In fact, the methods of quality management can be divided into two categories: one is organizational quality management based on the idea of total quality management; The second is quality control based on mathematical statistics.
Organizational quality management method refers to the method of quality management from the perspective of organizational structure, business process and personnel working methods. It is based on the idea of total quality management, and its main contents include formulating quality policy, establishing quality assurance system, carrying out QC group activities, sharing quality responsibilities of various departments, and making quality diagnosis.
Statistical quality control is the starting point of the control chart first proposed by Dr. W.A.Shewhart of Bell Telephone Laboratory in the United States in 1924. There has been great development for more than half a century. Now these methods can be roughly divided into the following three categories.
(1) Primary statistical management method: also known as common statistical management method. It mainly includes seven QC tools (or seven technologies of quality control) such as control chart, causal chart, correlation chart, Pareto chart, statistical analysis table, data layering method and scatter chart. Using these tools, we can systematically collect all kinds of data related to product quality from the ever-changing production process, and use statistical methods to sort out, process and analyze the data, and then draw various charts, calculate some data indicators, find out the law of quality change and realize quality control. Kaoru ishikawa, a famous Japanese quality management expert, once said that 95% of the quality management problems in an enterprise can be solved by all employees from top to bottom using these seven QC tools flexibly. The implementation of total quality management is also inseparable from the mastery and flexible use of these tools by personnel at all levels and departments of enterprises.
(2) Intermediate statistical management methods: including sampling survey methods, sampling inspection methods, functional inspection methods, experimental planning methods, method research, etc. These methods are not necessarily mastered by all employees of the enterprise, but are mainly used by relevant technicians and people in the quality management department.
(3) Advanced statistical management methods: including advanced experimental planning methods and multivariate analysis methods. These methods are mainly used for complex engineering analysis and quality analysis, and usually only professionals can use them with the help of computer.
Here is a brief introduction to seven methods commonly used in primary statistical quality management, namely the so-called "seven QC tools", for the reference of netizens.
(1) statistical analysis table
Statistical analysis table is a tool for sorting out data and preliminarily analyzing the reasons by using statistical tables, and its format can be varied. This method is simple, but practical and effective.
(2) Data layering method
The method of data stratification is essentially the same, and the data collected under the same conditions are summarized for comparative analysis. Because in actual production, there are many factors that affect the quality change. If these factors are not distinguished, it is difficult to get the law of change. Data layering can be carried out in many ways according to the actual situation. For example, according to different time and shift, according to the type of equipment used, according to the feeding time and raw material composition, according to the inspection method and use conditions, according to different defective products and so on. The data stratification method is often used in combination with the above statistical analysis table.
The application of data layering method is mainly a systematic concept, that is, if you want to deal with quite complex data, you must know how to classify and summarize these data systematically and purposefully.
Scientific management emphasizes management skills to make up for the shortcomings of previous management based on experience and visual judgment. This management technology not only needs to establish correct ideas, but also needs to apply data to analyze the work and take correct measures.
How to establish raw data and collect these data according to the required purpose is also the most basic work of many quality control methods.
For example, the aviation market in China has become more and more fierce with the opening up in recent years. In order to win the market, airlines have not only strengthened various measures, but also made great efforts in service quality. We can often see customer satisfaction surveys on the plane. The survey was conducted through a questionnaire. The design of questionnaire is usually divided into ground service quality and aircraft service quality. The ground is divided into reservation and waiting; Aircraft can be divided into flight attitude, catering, sanitation and so on. Through these surveys, we can collect these data and get where to strengthen the quality of service.
③ pareto chart (Plato)
Pareto chart, also known as Plato, was named after Plato, the inventor of this graph and an Italian economist in the19th century. Plato first used pareto chart to analyze the distribution of social wealth. He found that 80% of Italy's wealth was concentrated in the hands of 20%. Later, people found that this law was observed in many occasions, so it was called Pareto Law. Later, Dr. Zhu Lan, an American quality management expert, extended Plato's statistical chart and applied it to quality management. Pareto diagram is a tool to analyze and find the main factors affecting quality. Its form is a double rectangular coordinate diagram, and the left ordinate indicates the frequency (such as the number of pieces, etc. ) The ordinate on the right represents the frequency (such as percentage). The dotted line indicates the cumulative percentage, and the abscissa indicates various factors affecting the quality, which are arranged from left to right according to the degree of influence (that is, the frequency of occurrence). By observing and analyzing pareto chart, we can grasp the main and original factors that affect the quality. In fact, this method is very useful not only in quality management, but also in many other management work, such as inventory management.
In the process of quality management, there are many problems to be solved, but we often don't know where to start. But in fact, as long as we can find out a few influential reasons, we can solve more than 80% of the problems. Plato systematically classified the projects (levels) according to the collected data, and calculated the data (such as defective rate and loss amount) and the proportion of each project, and then arranged them in order of size, plus the graph of accumulated value.
In factories or offices, losses such as inefficiency, defects and defective products can also be converted into more than 80% of the loss amount according to their causes or phenomena, which is the so-called Plato analysis.
Plato's use should be based on the items (phenomena) of the hierarchical method, and Plato can draw according to the statistics after the order adjustment.
Steps of Plato's analysis;
(1) The things to be handled should be classified according to the situation (phenomenon) or reason.
(2) Although the vertical axis can represent the number of pieces, it is best to express it strongly by the amount.
(3) Determine the period of data collection, from when to when. As the basis of Plato's data, the period should be as regular as possible.
(4) Items are arranged on the horizontal axis from left to right at half the size.
(5) draw a histogram.
(6) Connect the cumulative curve.
Plato's method (key control method) provides us with important things and key things under the condition that we can't cover everything, and these important things are not judged by intuition, but based on data and reinforced by graphics. That is, the hierarchical method provides a statistical basis, and Plato's law can help us grasp the key things.
(4) Causality analysis diagram
Causality analysis chart is characterized by results, with reasons as factors, and the relationship between them is represented by arrows. Causal analysis chart is a good way to fully mobilize employees' brains, find out the reasons and brainstorm, especially suitable for quality democratic management of working groups. When there is a quality problem and the reason is unclear, we can mobilize everyone to find the possible reasons for the problem, so that everyone can speak freely and list all possible reasons.
The so-called causal analysis diagram is to explain many reasons for a certain result in a systematic way, that is, to express the relationship between the result (characteristic) and the cause (factor) with a diagram. Its shape is like fishbone, also known as fishbone map.
There must be a reason for the formation of a certain result, so try to find out the reason by graphic method. Dr. Kaoru Ishikawa, a Japanese quality control authority, first put forward this concept, so the characteristic reason map is also called [Ishikawa map]. Causal analysis chart can be used in all stages of general management and work improvement, especially in the early stage of establishing consciousness, which is easy to clarify the cause of the problem and design the steps to solve the problem.
(1) To Analyze Charts with Results
Step 1: Gather relevant personnel.
Gather experienced personnel related to this problem, preferably 4- 10.
Step 2: Hang a big piece of white paper and prepare 2-3 colored pens.
Step 3: Members of the General Assembly speak on the causes of the problems, and the contents of the speeches are recorded on the map. Criticism and questioning are not allowed in the middle. (brainstorming method)
Step 4: The time is about 1 hour, and it will be over after collecting 20-30 reasons.
Step 5: As for the collected reasons, which one has the greatest influence, then everyone takes turns to speak. After consultation, those that have the greatest impact will be circled in red.
Step 6: As in step 5, if you think it is the most important, you can circle it with two or three circles.
Step 7: Draw another reason chart, and remove the ones that are not circled. Columns with more circles are the first.
Causal analysis chart provides a tool to grasp the important reasons, so participants should include people who have experience in this work, which is easy to be effective.
(2) Causal analysis diagram and Plato's usage
To establish Plato, it is necessary to establish the required purpose statistics step by step. The purpose of establishing Plato is to master several important projects that have great influence on the overall situation. Reuse the characteristic cause diagram to discuss the causes of these projects one by one, and take the improvement countermeasures. Therefore, the causal analysis diagram can be used alone or in combination with Plato.
(3) Re-analysis of causality diagram
We should find the root of the problem, so as to solve it fundamentally. After finding out the main reason of the problem, the experiment is analyzed by the method of experimental design, and the specific experimental method is put forward to find out the best working method. The problem may be completely solved, that is, solving the problem and preventing the problem.
Everyone, any enterprise has its goals, but in the process of pursuing the goals, there will always be many tangible and intangible obstacles, and what are these obstacles, how to form them, how to solve them and so on. , is the main concept of chart method of cause analysis.
A manager, in his management work within the scope of the pursuit of goals, if specific induction, we can know that from the project is not much. However, for every project you pursue, there are primary and secondary reasons that will affect your goal. These reasons are variables that prevent you from finishing your work.
How to list the pursued projects one by one, sort out the main and secondary reasons that affect the achievement of each project, express them with causal analysis diagram, and strengthen them in a planned way according to these reasons, which will make your management work more handy.
Similarly, with these cause analysis diagrams, even if there is a problem, it can be faster and more reliable in the process of analyzing the problem.
(5) Histogram
Histogram, also known as histogram, is the main tool to represent data changes. Histogram can be used to analyze the regularity of chaotic data, visually see the distribution of product quality characteristics, and the central value or distribution of data is clear at a glance, which is convenient to judge its overall quality distribution. When making histograms, some statistical concepts are involved. First of all, data should be grouped, so how to group reasonably is the key problem. Grouping is usually carried out according to the principle of equal group spacing. The two key numbers are the number of groups and the distance between groups.
(6) scatter plot
Scatter chart, also known as correlation chart, is to point out two variable data that may be related on a coordinate chart to indicate whether there is correlation between a pair of data. This pairing data may be the relationship between feature-cause, feature-feature and cause-cause. Through observation and analysis, we can judge the correlation between two variables. This problem is also very common in practical production, such as the relationship between quenching temperature and workpiece hardness during heat treatment, and the relationship between the content of an element in the material and the strength of the material. Although this relationship exists, it is difficult to express it with accurate formulas or functional relationships. In this case, it is very convenient to analyze with correlation diagram. Suppose there are a pair of variables X and Y, X represents a certain influencing factor and Y represents a certain quality characteristic value. Through experiments or collection, the data of X and Y can be represented by points on the coordinate map, and the correlation between X and Y can be judged according to the distribution characteristics of the points.
In our life and work, many phenomena and causes are regularly related, while others are irregularly related. To understand it, we can judge the correlation between them with the help of scatter plot statistics.
(7) Control chart
Control chart is also called control chart. The control chart was first put forward in 1924 by Dr. W.A.Shewhart of Bell Telephone Laboratory in the United States. Since then, control chart has become an important tool for scientific management, especially quality management. It is a graph with control boundary, which is used to distinguish whether the cause of quality fluctuation is accidental or systematic, and can provide information about the existence of systematic causes, so as to judge whether the production process is under control. According to the purpose, control charts can be divided into two categories. One is the control chart for analysis, which is used to analyze the change of quality characteristic value in the production process to see whether the process is in a stable and controlled state; The other is the control chart for management, which is mainly used to find out whether there is any abnormal situation in the production process and prevent unqualified products.
Statistical management method is an effective tool for quality control, but the following problems must be paid attention to in application, otherwise the due effect will not be achieved. These problems are mainly: 1) data errors. There may be two reasons for data errors. One is that the wrong data is used artificially, and the other is that statistical methods are not really mastered. 2) The data collection method is incorrect. If the sampling method itself is wrong, then the subsequent analysis method is useless no matter how correct it is; 3) The record of data is copied by mistake; 4) Handling of abnormal values. Usually the data obtained in the production process always contains some abnormal values, which will lead to incorrect analysis results.
The above summarizes seven common methods of primary statistical quality management, namely the so-called "seven QC tools", which embodies the characteristics of "judging and managing based on facts and data" in quality management. Finally, it should be pointed out that these methods seem simple, but it is not a simple thing to apply them correctly and flexibly in practical work.