Learning to write without writing: Writing accurate descriptions of interactions after learning graph-printed description relations

Abstract

Interpreting and describing complex information shown in graphs are essential skills to be mastered by students in many disciplines; both are skills that are difficult to learn. Thus, interventions that produce these outcomes are of great value. Previous research showed that conditional discrimination training that established stimulus control by some elements of graphs and their printed descriptions produced some improvement in the accuracy of students’ written descriptions of graphs. In the present experiment, students wrote nearly perfect descriptions of the information conveyed in interaction-based graphs after the establishment of conditional relations between graphs and their printed descriptions. This outcome was achieved with the use of special conditional discrimination training procedures that required participants to attend to many of the key elements of the graphs and the phrases in the printed descriptions that corresponded to the elements in the graphs. Thus, students learned to write full descriptions of the information represented by complex graphs by an automated training procedure that did not involve the direct training of writing.