The study is quite straightforward. Two types of attribute misspecification (ie, exclusion of an essential attribute and inclusion of a superfluous attribute) were included within the framework of the rule space method. It was found from the simulations that exclusion of an essential attribute tends to lead to underestimation of examinees’ mastery probabilities while inclusion of a superfluous attribute leads to overestimates it.
The possible consequences of attribute misspecification were traced in knowledge states of the RSM. That’s fine. But they are not enough:
(1) The RSM consisted of four steps. Besides knowledge states (the second step), how to explain the possible consequence of attribute misspecification with respect to other three stages of the RSM?
(2) The consequences of misspecification were actually concluded from the simulation studies. But, can we predict the results according to the theoretical explanation of the changes in knowledge states because of attribute misspecificaition?