Describing the shape of graphical data

In describing a histogram, there are three main factors to take into account: the shape, the centre, and the spread.

Shape

The shape of a histogram can indicate whether there is a tendency for some data values to occur more than others, or not. A graph’s symmetry can be described as being approximately symmetrical, positively skewed, or negatively skewed.

Symmetrical

These are usually bell shaped or normally distributed. The mean of the data is usually equated to the median.

500px-Histogramme_proche.svg

 

Symmetrical distributions can also take in the form of two different peaks. In the case where there are two distinct peaks that are symmetrical about the dip, the distribution is referred to as being bimodal.

bimodal symmetrical

Positively Skewed

These graphs usually have the lowest values having the highest frequencies. In this case the mean will be greater than the median.

Histograma_alboratua_eskuin_01

Negatively Skewed

Negatively skewed data values have their highest data values with the highest frequency. Generally, the mean will be less than median in these data sets.

Histograma_alboratua_eskuin_02

Centre

There are three measures of centre that can be used to describe a histogram. They are the mean, median and mode, which can be inferred from a labelled histogram.

Spread

A distribution’s spread can be used to compare two distributions. The spread of the distribution is referred to as the maximum range of the distribution.

range = largest value – smallest value

The histogram below can be described as having a broad peak, or a wide spread.

wide spread narrow spread

On the contrary, the histogram above can be described as having a narrow peak.

See also: