As reliability is a central concept in classical test theory, *information* is central to modern test theory. The amount of information contained within an item reveals how much we learn about the trait of interest by observing a response to that item.

For example, say that we are interested in measuring the trait Extraversion. If we were to ask somebody "do you have many friends?", this should give us more information than if we would ask them "do you like to try new dishes?". The first question is more related to Extraversion, while the second is more related to Openness.

The total information for a test is calculated by summing the information for all of the items in the test. According to European standards, test information above 10 is deemed 'Excellent' and test information above 5 is deemed 'Good'.

There is a mathematical relationship between test information and reliability:

As can be seen from this formula, as the test information increases, so does the reliability.

There is also a relationship between test information and the Standard Error of Measurement (SEM) or uncertainty of the test:

โ

That is, as the test information increases, the uncertainty decreases.