Using computerized procedures for testing and training abstract comparative relations

Agata Vitale; Veronica Cullan

Abstract comparative reasoning is involved whenever one arbitrary stimulus (i.e. not defined by its physical properties) is related to another in terms of qualitative or quantitative relations. This kind of reasoning is part of our everyday life, and it is the substratum for other kinds of more complex related reasoning skills, such as hierarchical relations. Previous studies indicated that normally developed adults find difficult to solve some comparative relations, even when these simply involve three elements (e.g. X> U; U<B). The current paper describes the development of automated training procedures for testing and training abstract comparative reasoning. The use of automated procedures was essential to the empirical work reported in this paper. It addressed problems of validity such as the risk that individuals’ performances might be affected by the presence of the experimenter. It also improved accuracy in recording individuals’ pattern of responses. Furthermore it facilitated the development of more effective training techniques to improve this kind of reasoning. In Experiment 1 participants were exposed to the same comparative relations among three stimuli (e.g. A, B & C) in three subsequent identical phases (Phases1, 2 &3). This was done to verify if they could improve their performance simply by repeated exposures to the same comparative relations. Experiment 2 was identical to Experiment 1, except that in phase 2 individuals were exposed to training (this involved a combination of non arbitrary trials and written feedback). The results indicated that the training was effective, and that individuals in Phase 3 of Experiment 2 performed better than those in Phase 3 of Experiment 1.

Computer programming, more than other fields, requires flexibility of thinking in abstract terms (including abstract comparative reasoning). It is therefore hoped that the automated procedures developed in Experiments 1&2 may provide the basis for similar techniques to test and train comparative relations in individuals who approach this field for the first time (e.g. students). 

Type of Publication: Paper
Conference: PPIG 2009 - 21st Annual Workshop
Publication Year: 2009
Paper #: 10
TitleUsing computerized procedures for testing and training abstract comparative relations
Publication TypePaper
AuthorsVitale, A, Cullan, V
PPIG Workshop: