Search

Browse

• Communities
• Authors
• Year
• Titles
• Subjects

My Space

• Login

Statistical quality assurance and control to evaluate analytical measurement system performance

Tsachouridis Angelos

Full record

---

Summary

Quality is an important factor when it comes to any product or service. With the high market competition, quality has become the main differentiator for almost all products and services. Quality control and assurance are essential in building a successful business that delivers products that meet or exceed the expectations of the customers. The quality of data must be known and established before it can be used in any application. Data quality may be judged on the basis of its quantitative accuracy and on the confidence that can be placed in the qualitative identification of the parameters measured. In order to be able to use the results generated from a measurement process with confidence, the fitness – for – use of these results must be evaluated. This can be achieved by regular application of the measurement process to quality control samples, as well as engaging the measurement process in regular interlaboratory testing programs (round robin studies). This requires the production of data in a quality control program that permits the assignment of its statistically supported limits of uncertainty. The essential features of such a program, which consists of statistical quality control techniques, are discussed in this project that focuses on how an overall statistical quality control procedure should be conducted, fully or partially, by any engineer and in almost every laboratory environment, based on certified standards. The program that was written in Matlab was used for the fulltime evaluation of a Rock – Eval pyrolysis apparatus and a GC – MS apparatus. In order to describe as best as possible the general statistical quality control design, a description of a Ruggedness test is also included and finally, the statistical procedure with which firstly outliers can be detected in the results of an Interlaboratory – round robin test, and after their discarding, the final precision statement (in terms of repeatability and reproducibility) is also described. The precision statistics of a round robin study are indicatively implemented by a small Matlab code on a series of interlaboratory data results related to the different properties of an oil lubricant sample.

Available Files

Services