The effects that drive the variation that you are interested in assessing that have a fixed number of well-defined levels. They can also include nuisance variables that you need to consider in your model. Fixed effects include factors such as experimental treatment, disease status, age or developmental status of the test organisms, and gender. Variation due to fixed effects is the variation that you are interested in estimating and must be kept in the analysis.