Is there a difference between being in-control and in-spec? Yes there is, and it is a big difference. You can be in-spec but not in-control. And you can be in-control but not in-spec.
So, if you can have only one, which is better – being in-spec or in-control? Some may say “in-spec of course.” But unless the spec is very generous and forgiving, you will always be better off with a process that is in-control, but not completely in-spec. Obviously neither is the ideal situation – but there is a fundamental problem with a process that is in-spec, but not in-control. The problem is that you never know from one minute to the next if it will go out-of-spec. This means that you must inspect every part or product that is made – otherwise, how will you know that the product is still in-spec?
So, what is the advantage of a process that is in-control but not in-spec? Let’s assume that at least some of the product being made by this in-control process is in-spec. Given that some of the process output is in-spec and the process is in-control, we can actually statistically calculate how much of the product will be out-of-spec.
Of course, the goal is to have a process that is both in-control and in-spec – that is what we call a capable and centered process. What if you have a process that is in-control, but not in-spec? By using SPC, root cause analysis, and some common sense, an in-control process can usually be brought in-spec.
How do you know if your processes are in-control? A good first step to take is to create a histogram of your data and then analyze the pattern of variation to identify possible reasons for the pattern. A control chart is really just an extension of a histogram. Not only does it show you the pattern of variation, but it plots that pattern over time.