Examples of non-experimental (or observational) study designs are:
- cohort studies,
- case-control studies, and
|Cohort||Cohort studies , which are also known as longitudinal or prospective, focus on the development of a phenomenon like payment fraud.
Cohort members may be cardholders. They are studied over an extended period. Some members will be “exposed,” e.g., by being victims of identity theft, which is what the analysis aims to reveal.
On the other hand, a cohort can also measure an exposure to a positive outcome like that of repeat-purchases. The aim could be to measure the Client’s Lifetime Value (CLV), i.e., the total value of its repeat purchases over a period of time.
|Case control||Case-control studies  compare two groups: a group of cases that has been exposed to payment fraud, for example, and a group of controls that has not been exposed. These studies are common in machine learning (ML) and artificial intelligence (AI).|
|Surveys||Finally, surveys  assess whether the members of a group have been exposed to the phenomenon, e.g., payment fraud.
Alternatively, the survey could also aim to measure the intention to recommend a service to someone else, which is referred to as the Net Promoter Score (NPS).
In contrast to longitudinal studies, these studies are called cross-sectional because they are done at one point in time.
- A. Aschengrau and G.R. Seage. Essentials of Epidemiology in Public Health. Jones & Bartlett Learning, 2014.