Histogram


Histogram is a quality control tool that organizes individual measured values in a data set according to the frequency (number or percentage) of occurrence.

  • similar to a column chart
    • but the key differentiator is the ability to show a frequency of values that are grouped by bins, or class intervals
  • gives you an immediate picture of the spread or dispersion of your data
  • Each bin is aligned on the x-axis
  • metric we want to assess against is listed on the y-axis
  • size or height of each column portrays the volume of the numbers in each bin
  • no space between the columns of a histogram
  • 4 formatting best practices:
    1. Each bin should be the same size
    2. Outliers must be included in the bins
      • consider the lowest number in your data set and the highest number, and then choose the size of your group accordingly
    3. Whole numbers should be used for both the start and end of the bin
      • help make the chart easier to read
    4. The number of bins should not be exceedingly large
      • ultimately depends on your volume of data
      • Excel uses Sturges’ rule (1926)
        • developed by Herbert A. Sturges
        • uses a simplified approach to determining the number of bins based on the number of observations (data points)
        • will create fewer, but wider, bins
        • useful when working with decently-sized data sets
      • automated bin-width rule developed by Freedman-Dianconis (1981)
        • good for skewed data
        • is resistant to outliers because it leverages interquartile range (IQR) and ignores extreme values
      • David W. Scott’s Rule (1979)
        • uses standard deviation to determine bin sizes
        • works well for data that is normally distributed
        • is sensitive to outliers