and auditory stores ( Mayer and Moreno 1998). These dual coding or dual processing theories are compatible with notions of situation models, as part of Kintsch's ( 1998) broader theory of text comprehension.
In addition to fostering conceptual change, other aspects of the HMSS text design are supported by reading research. For example, illustrations are central to the text and designed to be interacted with, not merely to decorate. It is well known that illustrated text is more memorable than unillustrated ( Peeck 1989; Levie and Lentz 1982). The interactivity aspect between text and illustration should support dual coding of material into verbal and imagery stores, thereby fostering more accurate and memorable mental models of material.
Finally, hints are embedded into the computer text, often taking the form of questions, which is supported by reading research. For example, higher level questions can facilitate higher level and factual learning ( Andre 1979; Hamaker 1986). Unfortunately most questions that appear in science texts are rote level (e.g., Pizzini, Shepardson and Abell 1992) and the insertion of questions about illustrations in science texts can increase cognitive load in ways that appear to be detrimental ( Iding 1997).
In addition to addressing these design considerations, preliminary data was collected. There is a recorder built into the program that keeps a log of the time and type of user interactions. Data collection and review indicates that correlation with actual classroom topics being covered was high for the two high school classes observed. The instructor did not manage the student use of the computer. Most of the enrichment computer use was by boys in the coed class. Some of the interactions were rapid -- up to 15 per minute. The instructor commented that the program appeared easy for the students to use. Often four or five students would be gathered around the one computer in the classroom. Data is currently being collected from seven schools. We do not have reports yet from teachers who entered hypermedia content themselves. A content expert commented that it was a simple process to link hypermedia text and graphic examples and hints, as well as quizzes.
Considering the most effective format for constructivist computer multimedia science texts and addressing questions for effective incorporation of expert information are important components in the ongoing formative evaluation of the HMSS multimedia format. Questions related to authorship recognition and validity of information need to be examined. Truly, the ability of the student and teacher to add content to the multimedia text makes this a constructivist adventure.
Hamaker, C. ( 1986). "The effects of adjunct questions on prose learning". Review of Educational Research, 56, 212-242.