The relative effects of pre‐ and postattention directing devices on learning from a “walk‐through” museum exhibit

Publisher Website
Google Scholar
Abstract
Twenty‐eight seventh and eighth grade students were randomly assigned to two treatments and a control as they entered the Florida State Museum. All subjects were given set induction materials saying in effect: (1) Treatment‐Study the wall panel which is part of the cave exhibit and describes the many biologic relationships and organisms in the cave at the cave entrance. Continue into the cave and study the habitat. When you exit the cave you will be given a test to see how much you learned; (2) Treatment II‐Enter the cave and study that habitat. When you exit at the other end, study the wall panel which is part of the cave exit and describes the many biologic relationships and organisms in the cave. After studying this panel you will be given a test to see how much you learned; (3) Control‐You will take a walk through exhibits depicting a variety of Florida Habitats and later see a film on the Habitats of Florida that are represented in the museum. Then you will be given a test to determine how much you know about one type of habitat. Both treatments and the control spent equivalent time at their tasks and were confronted with either the cave exhibit as described or the equivalent control experiences. A 25 item criterion test was designed to measure the acquisition of conceptual and factual knowledge and specific attentional factors relative to both the cave and the instructional panel. A Kr‐20 indicated that this instrument had a reliability coefficient of 0.80. It was anticipated that the panel, used as an attention directing and controlling device prior to entering the cave (Treatment I) or upon exiting the cave (Treatment II), would function as a forward‐shaping or backward review prompting adjunct, hence both should be significantly more effective than the control. This was confirmed (df 2,28; F 8.09) p < 0.01. It was also expected that the forward shaping panel would be more effective than the backward review panel in the tradition of inserted questions in prose research. Although the differences were in this direction, they were not significant (p≤ 0.10).