Evaluating the Effectiveness of Classical and Bayesian Control Charts in Detecting Process Mean Shifts

The article aims to compare the efficiency of the traditional Shewhart average chart and the average chart using the Bayes method. Bayes relies on a dynamic technique to create the quality control limits and calculate the points drawn on this chart through the posterior distribution of parameters for normal distribution, which considers the initial information as a prior distribution and the sample observations as a likelihood function. Shewhart and Bayesian charts were compared based on average run length, standard deviation, the difference between upper and lower control limits, false alarm rate, and detection rate using simulation and real data representing the blood glucose levels of newborns at Valia Hospital in Erbil. The Bayesian chart was more accurate in detecting rapid blood glucose level changes, offering a higher detection rate and fewer false alarms compared to the Shewhart chart.