The effects of graphical overviews, prior knowledge, and self-concept on hypertext disorientation and learning achievement.

JI. of Educational Multimedia and Hypermedia (2003) 12(2), 117-134

To counter user disorientation and cognitive overload, hypertext environments often incorporate navigational aids in the form of graphical overviews of the hypertext structure. The aim of the present experimental investigation of learning in hypertext environments was to examine the impact of such graphical overviews on learning achievement in conjunction with the respective influence of individual learner variables. In a sample of N = 82 students working on a hierarchically structured hypertext about the psychology of memory, how domain-specific prior knowledge and the self-concept of computer-related ability affect perceived disorientation and learning outcomes as a function of access to a navigational aid was tested. Participants provided with a graphical overview experienced somewhat less disorientation, particularly those with high prior knowledge. A high computer-related self-concept also helped to curb disorientation. Both prior knowledge and computer-related self-concept proved to be significant predictors of complex learning achievement (i.e., the acquisition of structural knowledge). Participants with high prior knowledge, a high self-concept, and access to the graphical overview achieved better retention scores (i.e., acquired more purely factual knowledge).

**********

Multimedia systems that present information as a network of nodes and links, rather than in a traditional linear sequence, have generated particular excitement in the context of learning with multimedia. Because they present material as part of an interconnected network, these hypertext and hypermedia systems are regarded as a particularly effective way of supporting self-directed learning. Hypertexts allow users to elect which links to follow, and to work their way through the material along various paths of their own (in accordance with their learning preferences, e.g., Plass, Chun, Mayer, & Leutner, 1998). Hypertext has been termed "non-sequential" (Nielsen, 1990a) or "non-linear" in structure to differentiate it from forms of text with a predetermined structure.

However, it has been shown that people navigating through hypertext environments may encounter problems of disorientation (lost in hyperspace; e.g., Conklin, 1987; Edwards & Hardman, 1989; Elm & Woods, 1985) and cognitive overload that prevent them from concentrating on the content of the text and thus inhibit deeper-level information processing. Disorientation involves problems such as difficulty in gauging the scope of a hypertext, uncertainty about the route taken to arrive at the present node, uncertainty about where one is in relation to the other nodes of the hypertext, difficulty in deciding which node to move on to next, and not knowing which nodes have already been visited. Clearly, navigating through hypertext makes additional demands of the cognitive system and calls for resources that--given the assumed limitations of the working memory--may result in learning deficits (e.g., Chandler & Sweller, 1991; Kim & Hirtle, 1995; Niederhauser, Reynolds, Salmen, & Skolmoski, 2000; Tripp & Roby, 1990).

To help preclude this kind of disorientation, hypertext authors often include navigational aids in their designs. As formulated in the "supplantation" hypothesis (Salomon, 1979) for media as a whole, navigational tools are of assistance to learners when they replicate the cognitive processes necessary for, or beneficial to, learning and problem solving. Ausubel (1960) proposed that conventional texts should be preceded by "advance organizers" illustrating the structure and contents of the text to follow in a more general form. Indeed, most authors regard graphical overviews of hypertext structure as indispensable navigational tools. Like tables of contents in linear texts, their function is to reproduce the structure of the hypertext, thus helping the user to construct a mental map of this structure (Simpson, 1989). Chen and Rada (1996) conducted a meta-analysis of seven experimental studies, and concluded that graphical navigation aids are better than simple textual overviews. Hierarchically structured overviews, in particular, appear to reduce perceived disorientation (Beasley & Waugh, 1995). However, numerous experimental studies have shown that access to this kind of structural information does not automatically help users to understand or retain material any better (e.g., Dias & Sousa, 1997; Reynolds & Dansereau, 1990; Stanton, Taylor, & Tweedie, 1992; Wenger & Payne, 1994). In fact, Jonassen and Wang (1993) found that participants who worked on a hypertext with a navigational aid were outperformed by members of a control group without a navigational aid.

As yet, little attention has been paid to the effects of individual learner variables on the use of navigational aids. It would seem reasonable to start by considering the role played by prior subject knowledge, a powerful predictor of learning achievement. Prior knowledge is generally understood to be the factual (declarative) and practical (procedural) knowledge that a person possesses in a given domain (e.g., physics, mathematics, psychology; cf. Alexander, 1992). Numerous studies on learning with conventional media and learning in schools have provided a wealth of evidence to show that domain-specific prior knowledge facilitates learning (e.g., Byrnes, 1995; Schneider, Korkel, & Weinert, 1989; Weinert & Helmke, 1995; for an overview, Dochy, Segers, & Buehl, 1999). Persons with prior knowledge in a given domain are equipped with topic-specific frameworks that help them to integrate new information into the prior knowledge base, for example, and serve as a structural aid to recall. With increasing prior knowledge, the individual is better able to establish links with the existing knowledge base and to "file" new information accordingly (Alexander & Judy, 1988). Learners with less prior knowledge appear to have more difficulty navigating through a given content area than learners with higher prior knowledge in that domain (McDonald & Stevenson, 1998). Shin, Schallert, and Savenye (1994) found that users with little prior knowledge benefited from limited access to the contents of a hypertext (as opposed to free access to all content areas). No such differences emerged for users with high prior knowledge, however, thus substantiating an aptitude-treatment interaction (Cronbach & Snow, 1977) between the two variables.

A second differential characteristic worth considering is the self-concept of computer-related ability. Self-concepts are defined as generalized domain-specific appraisals of one's own ability; they are acquired through experiences of competence in the respective domains (Helmke & van Aken, 1995; Marsh, 1986; Moller & Koller, 2001a). The particular significance of academic self-concepts has been confirmed by numerous studies showing that they can explain and predict achievement behavior. There is, for example, broad consensus that academic self-concepts foster learning processes in school (e.g., Byrne, 1996; Helmke & van Aken, 1995; Marsh, 1990; Marsh & Hattie, 1996; Moller & Koller, 2001b; Shavelson & Bolus, 1982; see also the meta-analytic findings of Hansford & Hattie, 1982). Furthermore, domain-specific academic self-concepts remain powerful predictors of school achievement, even when controlling for key variables such as intelligence or prior knowledge (Koller, Klemmert, Moller, & Baumert, 1999). In view of these findings, it would appear logical to consider the self-concept as a predictor of computer-based learning. Yet the contribution of the computer-related self-concept to learning achievement has, as yet, been largely overlooked in this context. Instead, research has focused on aspects such as gender differences in the computer-related self-concept (see the meta-analytical findings of Whitley, 1997) and the effect of computer experience on the self-concept (Smith, Caputi, Crittenden, Jayasuriya, & Rawstorne, 1999). Some studies have investigated a related concept: computer self-efficacy. In contrast to self-concept, however, self-efficacy can be defined as a person's subjective appraisal of their ability to succeed in a particular task (Bandura, 1986). Self-efficacy--also defined as the learned expectation of success in specific, recurrent tasks--is thus relatively task-specific (for a computer self-efficacy scale, see e.g., Compeau & Higgins, 1995). The self-concept, in contrast, is more generalized, and allows for predictions to be made across broader behavioral domains (for further differentiation of the two constructs, see Bong & Clark, 1999).

The aim of the following investigation into learning with hypertext was thus to test how domain-specific prior knowledge and self-concept of computer-related ability interact with access to a navigational aid in predicting learning achievement and perceived disorientation. It was assumed that high prior knowledge and high self-concept would reduce feelings of disorientation and facilitate learning. Likewise, access to a navigational aid was expected to curb disorientation and foster learning. Interaction effects between the learner variables and access to the navigational aid were of particular interest.

METHOD

Sample. A total of N = 82 students from the Universities of Kiel and Bielefeld (74% women; mean age M = 23.23, SD = 3.08) participated in the study, either as partial fulfillment of the experimental session requirement or in return for a small payment.

Material. The hypertext, which was prepared and hierarchically structured (e.g., Mohageg, 1992) using the computer languages Hypertext Markup Language (HTML) and JavaScript, deals with the topic of "Learning and Memory" (based on Wendt, 1998). The hypertext consists of 25 main sections constituting an introduction "Learning and Memory" and two main chapters, "Temporal Structure of the Memory" and "Content Structure of the Memory." It also includes three excursions: "Neuronal Anchoring of the Memory," "Localization of Memory Contents," and "Simple Learning in the Snail." The hypertext consists of 51 nodes, 11 figures, and a total of roughly 4,300 words.

In this study, two navigation conditions ("overview" vs. "no overview") were implemented, and a corresponding version of the hypertext prepared for each. Figure 1 shows an example of a typical content page from the hypertext environment. The hypertext structure takes the form of unidirectional embedded links. Buttons for the functions "backtrack" (to the previous read nodes; chronological backtrack), "overview" (or "start"), and "help" (with the meaning of the functions) are positioned at the head of each node. Each node is presented on a single screen page, meaning that no scrolling is necessary. Only one node can be displayed at a time. Should the content of a node run to more than one screen page, this is indicated below the buttons (e.g., "Page 1 of 2"). Participants navigate between these pages using "forward" and "back" buttons. All buttons are dynamic--they "light up" when the cursor is moved over them. A small text field also appears, providing key information on the function or figure in question. The nodes that contain a link to a figure also feature a "figure" button at the end of the page.

[FIGURE 1 OMITTED]

Experimental design and procedure. In addition to the experimental factor overview, the two learner variables prior knowledge and self-concept of computer-related ability were included as predictors. Participants were allocated to one of the two experimental conditions at random. The two semi-experimental measures were treated as continuous variables in the statistical analyses. The factual and structural knowledge acquired from the hypertext as well as the disorientation experienced while working on it were examined as dependent variables.

The experiment was conducted in one- or two-person sessions. Participants in two-person sessions were seated such they could not see each other. The experimenter was present in the room for the entire duration of the test. After a few words of introduction from the experimenter, a questionnaire tapping prior knowledge and computer-related self-concept was administered. The online part of the experiment then began with a short introduction to the basic principles of hypertext (participants could opt to skip this section), a description of the hypertext used in the experiment (illustration of the layout using excerpts from a node; explanation of the navigational functions including the "start" and "overview" buttons, etc.), and information about the experimental procedure. The introduction was also designed to familiarize less computer-literate participants with the mouse and the navigation techniques required (clicking on buttons and embedded links). The participants were then given a time limit of 35 minutes to work through the hypertext. This time limit made it possible to compare individual performances and to trace these back to the independent variables. Participants were instructed to study the hypertext carefully, and informed that they would be asked questions about it. Thirty minutes into the test, a window appeared on the monitor informing the participants that they had five minutes left. Finally, they completed a questionnaire measuring factual and structural knowledge and perceived disorientation.

Operationalization of Variables

Prior knowledge. The variable prior knowledge was operationalized at the beginning of the experiment in the form of an 11-item questionnaire with four response options (multiple-choice questions). This questionnaire tapped the participants' factual (declarative) knowledge in the domain "Learning and Memory" (example item: "Who was the first to investigate the relationship between forgetting and time?"; response options: (a) George Sperling (b) Hermann Ebbinghaus (c) Saul Sternberg (d) George Miller). One or more of the answers to each multiple-choice question were classified as correct, giving a total of 22 correct answers. The position of the correct answers was varied at random. One point was given for every correct answer marked as such. An alternative scoring approach would have been to allocate a point for every incorrect answer not selected by the testtaker. Because there was no difference in the results yielded by the two scoring schemes, however, we will refer only to the first variant in the results section. The reliability of the prior knowledge scale (Cronbach's [alpha]) was .73.

Self-concept of computer-related ability. The self-concept of computer-related ability was assessed as a generalized appraisal of one's own capacity to work with computers. It was measured using eight items (e.g., "I have difficulty understanding a lot of things that relate to computers," response options: from 1 = "strongly agree" to 4 = "strongly disagree"). The reliability of the scale (Cronbach's [alpha]) was .72.

Graphical overview. Approximately half the participants (n = 43) worked on the hypertext without a navigational aid; the other half (n = 39) were provided with a three-part navigational tool. This graphical overview (Figure 2) is hierarchically structured parallel to the structure of the hypertext (Beasley & Waugh, 1995). It allows users to move directly to a specific node of the hypertext by clicking on the respective title. Links that have already been visited appear in paler font. At the head of the page, there is a "backtrack" button and a button leading to the overview of main topics.

Learning achievement: Factual and structural knowledge. The items used to measure prior knowledge were used again to assess the variable factual knowledge. To minimize recognition effects, the order of the items was changed. The correct answers could be drawn directly from the text. Eleven further multiple-choice questions (again, each with four response options) were formulated to assess structural knowledge. To answer these questions, participants needed to be able to recognize relationships and draw conclusions (e.g., "What kind of learning is involved in operant conditioning?"; Response options: (a) implicit learning (b) explicit learning (c) associative learning (d) procedural learning). In total, there were 28 correct answers. The answers could not be retrieved directly from the text, but required a certain understanding of its content. Reliability was [alpha] = .72 for factual knowledge and [alpha] = .69 for structural knowledge.

[FIGURE 2 OMITTED]

Disorientation. Perceived disorientation was operationalized using a 10-item scale (following Beasley & Waugh, 1995; Non-Linear Media Disorientation Assessment) with four response categories (from 1 = "strongly disagree" to 4 = "strongly agree"). Example items for the disorientation scale include: "I often felt lost in the text" and "I was often unsure about where I'd already been." The internal consistency of the scale (Cronbach's [alpha]) was .87.

The age and gender of the participants were also recorded, as was their computer experience (eight items on a four-point scale; e.g., "I have already worked with hypertexts"; response options: from 1 = "never" to 4 = "often"). The reliability of the computer-experience scale was [alpha] = .81. Because computer experience proved to have no impact on the results of the present study, it will not be reported in the following.

RESULTS

The descriptive statistics summarizing the relations between the variables considered are presented first. Then regression analyses are used to examine the impact that the variables prior knowledge and self-concept have on disorientation and learning achievement in the "overview" and "no overview" experimental groups.

Descriptive results. Table 1 presents the correlations between the variables under consideration, as well as their means and standard deviations. Prior knowledge does not correlate with either self-concept or perceived disorientation, but as expected, there are strong positive correlations between prior knowledge and the variables factual and structural knowledge. Self-concept of computer-related ability displays significant correlations with structural knowledge and perceived disorientation, but only a weak relationship with factual knowledge. A higher level of disorientation was expected to be associated with lower levels of factual and structural knowledge, but the observed correlations--although negative--are weak.

No gender differences emerge for computer-related self-concept, prior knowledge, factual knowledge, structural knowledge, or disorientation, as shown by t-tests for independent samples (all t(80) < .23; ns). Likewise, the "overview" group does not differ from the "no overview" group in terms of either prior knowledge (t(80) = .11, ns) or computer-related self-concept (t(80) = .62, ns).

Level of disorientation. A linear regression analysis was performed to predict the level of perceived disorientation (Table 2). In addition to the experimental factor (0 = no overview; 1 = overview) and the two learner variables (prior knowledge; computer-related self-concept), all interactions (i.e., the z-standardized product terms) of the three variables prior knowledge, self-concept, and navigational aid were simultaneously entered as predictors in the equation. It emerged that these predictors together explain 18% of the total variance (see the results for the total equation in Table 2).

Access to the graphical overview and self-concept of computer-related ability both proved to be significant predictors of perceived disorientation. Participants who were able to consult the overview were less likely to feel "lost in hyperspace." Likewise, higher scores on the self-concept scale were associated with less disorientation. The variable prior knowledge had no significant main effect, and there were no further interaction effects. The results do reveal a marginally significant interaction between prior knowledge and the "overview."

To further analyze the marginally significant interaction effect "overview" X "prior knowledge," separate regression analyses were performed for the two experimental groups. Whereas the correlation between the participants' prior knowledge and their perceived disorientation was negative in the "overview" X group (r(39) = -.13), the corresponding correlation in the "no overview" group was positive (r(43) = .19). The difference between the two correlations is marginally significant (z = -1.41, p < .08).

Effects on learning. The variables factual and structural knowledge were analyzed using the same regression approach and the same predictors as above (see Table 2). As expected, prior knowledge was a significant predictor of both variables. In addition, self-concept proved to be a significant predictor of deeper-level comprehension of the text (structural knowledge), with a higher self-concept being associated with much better results. The two-way interactions did not yield any significant results. Although the three-way interaction ("overview" X "prior knowledge" X "self-concept") had no significant effect on structural knowledge, it did have a statistically significant influence on factual knowledge (b = .20). The predictors explained a total of 53% of the variance in factual knowledge (and 35% of the variance in structural knowledge; see Table 2 for the F-values). To further analyze the interaction effect on factual knowledge, the regression analyses were repeated separately for the "overview" and "no overview" groups. In the "no overview" group, only prior knowledge proved to be a significant predictor of factual knowledge (Table 3). In the "overview" group, however, there was also a significant interaction between prior knowledge and self-concept, and the self-concept variable only just failed to reach the level of statistical significance (p = .08).

To further investigate the significant interaction between prior knowledge and self-concept in the "overview" group, median dichotomization (Md = 9) was used to split the sample into two sufficiently different groups, one with low prior knowledge (M = 5.49, SD = 1.54, N = 39), the other with high prior knowledge (M = 11.05, SD = 2.03, N = 43). In the "overview" group with low prior knowledge, the participants' self-concept was unrelated to their factual knowledge (r(18) = -.05, ns). In the "overview" group with high prior knowledge, in contrast, higher self-concept scores were associated with higher factual knowledge scores (r(21) = .50, p < .05).

DISCUSSION

The aim of the present study was to investigate how domain-specific prior knowledge and self-concept of computer-related ability impact on perceived disorientation and learning achievement in a hypertext environment. To this end, participants worked through a hierarchically structured hypertext about the psychology of memory, either with or without a navigational aid. The findings emphasize the need to include potential learner variables such as prior knowledge and self-concept in studies relating to learning with hypertext. Based on the well-documented finding that domain-specific prior knowledge facilitates learning with conventional media and in the school context, it was hypothesized that a similar effect would be observed for students learning with hypertexts. Our results indeed confirm that participants with high prior knowledge performed much better overall than participants with low prior knowledge (cf., e.g., Niederhauser et al., 2000). This holds for both factual and structural knowledge. These results are consistent with the findings of studies on information retrieval (e.g., O'Donnell, 1993). Self-concept of computer-related ability proved to be a decisive factor in the prediction of structural knowledge. It was also expected that a higher self-concept would be associated with larger achievement gains in factual knowledge. However, the results did not substantiate this hypothesis. In this study, the computer-related self-concept evidently only had a direct effect when it was not a case of simply remembering facts, but of processing material at a deeper level. Presumably, the lower level of disorientation among participants with a high self-concept contributed to this result. This finding is consistent with theories explaining the effects of the academic self-concept on scholastic achievement--these effects are also attributed to increased intensity of processing (Helmke & van Aken, 1995).

The overview factor did not have a significant main effect on learning outcomes. This result is consistent with the results of those studies mentioned in the introduction (e.g., Wenger & Payne, 1994), which have shown that access to an overview does not automatically result in increased learning achievement. Still, the kind of global navigational aid used in the present study does not seem to have actually inhibited knowledge acquisition (as reported by Jonassen & Wang, 1993). Rather, it emerged that the acquisition of factual knowledge was largely dependent on the interaction of the overview condition with the two learner variables. Access to the navigational aid was only associated with improved retention in participants with high prior knowledge if they also had a high self-concept. However, the advantages reported for persons with high prior knowledge and high self-concept only held for factual knowledge. A deeper understanding of the subject matter, measured using the structural knowledge variable, was not achieved simply by providing users with a graphical overview. It may be that fostering this kind of knowledge requires a second tier of local navigational aids, presenting more detailed overviews of individual subsections of a text as and when required. Moreover, instead of using total scores only, it seems reasonable to suggest for further studies to include a distinction of macropropositions and micropropositions (van Dijk & Kintsch, 1983) for both factual and structural knowledge questions (see Hofman & van Oostendorp, 1999). Examining the advantages reported for persons with high prior knowledge and high self-concept, this distinction would allow exploring in more detail to which aspects of knowledge these advantages refer.

Significant effects also emerged for perceived disorientation. Participants with access to a graphical overview were particularly unlikely to experience disorientation if they had high prior knowledge (cf. Moller & Muller-Kalthoff, 2000). It was expected that better learning outcomes would be more clearly mediated by a lower level of disorientation. However, the self-reports show only weak correlations between disorientation and learning achievement. This may indicate a problem with the valid measurement of feelings of disorientation. Astleitner and Leutner (1995) have already drawn attention to the lack of an established scale for measuring the degree to which users are "lost in hyperspace." Proximal, objective measurements of behavior may be better suited to tapping disorientation than subjective measures like self-reports.

Further studies are needed to investigate whether the present findings hold for more complex navigational aids (global and local overviews; e.g., Chou & Lin, 1998; Utting & Yankelovich, 1989). More detailed navigational aids might conceivably lead to increased disorientation among persons with low prior knowledge, for example, but enable persons with high prior knowledge to build on the existing knowledge base. It is also important not to overlook the size of the learning environment when investigating the use of navigational aids (Gupta & Gramopadhye, 1995). Zellweger (1989) concluded that graphical overviews can be particularly helpful in smaller hypertexts. In more extensive hypertext environments, one is soon faced with the paradoxical situation of having to provide navigational aids to help users find their way around the many overviews (Dillon, McKnight, & Richardson, 1993; Nielsen, 1990b).

In evaluating studies on hypertext systems, it is thus necessary to consider a variety of factors that make it difficult to compare studies and draw general conclusions on the value of, for example, navigational aids. In addition to the size of the hypertext (which, at 51 nodes, was relatively small in the present study), the subject matter of the text is relevant. One of the main aspects to be considered is the design of the hypertext. This may vary in numerous respects including the hypertext structure, the type and number of navigational functions implemented, and the design of the navigational aid (type, structure, presentation, etc). The impact of potential learner variables, which have been largely overlooked in previous studies, should also be considered (Dillon & Gabbard, 1998). This will not only provide important insights into factors impacting on the learning process, but well-founded guidelines for the development of more user-friendly, adaptive systems (Brunken & Leutner, 2000). The results of the present study suggest that studies investigating the effects of navigational aids should consider both prior knowledge and the self-concept of ability as powerful predictors of learning achievement.

Table 1 Correlations, Means, and Standard Deviations (in parentheses) of
the Learner Variables Prior Knowledge and Self-Concept, and of the
Dependent Variables Disorientation, Factual Knowledge, and Structural
Knowledge

                Prior        Self-concept  Disorientation   Factual
                knowledge                                  knowledge

Prior           8.40 (3.33)    .18           .05             .69**
knowledge
Self-concept                 23.62 (3.64)   -.27*            .21

Disorientation                             20.10 (5.64)     -.12

Factual                                                    12.74 (3.67)
knowledge
Structural
knowledge

Notes: N = 82.
* p < .05, ** p < .01

Table 2 The Dependent Variables Disorientation, Factual Knowledge, and
Structural Knowledge as Functions of the Factor Overview and of the
Learner Variables Prior Knowledge and Self-Concept (linear regression)

                   Disorientation (a)  Factual        Structural
                                       knowledge (b)  knowledge (c)

Predictors           b        SE         b       SE     b       SE
Overview           -.43*     .21       -.14     .16   -.05     .19
Prior knowledge     .13      .11        .69***  .09    .50***  .10
Self-concept       -.36**    .11        .12     .09    .23*    .10
Overview--prior
knowledge          -.21 (+)  .11       -.08     .09    .12     .10
Overview--
self-concept       -.03      .12        .13     .09   -.03     .10
Prior knowledge--
self-concept       -.19      .13        .14     .10   -.03     .11
Overview--prior
knowledge--
self-concept       -.05      .13        .20*    .10    .05     .12

Notes: N = 82; b = unstandardized regression coefficients for predictors
entered simultaneously in the total equation. Standardized values were
computed for the dependent variables and the learner variables. The
coding of the experimental factor is given in the text.
(a) Result for the total equation: F (7, 74) = 2.35, p < .05
(b) Result for the total equation: F (7, 74) = 11.99, p < .001
(c) Result for the total equation: F (7, 74) = 5.67, p < .001
(+) p = .06, * p < .05, ** p < .01, *** p < .001

Table 3 Factual Knowledge as a Function of the Learner Variables Prior
Knowledge and Self-Concept (linear regression), Computed Separately for
the Two Experimental Groups

                                  No Overview (a)      Overview (b)

Predictors                          b         SE        b       SE
Prior knowledge                    .77**     .12       .61**   .12
Self-concept                      -.00       .10       .26     .14
Prior knowledge X self-concept    -.06       .10       .35*    .17

Notes: n = 43 ("no overview" condition); n = 39 ("overview" condition);
b = unstandardized regression coefficients for predictors entered
simultaneously in the total equation.
Standardized values were computed for the variables.
(a) Result for the total equation: F (3, 39) = 16.14, p < .001
(b) Result for the total equation: F (3, 35) = 11.71, p < .001
* p < .05, ** p < .001

References

Alexander, P.A. (1992). Domain knowledge: Evolving issues and emerging concerns. Educational Psychologist, 27, 33-51.

Alexander, P.A., & Judy, J.E. (1988). The interaction of domain-specific and strategic knowledge in academic performance. Review of Educational Research, 58, 375-404.

Astleitner, H., & Leutner, L. (1995). Learning strategies for unstructured hypermedia--A framework for theory, research, and practice. Journal of Educational Computing Research, 13, 387-400.

Ausubel, D.P. (1960). The use of advance organizers in the learning and retention of meaningful verbal material. Journal of Educational Psychology, 51, 267-272.

Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall.

Beasley, R.E., & Waugh, M.L. (1995). Cognitive mapping architectures and hypermedia disorientation: An empirical study. Journal of Educational Multimedia and Hypermedia, 4, 239-255.

Bong, M., & Clark, R.E. (1999). Comparison between self-concept and self-efficacy in academic motivation research. Educational Psychologist, 34, 139-153.

Brunken, R., & Leutner, D. (2000). Neue Medien als Gegenstand empirischer padagogischer Analyse: Stand der Forschung und Perspektiven. [New media as the subject of empirical pedagogical analysis: State of research and research perspectives.] In D. Leutner & R. Brunken (Eds.), Neue Medien in Unterricht, Aus- und Weiterbildung. Aktuelle Ergebnisse empirischer padagogischer Forschung (pp. 7-16). Munster: Waxmann.

Byrne, B.M. (1996). Academic self-concept: Its structure, measurement, and relation to academic achievement. In B.A. Bracken (Ed.), Handbook of self-concept (pp. 287-316). New York: Wiley.

Byrnes, J.P. (1995). Domain specificity and the logic of using general ability as an independent variable or covariate. Merrill-Palmer Quarterly, 41, 1-24.

Chandler, P., & Sweller, J. (1991). Cognitive load theory and the format of instruction. Cognition and Instruction, 8, 293-332.

Chen, C., & Rada, R. (1996). Interacting with hypertext: A meta-analysis of experimental studies. Human-Computer Interaction, 11, 125-156.

Chou, C., & Lin, H. (1998). The effect of navigation map types and cognitive styles on learners' performance in a computer-networked hypertext learning system. Journal of Educational Multimedia and Hypermedia, 7, 151-176.

Compeau, D.R., & Higgins, C.A. (1995). Computer self-efficacy: Development of a measure and initial test. Management Information Systems Quarterly, 19, 189-211.

Conklin, J. (1987). Hypertext: An introduction and survey. IEEE Computer, 20, 17-41.

Cronbach, L.J., & Snow, R.E. (1977). Aptitudes and instructional methods. New York: Irvington.

Dias, P., & Sousa, P. (1997). Understanding navigation and disorientation in hypermedia learning environments. Journal of Educational Multimedia and Hypermedia, 6, 17-185.

Dillon, A., & Gabbard, R. (1998). Hypermedia as an educational technology: A review of the quantitative research literature on learner comprehension, control and style. Review of Educational Research, 68, 322-349.

Dillon, A., McKnight, C., & Richardson, J. (1993). Space--the final chapter or why physical representations are not semantic intentions. In C. McKnight, A. Dillon, & J. Richardson (Eds.), Hypertext: A psychological perspective (pp. 169-191). New York: Ellis Horwood.

Dochy, F., Segers, M., & Buehl, M.M. (1999). The relation between assessment practices and outcomes of studies: The case of research on prior knowledge. Review of Educational Research, 69, 145-186.

Edwards, D.M., & Hardman, L. (1989). Lost in hyperspace: Cognitive mapping and navigation in a hypertext environment. In R. McAleese (Ed.), Hypertext: Theory into practice (pp. 105-125). Norwood, NJ: Ablex.

Elm, W.C., & Woods, D.D. (1985). Getting lost: A case study in interface design. Proceedings of the Human Factors Society, 29, 927-931.

Gupta, M., & Gramopadhye, A.K. (1995). An evaluation of different navigational tools in using hypertext. Computers and Industrial Engineering, 29, 437-441.

Hansford, B.D., & Hattie, J.A. (1982). The relationship between self and achievement/performance measures. Review of Educational Research, 52, 123-142.

Helmke, A., & van Aken, M.A.G. (1995). The causal ordering of academic achievement and self-concept of ability during elementary school: A longitudinal study. Journal of Educational Psychology, 87, 624-637.

Hofman, R., & van Oostendorp, H. (1999). Cognitive effects of a structural overview in a hypertext. British Journal of Educational Technology, 30, 129-140.

Jonassen, D.H., & Wang, S. (1993). Acquiring structural knowledge from semantically structured hypertext. Journal of Computer-Based Instruction, 20, 1-8.

Kim, H., & Hirtle, S.C. (1995). Spatial metaphors and disorientation in hypertext browsing. Behaviour and Information Technology, 14, 239-250.

Koller, O., Klemmert, H., Moller, J., & Baumert, J. (1999). Leistungsbeurteilungen und Fahigkeitsselbstkonzepte: Eine langsschnittliche Uberprufung des Internal/External Frame of Reference Modells. [Achievement feedback and academic self-concepts: A longitudinal test of the internal/external frame of reference model.] Zeitschrift fur Padagogische Psychologie, 13, 128-134.

Marsh, H.W. (1986). Verbal and math self-concepts: An internal/external frame of reference model. American Educational Research Journal, 23, 129-149.

Marsh, H.W. (1990). A multidimensional hierarchical self-concept: Theoretical and empirical justification. Educational Psychology Review, 2, 77-171.

Marsh, H.W., & Hattie, J. (1996). Theoretical perspectives on the structure of self-concept. In B.A. Bracken (Ed.), Handbook of self-concept (pp. 38-90). New York: Wiley.

McDonald, S., & Stevenson, R.J. (1998). Effects of text structure and prior knowledge of the learner on navigation in hypertext. Human Factors, 40, 18-27.

Moller, J., & Koller, O. (2001a). Dimensional comparisons: An experimental approach to the internal/external frame of reference model. Journal of Educational Psychology, 93, 826-835.

Moller, J., & Koller, O. (2001b). Frame of reference effects following the announcement of exam results. Contemporary Educational Psychology, 26, 277-287.

Moller, J., & Muller-Kalthoff, T. (2000). Lernen mit Hypertext: Effekte von Navigationshilfen und Vorwissen [Learning with hypertext: The impact of navigational aids and prior knowledge]. Zeitschrift fur Padagogische Psychologie, 14, 116-123.

Mohageg, M.F. (1992). The influence of hypertext linking structures on the efficiency of information retrieval. Human Factors, 34, 351-367.

Niederhauser, D.S., Reynolds, R.E., Salmen, D.S., & Skolmoski, P. (2000). The influence of cognitive load on learning from hypertext. Journal of Educational Computing Research, 23, 237-255.

Nielsen, J. (1990a). Hypertext and hypermedia. San Diego, CA: Academic Press.

Nielsen, J. (1990b). The art of navigating through hypertext. Communications of the ACM, 33, 296-310.

O'Donnell, A. (1993). Searching for information in knowledge maps and texts. Contemporary Educational Psychology, 18, 222-239.

Plass, J.L., Chun, D.M., Mayer, R.E., & Leutner, D. (1998). Supporting visual and verbal learning preferences in a second-language multimedia training environment. Journal of Educational Psychology, 90, 25-36.

Reynolds, S.B., & Dansereau, D.F. (1990). The knowledge hypermap: An alternative to hypertext. Computers and Education, 14, 409-416.

Salomon, G. (1979). Interaction of media, cognition, and learning. San Francisco: Jossey-Bass.

Schneider, W., Korkel, J., & Weinert, F.E. (1989). Domain-specific knowledge and memory performance: A comparison of high- and low-aptitude children. Journal of Educational Psychology, 81, 306-312.

Shavelson, R.J., & Bolus, R. (1982). Self-concept: The interplay of theory and methods. Journal of Educational Psychology, 74, 3-17.

Shin, E.C., Schallert, D.L., & Savenye, W.C. (1994). Effects of learner control, advisement, and prior knowledge on young students' learning in a hypertext environment. Educational Technology Research and Development, 42, 33-46.

Simpson, A. (1989). Navigation in hypertext: Design issues. Online Information, 12/89, 241-255.

Smith, B., Caputi, P., Crittenden, N., Jayasuriya, R., & Rawstorne, P. (1999). A review of the construct of computer experience. Computers in Human Behavior, 15, 227-242.

Stanton, N.A., Taylor, R.G., & Tweedie, L.A. (1992). Maps as navigational aids in hypertext environments: An empirical evaluation. Journal of Educational Multimedia and Hypermedia, 1, 431-444.

Tripp, S.D., & Roby, W. (1990). Orientation and disorientation in a hypertext lexicon. Journal of Computer-Based Instruction, 17, 120-124.

Utting, K., & Yankelovich, N. (1989). Context and orientation in hypermedia networks. ACM Transactions on Information Systems, 7, 58-84.

van Dijk, T.A., & Kintsch, W. (1983). Strategies of discourse comprehension. New York: Academic Press.

Weinert, F.E., & Helmke, A. (1995). Interclassroom differences in instructional quality and interindividual differences in cognitive development. Educational Psychologist, 30, 15-20.

Wendt, D. (1998). Lernen und Gedachtnis [Learning and memory]. Kiel: Institut fur Psychologie.

Wenger, M.J., & Payne, D.G. (1994). Effects of a graphical browser on readers' efficiency in reading hypertext. Technical Communication, 41, 224-233.

Whitley Jr., B.E. (1997). Gender differences in computer-related attitudes and behavior: A meta-analysis. Computers in Human Behavior, 13, 1-22.

Zellweger, P.T. (1989, November). Scripted documents: A hypermedia path mechanism. Proceedings of Hypertext '89, ACM, New York, 1-14.

University of Kiel 

Germany

[e-mail]