Variance Reduction


Posted by Admin

Variance Reduction

Variance Reduction

When a researcher establishes a relationship between two variables in a study, he must take care before asserting that the effect is a causal one. It’s possible that some other extraneous or nuisance variable is causing the outcome. Consider the following three methods researchers can use to control for nuisance variables in a study.

History Threat

Researchers must show that some other event isn’t responsible for the study outcome. Research scholars call this the history threat. For example, while running a series of anti-smoking public service announcements in a city, a movie star shares that he’s just received a diagnosis of lung cancer from 40 years of smoking. Did the reduction in cigarette purchases come from the public service announcement or the reaction to the celebrity’s cancer? To control for this, the researchers shouldn’t do a single group study. The researchers should choose a similar city as a control group to rule out this threat to internal validity.