The theory of statistics is composed of highly abstract propositions that are linked in multiple ways. Both the abstraction level and the cumulative nature of the subject make statistics a difficult subject. A diversity of didactic methods has been devised to aid the student in the effort to master statistics, one of which is the method of propositional manipulation (MPM). Based on this didactic method, a corresponding assessment method has been developed. Basically, in using MPM for assessment purposes, the student is instructed to construct arguments using subsets of elementary propositions. In effect, the assessment procedure demands the student to display knowledge of the interrelationships between the propositions in a particular subset. Analysis of the student responses allows for scoring purely propositional knowledge, as well as conceptual understanding. In this paper we discuss research on the effectiveness of this assessment method, relative to assessment of conceptual understanding using concept mapping.
The CAUSE Research Group is supported in part by a member initiative grant from the American Statistical Association’s Section on Statistics and Data Science Education