Skip to main content
Fig. 24 | Journal of Big Data

Fig. 24

From: Cumulative deviation of a subpopulation from the full population

Fig. 24

Los Angeles County, reporting the number of people in the household, with scores being \(\log _{10}\) of the adjusted household income; \(n =\) 35,364; Kuiper’s statistic is \(0.06674 / \sigma = 9.605\), Kolmogorov’s and Smirnov’s is \(0.06495 / \sigma = 9.347\). Discerning the analogue of the dip in the plot of cumulative differences at the highest scores is possible yet difficult in the reliability diagrams, while being unmistakable in the cumulative plot; the reliability diagrams with enough bins do reflect the corresponding deviation at the highest scores, but are hard to interpret without the accompanying cumulative plot. The scalar summary statistics report very highly statistically significant deviation, largely since the number of observations from this largest county in California is so large

Back to article page