Matching can reduce the precision of your experiment if
A) the variable on which you match is not highly correlated with the dependent variable
B) the variable on which you match is too highly correlated with the dependent variable
C) the matching variable and the dependent variable are the same
D) None of these are true. Matching always increases the precision of an experiment.

Question

Quizplus · Accepted Answer

Matching can reduce the precision of your experiment if the variable on which you match is not highly correlated with the dependent variable. In this case, matching on that variable would not effectively control for its influence on the outcome and could introduce additional noise into the analysis. Matching on a variable that is too highly correlated with the dependent variable could also reduce precision by essentially removing variation in the matching variable that is also associated with the outcome. However, matching can still be a useful technique when done correctly and with careful consideration of which variables to match on. Option C is not true because if the matching variable and the dependent variable are the same, then there would be no need for matching. Option D is not true because matching does not always increase the precision of an experiment; it depends on how the matching is done and which variables are matched on.

Matching Can Reduce the Precision of Your Experiment If