There are two main types of medical research studies, Observational and Experimental. The key difference between observational study design and experimental design is how participants are assigned to a particular group. In experimental studies the researcher assigns the participants to a particular group (usually by using randomization). Unlike experimental research, in observational studies group assignment is controlled by “natural conditions.” In observational studies these natural conditions can include the study participant intentionally or unintentionally “assigning” themselves to a particular treatment or exposure. This group assignment can be based on personal preference, genetics, environment, social determinants and/or other factors. Researchers are essentially trying to be a “fly on the wall” without directly intervening in group assignment. They just want to OBSERVE what would have happened to the study participants naturally.
In an experimental research study the researcher may split participants into group A that receive a drug and group B that receives a placebo. In an observational study of this same drug the researchers are basically just observing people who would have taken the drug naturally. In this scenario you might observe patients who decided on their own volition to take the drug and compare them to a similar group of patients who are not currently taking the drug.
The experimental method is the “Gold Standard” for studying therapeutic interventions (treatments) or prophylactic interventions (prevention). The most common type of experimental research in medicine is the randomized controlled trial. Observational research is typically used to determine associations between health outcomes and exposures/risk factors. This can include studies on diagnosis, prognosis, etiology or harm. The most common types of observational research include cohort studies, case-control studies, case series, and cross-sectional studies.
Based on the higher amount of control the researchers have during experiments there is a higher level of validity and less bias than in observational studies. Well-designed experimental studies allow for us to determine causality (AKA this drug or exposure directly caused the change in outcomes). Alternatively, observational studies can only determine correlations between groups (AKA patients with a history of this exposure tend to have this outcome more often but we aren’t sure if the exposure caused the outcome). Observational studies have a much higher chance of confounding, because randomization cannot take place in these studies. Additionally, observational studies cannot be blinded.
However, there are some drawbacks to experimental research. These studies tend to require more resources (money) to perform. This also means it is not feasible to use experiments in order to study very rare outcomes since you would need a huge study population and that would be too expensive. In certain situations ethics also prevent experimental studies from being able to study exposures that are expected to cause harm. For example, if you wanted to study the effects of cigarettes you could not ethically assign patients to a group that required them to start smoking. By comparison some types of observational research are relatively cheap and fast. For example, if you wanted to study the effects of cigarettes you could likely find a data set that has already been collected by somebody else and just look at the group retrospectively. Since patients are getting exposures based on natural conditions you can also study certain exposures that are expected to cause harm without ethical concerns. If you are just observing somebody who decided to smoke on their own that isn’t an ethical concern like forcing a participant to smoke would be.