Final answer:
In the absence of random assignment, researchers compare preexisting groups using a non-experimental or observational design, which can be subject to confounding variables.
Step-by-step explanation:
When researchers are unable to use random assignment to divide participants among groups, they compare preexisting groups, which results in a non-experimental or observational design.
Experimental design involves having a control group and an experimental group, with the only intended difference between them being the factor under investigation or manipulation.
However, certain circumstances, ethical concerns, or the nature of the research question may preclude the use of random assignment, leading to the use of naturally occurring groups.
These observational designs can still provide valuable insights, but they often come with increased risk of confounding variables, which can limit the ability to draw strong causal conclusions.