InfinityQS is absolutely committed to the accurate and correct application of statistical process control. We are so committed to statistical accuracy that we employ several degreed Industrial Statisticians to work closely with our clients and developers to ensure statistical validity and accuracy.
Our commitment to statistical methods and their correct use is why our control limits are created by specifying a mean and a standard deviation (sigma). Why? Because Shewhart Control Chart limits are ONLY supposed to be based on the correct mean and standard deviation for a given subgroup size, n. By requiring users to type in a mean and standard deviation, InfinityQS ensures that the control limits are actually based on this extremely important statistical requirement: control limits must be based on the MEAN, and the STANDARD DEVIATION for a specific subgroup size.
A serious statistical flaw of all of our competitors is that they actually allow users to type in control limits without regard to standard deviation, and without regard even to subgroup size. Therefore, these supposed "control limits" represent what? A "best guess" on the part of the user, or maybe they represent the engineering specification limits. Either way, typing in control limits without considering the standard deviation and subgroup size is simply erroneous. It is a violation of everything that statistical process control is about.
It might be simple, but typing in control limits is not statistically sound. Those typed-in "control limits" most likely do not represent the actual process variation (as evidenced by the standard deviation), unless the user actually knows what the standard deviation, mean and subgroup size are. And if the user actually knows the mean, standard deviation and n, why would they ever want to type in control limits? If they already know the mean, standard deviation and subgroup size, why not just type them in at one time? In fact, doing this is far more efficient. Think about it. If the user must type in an upper and lower control limit and mean for the xbar chart, they must do the same thing for the associated range or sigma chart. InfinityQS needs, for variables charts, only a mean, standard deviation and n. This is a far more correct, even efficient way of handling the creation of control limits. Related control limits are back-calculated automatically given the mean, sigma and n--even if the subgroup size varies.
How could our competitors have made the incredibly serious error of allowing users to type in control limits? Because most of our competitors do not have the depth of knowledge in statistical methods that InfinityQS possesses. In fact, few if any competitors actually employ a statistician--much less several Industrial statisticians.
It may be different compared to our competition, but InfinityQS uses a statistically valid and mathematically correct method of allowing users to specify control limits.
Back to top