Dot Plots and Histograms

A dot plot puts one mark above each value. Nothing is lost, so it works well for small data sets.

Histograms

A histogram groups values into intervals and draws a bar for each count. It handles large data sets that a dot plot could not.

Bin width matters

Too few bins hides the shape; too many leaves a jagged mess. The same data can look quite different under two bin widths.

Shapes

A distribution may be symmetric, skewed right with a long tail of high values, skewed left, or have two peaks.

Skew names the tail

Skewed right means the tail stretches right, even though most of the data sits on the left. The name follows the tail, not the pile.

Outliers

An outlier sits far from the rest. It may be an error worth removing or the most interesting value in the set, so investigate before deciding.

Reading the shape

Symmetric, skewed, uniform or bimodal. Skew is named for the direction of the long tail, which is the opposite of where most of the data sit — a naming that catches nearly everyone out at first.

The bin width changes the picture

The same data with wide bins can look like one lump and with narrow bins can show real structure. Choosing bins is a decision that affects what a reader concludes, and it should be made honestly.

A histogram is not a bar chart

Histogram bars touch because the horizontal axis is a continuous number line; bar chart bars have gaps because the categories have no order. The touching bars are a signal about what the axis means.

Dot plots keep every value

A dot plot shows each observation individually, so nothing is lost. That makes it ideal for small data sets, where a histogram's grouping would hide more than it clarifies.