Which approach helps mitigate selection bias in behavior-change research?

• A. Randomization and control groups
• B. Using a nonrandom volunteer sample
• C. Collecting data from a single site only
• D. Not preregistering the study

Answer: (A)

Mitigating selection bias in behavior-change research relies on making groups comparable and ensuring that differences between them aren’t driving results. Randomization assigns participants to intervention or comparison groups by chance, balancing both known and unknown characteristics across groups. This reduces the risk that preexisting differences influence outcomes, so observed effects are more likely due to the intervention itself.

A control group provides a baseline for what would happen without the intervention, offering a counterfactual to compare against the treatment group. This helps separate the program’s impact from changes caused by selection or external factors.

Choosing a nonrandom volunteer sample introduces self-selection, meaning participants who volunteer may differ in motivation, health status, or other factors that affect outcomes, thereby increasing bias. Collecting data from a single site limits generalizability and can reflect site-specific characteristics that don’t represent broader populations. Not preregistering the study doesn’t address how participants are assigned or how groups are compared and primarily affects transparency and reporting practices rather than bias from nonrandom assignment.