INTERACTING WITH HYPERTEXT - A METAANALYSIS OF EXPERIMENTAL STUDIES

Authors
Citation
Cm. Chen et R. Rada, INTERACTING WITH HYPERTEXT - A METAANALYSIS OF EXPERIMENTAL STUDIES, Human-computer interaction, 11(2), 1996, pp. 125-156
Citations number
61
Categorie Soggetti
Computer Science Cybernetics
Journal title
ISSN journal
07370024
Volume
11
Issue
2
Year of publication
1996
Pages
125 - 156
Database
ISI
SICI code
0737-0024(1996)11:2<125:IWH-AM>2.0.ZU;2-T
Abstract
The meta-analysis compared and synthesized the results of 23 experimen tal studies on hypertext. The analysis was based on 56 pairs of effect sizes and significance levels of the impact of users, tasks, and tool s on interactions with hypertext. This analysis focused on three facto rs that prevailingly influence the use of hypertext: the cognitive sty les and spatial ability of users; the complexity of tasks; and the str ucture of information organization and the visualization of the struct ure. The meta-analysis found that this group of experimental studies r eported significantly discrepant findings, indicating that substantial differences exist among individual experiments. Individual difference s in cognition did not yield enough evidence to conclude that the effe ct sizes are significantly apart from zero. The meta-analysis showed t hat the overall performance of hy pertext users tended to be more effe ctive than that of nonhypertext users, but the differences in efficien cy measures were consistently in favor of nonhypertext users. Users be nefited more from hypertext tools for open tasks. Overall, the complex ity of tasks has the largest combined effect sizes. Graphical maps tha t visualize the organization of hypertext have significant impact on t he usefulness of a hypertext system. This meta-analysis raised two iss ues concerned with the present hypertext literature: (a) the absence o f a taxonomy of tasks for analyzing and comparing hypertext usability across studies, and (b) the weaknesses of the connections between abst ract hypertext reference models and specific hypertext systems. These weaknesses may considerably undermine the significance of individual f indings on hypertext usability. Results of the meta-analysis suggest t hat the discrepancies among empirical findings are related to these we aknesses. Future work on hypertext usability should emphasize task tax onomies along with longitudinal and ethnographic studies for a deep un derstanding of the interactions between users and hypertext. Recommend ed research issues for the future are highlighted in Section 5.