How to read this graph:

Halves divide a dataset into two equal parts, with each part representing 50% of the total observations. For example, cities can be ranked by population size and divided into two equal groups, or havles. The first group, or half, would include the 50% most populous cities. The least populated 50% would fall into the last half.

Thirds divide a dataset into three equal parts, with each part representing 33% of the total observations. For example, cities can be ranked by population size and divided into three equal groups, or thirds. The first group, or third, would include the 33% most populous cities, the second third would include the next 33%. The least populated 33% would fall into the last third.

Quartiles divide a dataset into four equal parts, with each part representing 25% of the total observations. For example, cities can be ranked by population size and divided into four equal groups, or quartiles. The first group, or quartile, would include the 25% most populous cities, the second quartile would include the next 25%. The least populated 25% would fall into the fourth quartile.

Quintiles divide a dataset into five equal parts, with each part representing 20% of the total observations. For example, cities can be ranked by population size and divided into five equal groups, or quintiles. The first group, or quintile, would include the 20% most populous cities, the second quintile would include the next 20%, and so on. The least populated 20% would fall into the fifth quintile.