What procedure is used in experimental research to reduce bias?

The procedure used in experimental research to reduce bias is random assignment. This method involves assigning participants to different groups or conditions in a way that is determined completely by chance. This ensures that each participant has an equal opportunity to be assigned to any group, which helps to minimize pre-existing differences between groups.

By using random assignment, researchers can effectively control for variables that might influence the outcomes of the study, allowing for a more accurate assessment of the relationship between the independent and dependent variables. This approach is fundamental in establishing causation since it helps ensure that any observed effects are likely due to the experimental manipulation rather than other confounding factors.

In contrast, the other options serve distinct purposes within research but do not directly address the reduction of bias through randomization. Control groups serve as a benchmark for comparing the experimental group, double-blinding prevents both participants and researchers from knowing which participants are receiving the treatment or control, which helps reduce expectancy effects, and sample stratification ensures representation from various subgroups but does not specifically involve random assignment to conditions.