MTMM experiments

In every round of the European Social Survey experiments are built in to evaluate
the quality of the questions by statistical estimation. Each of these experiments
consist of repeating at least three different questions in at least three
different ways.

This way one can separate how much of the variance of each
ESS question in the experiment is due to the ‘trait’ to which the question is
supposed to refer, and how much is due to the ‘method’ of asking it. Therefore
the type of experiment is also called ‘multitrait-multimethod’
(Campbell & Fiske 1959).

The program documentation
provides a more elaborate explanation of MTMM experiments with examples.