Panel data models use individual effects to control for differences between subjects. These effects can be random or fixed, depending on whether they are related to the variables being studied. A test called the Hausman test helps decide between these two types of effects. If the test rejects the random effects model, it means the fixed effects model is more appropriate.