These studies lack the ecological validity or representativeness of the contextualized research described above, but they are a powerful aid in helping to derive design principles for constructing new applications. These studies differ from the "horse race" studies described earlier in that they do not try to compare two complicated combinations of technology, content, and instructional methods but rather to relate a specific feature of a technology to a specific outcome.
This kind of research can be illustrated with experimental studies of microcomputer-based laboratories (MBLs). As described earlier, MBLs combine a microcomputer-based program for graphing analog data with a set of external measurement devices (e.g., motion sensors, temperature probes) that can be used to collect the data. Previous research had found that junior high school students often misinterpret graphs, confusing the depiction of absolute level with the concept of change (slope). Mokros and Tinker (1987) found that students using MBLs reduced this type of error relative to students who collected data and graphed it manually. By comparing conditions in which two groups collected data through MBLs and saw it displayed either in real time or after a 20-second delay, Brasell (1987) was able to show that 90 percent of the improvement of the regular MBL group was dependent on their ability to see the data graphed in real time. Even a delay as brief as 20 seconds appears to seriously disrupt students ability to see the connection between graphed data and real-world physical events.
Spoehr (1992) followed up her original work on ACCESS with an experimental study contrasting two different versions of the database. She found that giving students a hierarchical structure for the corpus, through overview cards and labeling of the links between nodes, appears critical in producing deeper conceptual understanding of the content under study.
Zellermayer, Salomon, Globerson & Givon (1991) investigated the effects of Writing Partner, an interactive computer tool designed to help student writers by providing guidance regarding the metacognitive aspects of writing (e.g., attention to the characteristics of the intended audience). Secondary school students who wrote essays with guidance from this tool subsequently wrote better essays and showed evidence of having internalized the guidance when writing independently (without the computer tool). Students who wrote the same number of essays with regular word processing software or who used a version of Writing Partner that gave guidance only when it was requested showed no improvement.
This page was last updated December 27, 2001 (jca)