Dot plots come in different variations.
First, we have the William Cleveland Dot Plot which is similar to a bar graph but uses position to encode data rather than height or length.
Second, we have the Unit charts which are also similar to bar charts—in the sense that they use length/height to encode data—but display each unit of measure as a stack of single mark or symbol instead of a continuous bar.
Third, there is the Wilkinson Dot Plot, which is a histogram where individual observations are displayed—using a dot or other symbol—on a continuous scale.
Finally, we have the Histodot plot—a name coined by Leland Wilkinson—to differentiate it from its more versatile sibling the Wilkinson Dot Plot. The difference between the two plots is that the latter displays the data on a continuous scale while the former uses simple histogram binning instead.
The graph below is a Histodot plot. It shows the distribution of the time required in minutes to complete various incoming calls. The data collected includes 150 random calls that were monitored and timed. The calls are rounded to the nearest minute.