important?

Just how reliable is your data? Everyday, devices are used to measure things and give us the data. Colleagues evaluate things and make decisions about goodness and badness. Then, all this information ends up in a computer where the data can be crunched and plotted using a software tool. We can then evaluate the results and make those all important data driven decisions. That’s what we want to do, but what if the data isn’t reliable?

To answer that question the often overlooked Measurement System Analysis (MSA) comes into play. We can all agree that using data and statistics to guide our decision making process makes sense. Assuring that our data has integrity is therefore highly important. That is just what a MSA can do for you.

MSA can be applied whether the measurements are taken with a device or assessed based upon some criteria. For the data to be considered reliable for use in a statistical analysis for making decisions it must be captured by a process that is repeatable and reproducible. Repeatable means that if I evaluate the same item multiple times I will get the same answer each time. Reproducible means that if you, me, and your best friend were to evaluate the same item multiple times we would all get the same answer each time.

The reliability of our data is very important because taking corrective actions or making data driven decisions can make or break an organization. It gets back to the old saying, “If it ain’t broke don’t fix it.” The MSA study does take a little time and some planning. It is basically a hypothesis test to determine if the measurement system can determine if there is a difference or not. We want to prove that all the variation in the data is associated with the differences in the items being evaluated, not in the measurement system that is being used for the evaluation.

Time and again we have found that the first place to take corrective action in a process is the measurement system itself. Be wary of data that came from the information system. Just how did that data get into the system in the first place? Make sure that time is aligned with the data. Measurement System Analysis is important, especially if you want to make good business decisions based upon data.