Next, by looking at group E (1, 4, 4, 4, 7), the arithmetic average is 4. When one subtracts the arithmetic average from each value, the values are (1-4 = -3), (4-4 = 0), (4-4 = 0), (4-4 = 0) and (7-4=3). When one averages the size of deviation from the arithmetic average, the rough standard of variations can be obtained. But when one adds -3, 0, 0, 0, 3, then the solution is 0. Therefore, we square each value and average the values to get an average of 3.6 (by dividing by 5).
However, when the original unit is cm, it becomes cm2 by squaring, therefore we extract 3.6 and restore it, then the size of variations is √3.6 cm2≒1.89 cm. Therefore, group D is bigger than group E in terms of variations.
Below, I will examine the characteristics seen from the relational database of “Jingoro Sahashi” by using the standard deviation formula.
花村嘉英（2017）「森鴎外の『佐橋甚五郎』データベースとバラツキによる分析」より translated by Yoshihisa Hanamura