MGMT 382 Jeremy T. Navarre, Ph.D. Quality Control Chapter 10 Quality Control Quality control is a process that measures output relative to a standard that may or may not necessitate corrective action. If results are acceptable compared to the specified standard, no corrective action is required whereas results not meeting the standard yield corrective action(s). Quality Assurance & Control Quality assurance relies on inspection of lots, or batches, and is referred to as acceptance sampling. Quality control processes that occur simultaneous to production are referred to as statistical process control. Quality Control Quality control evaluates the quality of conformance related to a process. Specifically, quality control evaluates whether a process conforms to the intended design of the process. Statistical Process Control Statistical process control employs statistical analyses to evaluate if a process conforms to the intended design or, in contrast, is out of control and, consequently, requires corrective action. Process Variability Process variability may be derived from random, or common, variability, which is deemed acceptable based on specified criteria. Assignable, or special, variation includes variation derived from an identifiable and meaningful phenomenon or phenomena. Assignable Variation Statistical process control focuses on assignable variation, which is deemed inherently problematic, is identifiable, does not conform to the intended design and able to be corrected. Assignable variation is typically identified from a sampling distribution’s statistical characteristics. Central Limit Theorem The central limit theorem states that as the sample size increases, the distribution of sample means approaches a normal distribution regardless of the shape of the sampled population. As the sample distribution becomes increasingly concentrated, the likelihood that a sample statistic is close to the true value in the population is higher for larger samples relative to smaller samples. Statistical Process Control Control Process Define Define Characteristics to be Controlled Measure Define Measurement Procedures Compare Compare Measurement Results to Specified Standards or Level of Standard Evaluate Interpret Results Based on Established Criteria for Control Status Correct Employ Corrective Measures, or Don’t Monitor Sample and Monitor Effectiveness Statistical Process Control Statistical Process Control Central Limit Theorem Control limits, lower and upper, establish parameters that effectively differentiate between random and nonrandom variation. Notably, there exists the potential that Type I or Type II errors are prevalent. Type I and Type II Errors A Type I error occurs when a sample statistic indicates assignable, or special, variation is present when only random, or common, variation is present. A Type II error occurs when a sample statistic indicates assignable, or special, variation is not present when, in fact, assignable, or special, variation is present. Control Charts Control charts, mean and attribute, are useful tools to visualize how sample statistics relate to control measures, or parameters. Typically, control limits are referred to as either an upper control limit (UCL) or a lower control limit (LCL). Mean Control Limits Upper Control Limit (UCL) Lower Control Limit (LCL) An upper control limit is an upper bound above the average of sample means, 𝑥. A lower control limit is a lower bound below the average of sample means, 𝑥. UCL = 𝑥 + 𝑧𝜎𝑥 LCL = 𝑥 − 𝑧𝜎𝑥 where where 𝜎𝑥 = 𝜎/ 𝑛, or standard deviation of sample means σ = estimated process standard deviation n = sample size z = number of standard deviations yielding control limits 𝑥 = average of sample means 𝜎𝑥 = 𝜎/ 𝑛, or standard deviation of sample means σ = estimated process standard deviation n = sample size z = number of standard deviations yielding control limits 𝑥 = average of sample means P-chart Control Limits Upper Control Limit (𝑼𝑪𝑳𝒑 ) Lower Control Limit (𝑳𝑪𝑳𝒑 ) An upper control limit is an upper bound above the average fraction defective in a population, p. A lower control limit is a lower bound below the average fraction defective in a population, p. 𝑈𝐶𝐿𝑝 = 𝑝 + 𝑧𝜎𝑝 𝐿𝐶𝐿𝑝 = 𝑝 − 𝑧𝜎𝑝 where where 𝜎𝑝 = 𝑝(1−𝑝) , 𝑛 or standard deviation of sample distribution n = sample size z = number of standard deviations yielding control limits 𝑝 = estimated average proportion of defects in population, p 𝜎𝑝 = 𝑝(1−𝑝) , 𝑛 or standard deviation of sample distribution n = sample size z = number of standard deviations yielding control limits 𝑝 = estimated average proportion of defects in population, p C-chart Control Limits Upper Control Limit (𝑼𝑪𝑳𝒄 ) Lower Control Limit (𝑳𝑪𝑳𝒄 ) An upper control limit is an upper bound above the average occurrence of defects per unit in a sample. A lower control limit is a lower bound below the average occurrence of defects per unit in a sample. 𝑈𝐶𝐿𝑐 = 𝑐 + 𝑧 𝑐 where n = sample size z = number of standard deviations yielding control limits 𝑐 = occurrence of defects per unit in sample 𝐿𝐶𝐿𝑐 = 𝑐 − 𝑧 𝑐 where n = sample size z = number of standard deviations yielding control limits 𝑐 = occurrence of defects per unit in sample Application of Charts P-Chart (𝐔𝐂𝐋𝐩 & 𝐋𝐂𝐋𝐩 ) C-Chart (𝐔𝐂𝐋𝐜 & 𝐋𝐂𝐋𝐜 ) P-charts are prescribed when: C-Charts are prescribed when: Binary Observations Occurrences Per Unit are Counted 0,1 Arrivals Per Hour Pass/Fail Calls Per Hour Examples… Examples… Process Capability Process capability refers to the capability of yielding output within specified parameters, which is imperative for organizations producing products or providing services. Among the aspects of variability to be measured and evaluated are specifications, control limits and process variability. Specifications Specifications, also referred to as tolerances, indicate a range of values that individual output values must fall within. Examples… Control Limits Control limits are statistical limits that reflect the extent to which sample statistics, such as means and ranges, can vary due to randomness. Examples… Process Variability Process variability represents the natural, or inherent, variability of a process and is measured via standard deviation. Process Capability Processes may yield output that are within parameters prescribed by control limits while not adhering to specifications. In this case, the process may not be capable of meeting specified specifications. Determining whether a process is capable of adhering to specifications, a capability index is computed and evaluated. Process Capability Index, 𝑪𝒑 Process capability index is computed via: Process capability index, 𝐶𝑝 = 𝑠𝑝𝑒𝑐𝑖𝑓𝑖𝑐𝑎𝑡𝑖𝑜𝑛 (𝑠𝑝𝑒𝑐.) 𝑢𝑛𝑖𝑡(𝑠) 𝑝𝑟𝑜𝑐𝑒𝑠𝑠 𝑢𝑛𝑖𝑡 (𝑠) or 𝑢𝑝𝑝𝑒𝑟 𝑠𝑝𝑒𝑐. − 𝑙𝑜𝑤𝑒𝑟 𝑠𝑝𝑒𝑐. 6𝜎 𝑜𝑓 𝑝𝑟𝑜𝑐𝑒𝑠𝑠 𝑢𝑛𝑖𝑡(𝑠) Process Capability Index, 𝑪𝒑𝒌 If the process mean is not perfectly centered between the specification range, the modified capability index is computed via: 𝐶𝑝𝑘 = 𝑢𝑝𝑝𝑒𝑟 𝑠𝑝𝑒𝑐𝑖𝑓𝑖𝑐𝑎𝑡𝑖𝑜𝑛 − 𝑝𝑟𝑜𝑐𝑒𝑠𝑠 𝑚𝑒𝑎𝑛 3𝜎 and 𝐶𝑝𝑘 = 𝑝𝑟𝑜𝑐𝑒𝑠𝑠 𝑚𝑒𝑎𝑛 −𝑙𝑜𝑤𝑒𝑟 𝑠𝑝𝑒𝑐𝑖𝑓𝑖𝑐𝑎𝑡𝑖𝑜𝑛 3𝜎 Process Capability Improvement Process capability can be improved through a reduction in process variability, which may be accomplished through simplification, standardization and automation. Quality Control Questions…