GET-IT Glossary
  

Browse the glossary:

A    B    C    D    E    F    G    H    I    J    K    L   
M    N    O    P    Q    R    S    T    U    V    W    X    Y    Z

multiple statistical comparisons

— The application of many tests of the statistical significance of associations when analysing the results of a study

Full explanation:

Multiple statistical comparisons (many tests of association) on the results of a study increase the likelihood of one or more associations (e.g. between a treatment and an outcome) having a low p-value (probability of having occurred by chance) when, in fact, the association did occur by chance.

Multiple statistical comparisons increase the likelihood of attributing a difference to a treatment effect when chance is the more likely explanation. This mistake is sometimes referred to as a Type 1 error.

See also:

Type 1 error

Still don't get it?

If you feel that this definition hasn't helped you to understand the term, click on our monkey to let us know.


← multicentre study natural course of health problems →