What is a quasi-experiment?

A quasi-experiment is a research method similar to a classical experiment but without random assignment of participants to groups. In the context of A/B testing, this means that user groups are not formed randomly, which can lead to bias and reduced reliability of the conclusions.

For example, if you test a new feature only on users from one region and do not randomly assign all users, this would be a quasi-experiment. This approach is used when randomization is impossible or difficult, but it is important to consider the limitations in interpreting the results.