I gathered the heights of the fourteen players on the Heat roster as well as their career steals per game. With these two individual data sets, I built two frequency distributions independent of each other. The first table is a frequency distribution of player heights. To determine the number of classes I used Sturgis’s Rule of Thumb, which indicated that there should be 5 in total.
Sturgis’s Rule of Thumb:
Round up [1 + 3.3xlog(number of data points)] to the highest integer
The number of data points is 15, so… 1 + 3.3xlog15 = 4.881101155 —> round this up to 5
Then I found the lowest data point and the class width. Of the fourteen players, the shortest height is at 73 inches (Eddie House) and the tallest is at 87 (Zydrunas Ilgauskas). The difference of these two heights, 14 inches, leads me to the class width: 3inches.
Round up [(difference of the highest and lowest data points) / (number of classes)] to the highest integer
The difference is 14, the number of classes is 5, so… 14/5 = 2.8 —> round this up to 3
With 5 total classes, a class width of 3 inches, and the lowest height starting height at 73 inches (6’11”), I made my first frequency distribution:
The second table is a frequency of career steals per game. Because I am using the same number of data points (15 players on the roster), the number of classes is again 5. Career steals per game on the miami heat roster range from 0 to 1.8*. So to the determine the class width, I just divided the difference of the lowest and highest data points by the number of class (1.8/5=0.36) and rounded that up to 0.4 career steals per game.
Because my sample (the Miami Heat roster) contains only fifteen data points, It is hard to make any significant inferences about all NBA players in the comparison of height and steal success.
*Dwayne Wade holds the highest number of career steals per game at 1.8, LeBron James follows closely with 1.7, and Dexter Pittman lags behind with 0 career steals in his entire NBA career.
Data Gathered Here