Overview

The index of qualitative variation, IQV, measures how evenly cases are spread across categories, such as survey answers or ethnic groups. It runs from 0, everyone in one category, to 1, perfectly even.

Enter the count for each category.

Worked examples

InputsResult
Count for each category12, 16, 14, 21, 19Index of qualitative variation (IQV): 0.99Simpson diversity baseline: 0.79
Show 2 more examples
Count for each category50, 1, 1, 1, 1Index of qualitative variation (IQV): 0.18Simpson diversity baseline: 0.14
Count for each category10, 10, 10, 10, 10Index of qualitative variation (IQV): 1Simpson diversity baseline: 0.8

How it works

Each category's share is its count divided by the total. Simpson's diversity, 1 minus the sum of the squared shares, measures spread. IQV rescales it by k ÷ (k − 1), where k is the number of categories, so the maximum possible spread equals 1.

The formula

IQV = k ÷ (k − 1) × (1 − Σpᵢ²)
kNumber of categories, 5 here
pᵢShare of cases in category i

Practical tips

  • Counts of 12, 16, 14, 21 and 19 give an IQV of about 0.99, very even.
  • Counts of 50, 0, 0, 0 and 0 give 0, with no variation at all.
  • Enter 0 for categories with no cases.

Frequently asked questions

How is IQV different from Simpson's index?

Simpson's index depends on the number of categories. IQV rescales it so 1 always means perfectly even.

Can I use percentages instead of counts?

Yes. Only the shares matter.

What if I have fewer than five categories?

The calculation assumes five, so the scaling will be off. Use Simpson's index instead.