Expository hypertexts may contain specific types of signals such as navigable topical overviews and hyperlinks that map conceptual relationships between text contents. Two experiments with German university students (N ϭ 130, 75% female, mean age 25 years) were conducted to test the hypothesis that hypertext-specific signals particularly support learners with badly routinized reading skills in organizing and integrating complex learning materials. The experiments were based on naturalistic texts and essay-writing tasks typical for exam preparation. Learning outcomes were measured by characteristics of participants' essays (amount of knowledge, knowledge focusing, knowledge integration). In both experiments, a hypertext with a high amount of signaling yielded better learning outcomes than did a linear text for readers with a low level of skill, whereas there were no differences for readers with a high level of skill (⌬R 2 from .03 to .08 for the interaction). In Experiment 2, the same interaction pattern was found for hypertext with a high versus a low amount of hypertext-specific signals (⌬R 2 from .04 to .10). Moreover, a lack of signals led to less efficient navigation behavior. These results demonstrate that hypertexts equipped with hypertext-specific signals may compensate for deficits in reading skill.

Math self-concept ϭ ␤ 0ijk Constant ϩ ␤ 1 linear ability ϩ ␤ 2 quadratic ability ϩ ␤ 3 school-average abilityModel 5: The interaction of school-average ability and individualism-collectivism was added:Model 6: Economic development was added as a fixed effect:Model 7: The interaction between school-average ability and economic development was added:Math self-concept ϭ ␤ 0ijk Constant ϩ ␤ 1 linear ability ϩ ␤ 2 quadratic ability ϩ ␤ 3 school-average ability ϩ ␤ 4 economic development ϩ ␤ 5 economic development ϫ school-average interaction ϩ v 0k ϩ v 1k ϩ v 2k ϩ v 3k ϩ u 0jk ϩ e 0ijk .