Explain how Campbell and Fiske’s Multitrait-Multimethod Matrix (MTMM) provides evidence of construct validity. Give examples and draw a MTMM matrix.

Question

What will be an ideal response?

Answer 1

• Donald Campbell and Donald Fiske (1959) cleverly combined the need to collect evidence of reliability/precision, convergent evidence of validity, and discriminant evidence of validity into one study. They called it the multitrait–multimethod (MTMM) design for investigating construct validity.

• The researcher chooses two or more constructs that are unrelated in theory and two or more types of tests—such as objective, projective, and a peer rating—to measure each of the constructs.

• A multitrait–multimethod correlation matrix presents all the correlations between a group of different traits, attributes, or constructs, each measured by two or more different measurement methods or tests.

• There are four different pieces of information contained in the matrix.
o Heterotrait-heteromethod correlations: These are the correlations between two different (hetero) traits measured by two different (hetero) methods. The measures share neither a trait nor a method in common, so you would expect this correlation to be low relative to the other correlations in the table.
o Heterotrait-monomethod correlations: These are correlations between two different (hetero) traits measured by the same (mono) method. In this case, the correlations are measuring the relationship between different traits that were measured using the same method so the correlations should be moderately large. Even though the traits are different, the fact that both were evaluated using the same method of measurement can create a correlation between the scores. This correlation is often referred to as method variance, or sometimes method bias.
o Monotrait-heteromethod correlations: These are the correlations between the same (mono) traits measured by two different (hetero) methods. In this case, we are correlating two traits, which are both supposed to measure the same construct using two different methods. We would expect this correlation to be larger than most in the MTMM matrix. This is our evidence for convergent validity.
o Monotrait-monomethod correlations: These are the correlations between the same (mono) traits using the same (method). This is equivalent to correlating a test with itself, so it is really a reliability coefficient. Because a test should not be able to correlate more highly with any other variable than it does with itself, these correlations will be the highest in an MTMM matrix.