How do Histograms Work?
Histograms are a valuable tool for quality improvement, as long as you know how to use them properly. First, you have to pick a process that you would like to measure. This can be anything from number of items output per week, to the number of calls incoming per day. Basically anything that occurs over an extended period of time. You need to be able to collect at least 100 data values. The more data values, the better. After you collect all of your data, you need to assemble a table of data values. It's important to take into account the frequency of data values.
The next part in using a Histogram is to calculate some statistics so you can make a chart. You need to calculate the mean, minimum, maximum, standard deviation, class width, number of classes, skewness, and kurtosis.
Mean is the average of all values. Minimum is the smallest value. Maximum is the biggest value. Standard Deviation is how widely spread the values are around the mean. Class Width is the x-axis distance between the left and right edges of each bar in the histogram. Number of Classes is the number of bars in the Histogram. Skewness is the alignment of the Histogram. Kurtosis is a measure of the "peakness" of the distribution. After you calculate these statistics, you can create the actual histogram - it will take one of five following shapes:
- normal distribution
- positively skewed
- negatively skewed
- bi-modal distribution
- multi-modal distribution