Chakravarty, Sadhan C. (1989, ASQC) Indian Statistical Institute, New Delhi, India
The word 'Quality' is considered in its broadest sense as 'the totality of features and chaacteristics of a product, process or service that bear on its ability to satisfy stated or impled needs' (ISO). Quality problems are those hurdles which surface from time to time or remain submerged while trying to achieve better quality more economically. These hurdles take the form of defects, wastage, rejection, rework, inefficiency, delay etc. And lead to poor profitability and economy in the long run.
In order to discover the root causes of these quality problems in the shortest possible time, there is need for a systematic approach against the trial-and-error method which is usually associated with enormous wastage of time and efforts. As in any learning process, the scientific approach to gain knowledge goes through the stages of (I) observation, (ii) formulation of hypothesis, (iii) prediction on the basis of hypothesis, (iv) verification of hypothesis through data and (v) reformulation of hypothesis if there is poor agreement between prediction from the hypothesis and observed facts. The method followed in solving quality problems in no different. The science of statistics plays a very important role in the different phases of the investigative efforts.
An attempt has been made in this paper to evolve a systematic procedure for solving quality problems which have been classified into distinct categories like (I) assessment of product/materials quality for acceptance/verification (e.g. assessment of process capability, lot quality, reliability etc.), (ii) poor level of product, process and service quality, (iii) high variability in quality characteristics, and (iv) high level of wastage of resources like time, money, materials, manpower etc. In individual cases, there may be overlapping of these categories. Although preliminary investigations follow almost identical paths in respect of all the different categories of problems, at some advanced stage they have to follow specific paths using specific tools depending upon the typical nature of process or problem.For investigation of any quality problem, the initial step by step approach for collection of necessary information (including data) and subjecting the data to analysis from different angles with the help of statistical and related tools have been detailed. It is also very important to understand whether the problem is one of 'Control' or 'Breakthrough' as defined by Juran.
If the problem is one of breakthrough, it may require the help of more sophisticated Statistical tools of Correlation & Regression Analysis, Design of Experiments and Multivariate Statistical tools like Discriminant Functions, Principal Components Analysis etc.
Many a time, solutions to problems belonging to the category of 'Wastages of resources' can be obtained through application of Operations Research Methods which help in making optimal decisions in the face of many alternatives and limited resources. There are well known techniques to solve problems of this type which may belong to one or more of the classes known as Programming, Queueing, Inventory, Sequencing, Scheduling, Replacement etc. The approach here generally involves
It must be noted that it is not enough to know the techniques to discover the root causes of problems. It is also very important that proper organizational support is given to the study team to smoothen out the rough edges in the course of its 'diagnostic journey' and subsequently during implementation of recommendations arising out of the studies.
Includes typical list of quality problems (departmentwise) and tools and techniques with brief description.
Human resources (HR)