"Statistical Control" is a phrase which is very popular in business circles. What does it mean? Let us examine two types, the statistical control of the quality of a manufactured product, and the statistical control of a purchasing department. In the former case, over a sufficient period of time, records are made of the number of defects per 1000 (say) found in the product manufactured. From these records statistical standards are set up consisting of the averages and distributions of the occurrence of defects. By comparing any subsequent sample with such standards, the statistician is able to assert that the occurrence of defects in this sample is within or without the ordinary limits of random sampling. If without, the manufacturing department is notified of the excess of defects and is expected to remedy the situation. In the case of the purchase of commodities, the standard (price to be paid) is ill defined or lacking. In this type of control, the distinguishing feature in the estimation of the proper price to be paid is the collection and use of a group of statistical facts concerning correlated, independent variables. Some such are (1) the trend of prices in the commodity itself with due attention to seasonal and cyclical fluctuations; (2) the level of general commodity prices, (3) the facility of credit, etc. From such data, the statistician fixes an average price to be paid for the commodity and a proper time for its purchase.
Proceedings of the Iowa Academy of Science
©1930 Iowa Academy of Science, Inc.
Snedecor, George W.
"Statistical Control of a Grading System,"
Proceedings of the Iowa Academy of Science, 37(1), 290-293.
Available at: https://scholarworks.uni.edu/pias/vol37/iss1/68