W. Edwards Deming was one of America’s most famous statisticians. He was best known for the role he played after World War II in teaching the Japanese how to improve the quality of their products by monitoring and continually improving their production processes. In his book Out of the Crisis (1986), Deming warned against the knee-jerk (i.e., automatic) use of histograms to display and extract information from data. As evidence, he offered the following example. 

Fifty camera springs were tested in the order in which they were produced. The elongation of each spring was measured under the pull of 20 grams. Both a time series plot and a histogram were constructed from the measurements. They are shown in Figure 2.38, which has been reproduced from Deming’s book. If you had to predict the elongation measurement of the next spring to be produced (i.e., spring 51) and could use only one of the two plots to guide your prediction, which would you use? Why? 

Data from figure 2.38

Transcribed Image Text:

Elongation .0014 .0012 .0010 .0008 .0006 0 00 10 ဂစ "కూ పిలంలో ఆ ఆకల్ని 20 30 Order of manufacture 40 50 .0014 .0012 .0010 .0008 .0006