Introduction
Statistical quality control (SQC) or statistical process control (SPC) refers to a group of analytical instruments and techniques used by process engineers to guarantee the quality of a process or product. Statistical quality control helps process managers to evaluate the quality of manufacturing course and solve emerging problems (Evans and Lindsay 34). Evans and Lindsay define statistical process control as “the use of statistical methods in monitoring and maintaining the quality of products and services” (30). The techniques of statistical quality control contribute significantly to analyzing the quality of a product or service.
There are three techniques of statistical quality control. They include “process control, acceptance sampling, and hybrid procedures” (Evans and Lindsay, 35). Statistical quality control is mostly applied to manufacturing processes. Statistical process control is aimed at monitoring the quality of a process and ensuring that it meets predefined standards (Mastrangelo, Runger, and Montgomery 204). Statistical quality control does not enhance an inadequately planned product’s dependability. Instead, it helps to preserve the uniformity of the procedure used to manufacture a product, and therefore guarantee the consistency of the manufactured goods.
The success of every company depends on the quality of the products or services it delivers. Hence, statistical process control is essential for all organizations. It not only helps to enhance the quality of products and services but also to establish a quality-driven culture. The primary objective of SQC is to guarantee, in a cost-efficient way, that the goods delivered to consumers are of the right quality (Mastrangelo, Runger, and Montgomery 205). Scrutinizing all goods is bungling and expensive. However, the effects of delivering faulty products can be severe. Therefore, statistical quality control enables manufacturers to improve the quality of the products, therefore enhancing their relationship with customers.
The inception of Statistical Quality Control
Quality control has been in existence for a long time. One would not be wrong to allege that quality control came as a result of competition in the business world. Customers buy goods based on their quality. They frequently evaluate the quality of comparable products before purchasing (Mastrangelo, Runger, and Montgomery 207). Therefore, whenever a company realizes that its products are not doing well in the market, it tries to improve its quality or lower the prices.
Companies strive to improve the quality of products as well as steps used to manufacture the goods. Today, there are institutions that are mandated with ensuring the quality of products and services offered to consumers (Mastrangelo, Runger, and Montgomery 207). Thus, quality control has been in existence since the era of the industrial revolution.
Conversely, statistical quality control is a relatively novel concept. The application of statistical approaches in product development dates back to only two centuries. However, it was during the 20th century that most organizations started using statistical quality control to improve their operations (Mastrangelo, Runger, and Montgomery 208). Initially, statistical quality control was only applied in physics and astronomy. It was also sparingly used in social and biological sciences. Evans and Lindsay argue, “It was not until 1920s that statistical theory began to be used effectively to quality control as a result of the development of sampling theory” (61).
Walter Shewhart was the first scientist to use the lately discovered statistical approaches to handle issues of quality control. In 1924, Walter developed a draft of the contemporary control chart. He kept building on his draft, and in 1931, he wrote the first book on statistical quality control. The book led to the adoption of statistical methods in controlling and managing the quality of industrial processes. Later, Dodge and Romig used Walter’s ideas to come up with a way of using statistical theory to analyze samples. Their work gave birth to the modern use of statistical quality control to enhance industrial processes and improve the quality of goods and services.
Benefits of Statistical Quality Control
One of the benefits of statistical quality control to organizations is that it reduces the cost of inspections. Statistical process control uses sampling techniques that are cheap. Consequently, it not only reduces the cost of control but also the cost of production. As aforementioned, statistical control methods entail acceptance sampling that ensures companies use quality raw materials (Mastrangelo, Runger, and Montgomery 208).
Hence, it eliminates cases of organizations having to repeat an entire production process due to the use of inferior materials. Moreover, statistical quality control involves process management. Therefore, it enables workers to monitor production procedures and make necessary changes whenever they suspect that a process is not running as expected. Another benefit of statistical process control is that it boosts the profit of a company.
Statistical process control minimizes the rejection of products by enabling a company to manufacture standard goods (Evans and Lindsay 70). Consequently, it boosts a company’s sales volume, therefore adding to its profit. Every company manufactures products with the hope that it will sell them and make a profit. However, at times, companies do not make considerable sales due to competition and the quality of the products. Unlike in the past when customers considered the cost of goods before buying, today, customers go for quality. They prefer purchasing quality products at a high cost to buying inferior goods at low prices. Hence, they are discriminating when it comes to quality of products.
Statistical quality control enhances employee efficiency and creates quality consciousness within an organization. It instills the culture of excellence in workers (Evans and Lindsay 72). Therefore, they make sure that all manufacturing processes are discharged in line with designed quality specifications. In return, it enhances organizational efficiency by eradicating cases of defective operations. Moreover, statistical quality control improves employee commitment and boosts their productivity. Workers feel motivated when they engage in productive activities (Evans and Lindsay 70). Statistical process control minimizes defects and variations, thus reducing chances of process failure. The success of a production process motivates employees and saves a company from avoidable costs.
Another benefit of statistical quality control is that it builds the image of a company and improves customer relations. The utilization of statistical process control results in achievement of preset quality of the products. By extension, customers get the preferred quality of products. The goal of every customer is to get value for his/her money. Hence, they like to associate with enterprises that offer excellent products (Mastrangelo, Runger and Montgomery 209).
Statistical quality control augments the goodwill of a company among the consumers by boosting the quality of its products. Moreover, SQC ensures continuous and uninterrupted manufacture of products by avoiding the breakdown of machinery or interruption of production course. Statistical quality control helps in troubleshooting. Consequently, organizations are assured of unrestricted production.
Limitations of Statistical Quality Control
Statistical process control advocates timely discovery and deterrence of problems, therefore enabling organizations to manufacture quality products (Mastrangelo, Runger and Montgomery 210). Nevertheless, it is hard to implement in manufacturing setting since it requires a lot of time. Operators need adequate time to observe processes and complete flow charts. Statistical quality control is fitted in the production system of an organization.
Consequently, an organization needs time to train workers on how to use the system and also enlighten other staff members. Ultimately, statistical process control leads to some operations being delayed so as to give workers time to focus on training sessions. Implementation and sustenance of statistical quality control program is a costly venture. Many organizations do not have experienced workers. Hence, they are forced to hire consultants to train their personnel. Moreover, organizations are supposed to adjust their operations so as to accommodate SQC system (Mastrangelo, Runger and Montgomery 210).
The adjustments do not come without an extra cost. Once a statistical quality control program is integrated into a company’s production system, the company begins to produce a high number of rejects. The challenge continues for some time until all production processes are streamlined. The increase in the number of rejects amounts to losses and affects the company’s profit margin.
Statistical process control can only be utilized in a manufacturing setting that allows workers to appraise production line machinery for possible departure from process norms. Even though SQC enables operators to detect when a production process does not conform to predefine standards, it does not indicate the extent of nonconformity. Consequently, it is hard for engineers to understand the magnitude of defect of the products (Montgomery et al. 80).
Montgomery et al. allege, “Statistical quality control does not give an integer number for the out-of-tolerance dimensions of product pieces, which would require precise measurements” (81). Therefore, it is hard for operators to know the course and extent of corrections they need to make. Statistical quality control relies on the information that workers collect through flow charts. Operators have to monitor regularly a production line so as to obtain accurate data. Thus, implementation of SQC system comes as an extra responsibility for workers. On the other hand, information gathered through charts may affect co-operation between employees.
A high number of rejects may lead to employers thinking that workers are unable to handle their assignments. Moreover, it may lead to some employees feeling demoralized and lacking confidence about their jobs. The management should enlighten workers on the role of statistical quality control to avoid mistrust. Failure to expound on the role and benefits of SPC may make some operators refrain from reporting errors (Montgomery et al. 83). As a result, an organization may incur avoidable costs.
Principles of Statistical Process Control
In manufacturing context, statistical quality control is based on the principle that the management of quality enhances production of salable products. Industries manufacture products with the sole purpose of selling them. Consequently, they endeavor to make products that are appealing to customers. Industries achieve this by ensuring that they monitor and control production processes to guarantee quality products. Hence, operation managers believe that statistical process control have the capacity to improve the sales volume of a company since it facilitates production of superior goods (Montgomery et al. 85).
Moreover, statistical quality control is governed by the belief that it lowers production costs and improves distribution. Institutions maintain that control of quality decreases production costs. Statistical process control enables companies to consider quality when designing products and processes. Besides, it allows organizations to identify potential challenges before a production process commences. Hence, it not only helps to avoid cases of process disruptions, but also costs associated with poor products. Although production operations observe high standards of quality, machines habitually wear out, workers make blunders, and acquired materials have deficiencies (Montgomery et al. 87).
These factors can lead to a company manufacturing inferior products. Fortunately, statistical quality control comes in handy in case of such challenges. It enables operators to make adjustments to a production process, therefore enhancing the quality of the final products.
A company cannot enhance the quality of its products after a production process is over. Therefore, statistical quality control is governed by the principle that an organization should ensure that a production process conforms to products’ design specifications. Companies should avoid using non-conforming materials to manufacture products (Montgomery et al. 88). Sorting of finished products does not enhance conformity (Oakland 111). Therefore, it is imperative to use appropriate production process and materials.
Statistical process control should be aligned to the circumstances at hand. To achieve this, organizations have to observe certain basic principles of statistical quality control. First, organizations should recognize that inconsistency is present in every rhythmic procedure. Statistical quality control enables organizational leaders to establish the nature of the possible or predictable variance of a system (Oakland 114). Therefore, it helps leaders to separate disproportionate differences that are as a result of unavoidable causes from those that are as a result of avoidable causes. The causes may then be analyzed and used to correct or streamline an operation. In other words, statistical quality control helps management to enhance the efficiency of subsequent processes by eliminating assignable causes and trying to overcome unassignable causes.
Another assumption that governs statistical process control is that it is hard for a production procedure to manufacture goods that “remain consistently within their range of chance variation” (Oakland 116). For a company to produce uniform products, it has to apply statistical quality control techniques. Therefore, champions of statistical quality control maintain that it contributes to the quality and consistency of products. In addition, there is the opinion that quality should be ingrained into a product.
Quality assurance specialists believe that inspection cannot add quality to a product. Hence, they advocate the use of statistical quality control techniques in the initial stages of a production process to guarantee that a process focuses on quality. The specialists maintain that it is not the quantity of a product that matters, but the quality. Statistical quality control is based on the opinion that operations cannot be efficient without of quality control. Hence, it enhances organizational efficiency.
Statistical Quality Control Techniques
Process Control
The foundation of process control is the assumption that all manufacturing processes are vulnerable to a certain degree of deviation because of inevitable grounds that are referred to as chance variations. Qin claims “When fluctuations in the measurements are within the range of chance variation, the process is said to be in a state of statistical control” (481). Such process exhibits a singular mode that is symmetrical or relatively skewed. A distribution of this kind of a process can be distinguished by computing its variance and mean. Moreover, it can be identified using other measures of dispersion.
The process remains in a state of statistical control as long as no assignable causes disturb it. However, when vital assignable causes work to upset the process, measurements show excessively significant variations from the mean of observations (Qin 483). The control chart indicates these deviations by showing points that fall beyond the control borders. Production managers should determine the assignable causes and address them accordingly. The idea of control charts is widely used in mechanical processes. Control charts help to ensure that mechanical processes run within the established control. In case a process is found to be out of control, plant managers use control charts to make necessary adjustments and restore a process back to a level of statistical control.
A machine may be considered to have a “natural degree” of inaccuracy within boundaries at which it can run processes and produce desirable results (Qin 483). Therefore, the machine is said to be out of control if the magnitude of its error is beyond the established limits. In most cases, the limits of an error are determined based on the nature of a mechanical process. Hence, a process is monitored regularly to ensure that it runs within the set boundaries. Whenever the mechanical process is found to deviate from the established controls, necessary steps are taken to bring into control the mechanical process. Mostly, supervisors replace a machine with one that can run a process within the established statistical controls.
Acceptance Sampling
The characteristics of acceptance sampling are determined using two curves. The curves are “average outgoing quality (AOQ) curve and the operating characteristic (OC) curve” (Qin 484). Three variables are used to determine the OC curve of a single sampling plan. The variables are the sample size, the work lot size, and the acceptance number. Spots on the operating characteristic curve stand for coordinates of the likelihood of acceptance and the actual fraction error.
If the lot size is smaller than the sample size, the former has an impact on the nature of the operating characteristic curve. Besides, it affects the method of sampling from a limited sample. In most cases, the sample size is always smaller than the lot size. Hence, the sampling technique does not have significant impacts on the operating characteristic curve. Increasing the sample size enhances the ability to differentiate between good and bad producers or lots. However, it comes at an increased cost.
The average outgoing quality curve is used to evaluate the quality of a process. The curve has two variables, which are incoming and outgoing percent defectives. For a quality process, the incoming and outgoing class echelons are almost identical. It arises due to the limited number of lots in a sample that exhibit numerous defectives. In the industrial setting, acceptance sampling entails randomly analyzing multiple items from a batch or lot in order to determine whether to use the batch or discard it entirely (Qin 485).
Acceptance sampling is conducted to determine the level of incompetence of industrial materials. Qin alleges, “The distinction between statistical process control and sampling acceptance is that the latter is performed prior or after the process” (485). Acceptance sampling has no benefit if done during a process. The acceptance sampling performed before a process entails sorting and analyzing raw materials obtained from suppliers.
For instance, it may involve sampling metal castings to be used in factory or bales of clothes to be sold in a supermarket. On the other hand, the sampling conducted after a process entails randomly examining finished goods before they are introduced into the market (Qin 487). For instance, in a computer manufacturing plant, staff members may sample numerous machines from a batch to guarantee that they work as expected.
Hybrid Procedures
In a manufacturing setting, inspection of mechanical processes benefits from both acceptance sampling and process control. Hence, “the procedures are of a hybrid type in which there are both process control limits and lot rejection limits” (Stuart, Mullins and Drew 204). When a process that is being executed by a machine deviates from statistical control, supervisors conduct a random test of the things in the work lot. The procedure of testing and adjusting work lots, essentially within the boundaries of acceptance sampling, runs in spite of whether the industrial process follows the established statistical control or not. When results produced by an industrial process are outside the control limits, necessary steps are taken to determine the origin of the problem and restore the process to the correct statistical control.
Errors and defects that emanate from an industrial process affect the period and cost of a process. Moreover, they affect the quality of the final products. Unfortunately, it is hard to detect all the defects or errors. Therefore, process supervisors use intuitions and judgments to come up with statistical quality control measures (Stuart, Mullins and Drew 205). The supervisors have to specify the anticipated levels of quality and use statistical theory to make their judgments. Statistical theory helps supervisors to determine if a process meets the desired control level. In case it does not, they use analysis to make adjustments.
Process supervisors choose and evaluate statistical process control plans based on their costs and the ability to distinguish between “acceptable and unacceptable work lots or between the creators of these lots” (Woodall 345). If the prevalence of the producers and lots are satisfactory, the supervisors use sample verification to determine the principal cost. Otherwise, they compute the primary value using the measures adopted to enhance outgoing quality. In most cases, supervisors emphasize on process control as a way to guarantee the quality of the final results. They prefer preventing defects to correcting them. Consequently, most statistical quality control measures are designed to ensure that industrial processes run within the desired guidelines.
Tools for Statistical Quality Control
There are seven basic tools used in statistical process control. They include the histogram, cause and effect diagram, Pareto chart, control chart, defect concentration diagram, check sheet, and scatter diagram.
Pareto Chart
The Pareto chart is used to demonstrate different sets of problems graphically in order for operators to come up with a prioritization schedule. There are numerous varied features of a production process or system that require improvement. They include cost savings, time allotment, and number of faulty products among others. Each element constitutes numerous minute challenges, making it hard for operators to decide how to handle the problem. Stuart, Mullins, and Drew argue, “A Pareto chart or diagram indicates which problem to tackle first by showing the proportion of the total problem that each of the smaller problems comprise” (206).
Therefore, it enables operators to make a quick and informed decision when dealing with production challenges. A Pareto chart contains vertical bar graphs that arrange problems in an ascending order based on their importance. The tallest bar graph represents the most significant problems, and production operators ought to start by addressing these challenges. Nevertheless, it is import for operators first to determine the problems that might derail a company’s long-term goals. At times, the issues ranked in the tallest bar graph may not have adverse effects on long-term goals of an organization.
Check Sheet
Check sheets refer to charts used to collect data for a quality control program. If check sheets are prepared modestly and accurately, they assist operators to collect precise and relevant data. Moreover, the gathered information is decipherable and applicable (Stuart, Mullins and Drew 209). To ensure the accuracy of collected data, check sheets are supposed to be used by production line operators. Information obtained from check sheets is used to create histograms, which shows the effectiveness of a production procedure.
Histogram
Quality control experts use the information gathered through check sheets to create histograms. A histogram is a pictorial representation, which shows the outcome of a process. Alternatively, a histogram may depict the deviation of a product or production course. A bell-shaped curve indicates that a process or product does not deviate from the intended specifications (Woodall 347). The histogram allows workers to analyze and monitor what is taking place in a production process, therefore initiating necessary changes where they realize that the process is deviating from the set controls. A histogram has numerous benefits to process managers.
First, it helps managers to determine the reason a process or product deviates from the desired conditions (Woodall 347). Second, process managers can use a histogram to determine if a production course will yield the desired results. In case they notice any variation, they can take the necessary steps to return a process back to the defined statistical control. Finally, a histogram gives information about the improvements that a process needs so as to meet specifications. Hence, it improves efficiency since process managers do not spend a lot of time to analyze a process and decide on the necessary improvements.
Control Chart
Organizations do not achieve constant results in a majority of the processes (Woodall 349). There always have to be variations whenever the operations are repeated. Variability or fluctuation is an unavoidable feature of all processes and is projected to occur naturally as a result of diverse chance events. Nevertheless, a deviation that lies outside a steady model may be a sign that the process is not running in a reliable way. Control charts help production managers to detect processes that are out of control. The charts are designed to show the standard pattern that a process is expected to follow. Any point that lies either below or above the control path indicates that a process is out of control. Control charts help production line operators to expose variations and deviant fluctuations. Therefore, they help an organization to deal with the variation so as to minimize operation costs and maximize profit. The charts show the limits within which a process is expected to yield positive results.
Cause and Effect Diagram
The cause and effect diagrams are also referred to as fishbone diagram. They are used to “display the relationships between different causes for the effect that is being examined” (Qin 491). They facilitate preparation of a brainstorming process. Cause and effect diagram helps process engineers to recognize the numerous components of a problem. It enables the production line workers to understand how people, equipment, procedures, and raw materials influence the success of a process. The cause and effect diagram is an essential tool for determining the possible cause of a problem facing a production process. It brings together workers with diverse skills and enables them to dissect a process in a bid to determine the cause of the problem.
Scatter Diagram
Scatter diagram is another statistical quality control instrument that is used to analyze processes. Another name for scatter diagrams is correlation charts. They help production workers to find out the cause of variations in manufacturing processes. A scatter diagram comprises two variables plotted against each other on a Cartesian plane (Qin 497). Production workers use a scatter diagram to determine the correlation between different variables involved in a production process. Workers use information gathered through scatter diagrams to adjust variables that inhibit or alter the production course.
Defect Concentration Diagram
A defect concentration diagram comprises the image of a product. It displays all views of a product. The different forms of defects are then pinpointed on the image. In most cases, production workers use the image to determine the cause of errors (Woodall 157). They use information such as the position of the defect to figure out its cause. Defect concentration diagrams help organizations to determine the exact cause of the flaw, therefore rectifying the source of the problem rather than overhauling an entire process.
Conclusion
Statistical process control refers to a set of statistical techniques and instruments that specialists utilize to enhance the quality of products or processes. Statistical quality control enables production operators to examine the quality of a manufacturing process and address emerging variations. The idea of statistical process control is relatively new among the manufacturing companies. Formerly, statistical process control was used in physics and astronomy.
Today, many organizations use SPC to enhance the quality of their products or services. Indeed, statistical quality control has tremendous benefits to both manufacturing and service industries. One of the advantages of statistical process control is that it minimizes inspection costs. Moreover, it improves employee commitment and instills a culture of quality consciousness. On the other hand, statistical process control may stall employee involvement if not well managed.
Employers may interpret a high number of defective products or processes as a sign of negligence on the side of employees. Consequently, statistical process control may lead to employees losing confidence in their skills, and therefore being unable to handle assignments. In a manufacturing context, statistical quality control is based on the view that quality control guarantees production of salable products. Moreover, it is based on the belief that the quality of the products is more important than the quantity. The techniques used to control the quality of products or processes include acceptance sampling, process control, and hybrid procedure.
The three techniques enable operators to monitor and adjust production processes to meet the desired specifications. In addition, numerous statistical instruments are used to determine the causes of variations in processes and solve them.
Works Cited
Evans, James, and W. Lindsay. The Management and Control of Quality, New York: West Publishing Company, 2012. Print.
Mastrangelo, Christian, George Runger and Douglas Montgomery. “Statistical Process Monitoring with Principal Components.” Quality and Reliability Engineering International 12.3 (1996): 203-210. Print.
Montgomery, Douglas, Dert Keats, George Runger and William Messina. “Integrating Statistical Process Control and Engineering Process Control.” Journal of Quality Technology 26.2 (2007): 79-88. Print.
Oakland, John. Statistical Process Control, Oxford: Butterworth-Heinemann, 2008. Print.
Qin, Joe. “Statistical Process Monitoring: Basic and Beyond.” Journal of Chemometrics 17.9 (2003): 480-502. Print.
Stuart, Michael, Wamonn Mullins and Eileen Drew. “Statistical Quality Control and Improvement.” European Journal of Operational Research 88.2 (2007): 203-214. Print.
Woodall, William. “Controversies and Contradictions in Statistical Process Control.” Journal of Quality Technology 32.4 (2000): 341-350. Print.
Woodall, William. “The Statistical Design of Quality Control Charts.” The Statistician 34.2 (2004): 155-160. Print.