Science and Technology Education Center

Tel-Aviv University - School of Education

Knowledge Technology Laboratory







David Mioduser, Rafi Nachmias, Orly Lahav, & Avigail Oren





Tel-Aviv University

School of Education

Ramat-Aviv, Tel-Aviv, 69978



Research Report # 54




In Press in International Journal of Research in Computers in Education

[For internal teaching and research use at Tel-Aviv University]

[© by the journals or books publishers]  




The Web, today, is a firmly established (virtual) reality. A few years after its impressive break through, from limited professional circles to everybodyís working and social life, the Web constitutes an additional space for people to communicate, work, trade or spend leisure time. And increasingly, too, a place to learn. Educators, well aware of the potential of the technology have adopted it for creating new Web-based learning environments (WBLE). This paper presents a study of the characteristics of Websites as teaching and learning environments. The major questions addressed in this study were: What characterizes educational Websites at the content, teaching, learning and communication levels? How do key teaching and learning issues appearing on educational Websites relate to educators' expectations from the new technology? What can a consideration of the current state of affairs teach us about further development and implementation of educational Websites? To answer these questions we developed a classification scheme (the Taxonomy of WBLE), implemented it for the study of 436 educational Websites focusing in mathematics, science, and technology learning, and elaborated on practical implications of the studyís results. The overall picture we have unveiled may sound dissapointing, and can be summarized as ìone step ahead for the technology, two steps back for the pedagogyî. But a more thoughtful discussion of the results suggests directions for the research and development of novel Web-based educational models.





The Web, today, is a firmly established (virtual) reality. A few years after its impressive break through, from limited professional circles to everybodyís working and social life, the Web constitutes an additional space for people to communicate, work, trade or spend leisure time. And increasingly, too, a place to learn (Berenfeld, 1996; Khan, 1997). ThisÜpaper presents a study of the characteristics of Websites as teaching and learning environments.

Educators (teachers, developers, researchers, students), well aware of the potential of Web technology, have adopted it for creating new learning environments, thus yielding a huge repertoire of educational Websites. The rationale behind this creative endeavor is based on the expectation that unique features of the technology (e.g., powerful information manipulation tools, communication means) will substantially contribute to teaching/learning processes. Let us briefly review some salient technological features, which are relevant to educational processes.

The first and obvious key feature of the web is the support for sophisticated manipulation of information. Information manipulation functions (e.g., generating, transmitting, storing, processing, retrieving of information) are at the heart of educational transactions. The possibility to contribute to, or to access, on-line libraries, databases, journals, museums, and other public information repositories on the Internet may therefore qualitatively affect education.

The network serves more and more as communication facilitator. Computer-Mediated Communication (CMC) constitutes a powerful interaction means (e.g., e-mail, group conferencing, IRC's), which enables students to communicate with peers, teachers, and experts, and also to conduct collaborative work (Berge, 1995; Harasim et al. 1995).

The web is also increasingly becoming a creation environment . A considerable number of user-friendly tools for the creation of Web-deliverable materials are currently available. These tools may support students' creativity and initiative, allowing them to generate and publish their own Web units without mediators and with minimal technical assistance.

Finally, the Web also serves as instruction delivery medium . Numerous Websites provide digital educational activities and netcourses for all grade levels in a large number of subjects (Hackbarth, 1997; Khan, 1997). The conception of the Web as learning environment is instantiated in varied forms, from online versions of traditional CAI to innovative individual and group virtual-learning.modes.

Within the global network, already a vast quantity of Web-based learning environments (WBLEs) has been developed. These WBLEs reflect educatorsí attempts to wrap together knowledge in specific content areas and the above-mentioned technological features, in pursue of learning goals. WBLEs differ from each other in a variety of aspects. For instance, they differ in terms of the identity of their originators (e.g., teachers, students, development centers, research centers), their goals and target population. They reflect a diversity of pedagogical conceptions and beliefs of the developers, either explicitely stated or implicitly embedded in the siteís design. They also differ in regard to the configuration of technological features (e.g., communication tools, information resources). Obvious outcome of this state of affairs is the high variability and the uneven educational value and quality that characterizes the growing aggregate of educational sites.

Given the continuous increase in quantity and diversification in quality, and the high level of expectations in the educational community regarding the educational potential of the Web technology, we have reached a stage at which a mapping of the WBLE landscape is required. Regarding this need, the major questions addressed in this study were:

- What characterizes educational Websites at the content, teaching, learning and communication levels? (detailed mapping of relevant features)

- What are the key teaching and learning issues appearing on educational Websites in correspondence with the educators' expectations? (overall analysis and evaluation of trends and solutions)

- What can a consideration of the current state of affairs teach us about further development and implementation of educational Websites? (practical implications and conclusions)

To answer these questions we developed a classification scheme, the Taxonomy of WBLE (Nachmias, Mioduser, Oren & Lahav, in press), implemented it for the study of 436 educational Websites focusing on mathematics, science and technology learning, and elaborated on practical implications of theÜstudyís results. In the following sections we will briefly describe our classification scheme, we will present the study and its findings, and finally we will discuss these results and suggest directions for the further development of WBLE.


For this study we developed a practical tool for describing the complexity of the educational kaleidoscope that has been generated in the Web. Our tool continues a research line focusing on the systematic analyses and evaluations of technology-based learning materials, from the early years of computer-based instruction (e.g., Blease, 1986; Shuell, & Shueckler, 1989) to more recent work on Web-based instruction (e.g., Berenfeld, 1996; Khan, 1998; December, 1998), as well as on the accumulated experience and knowledge in the field of instructional design (e.g., Gagne, Briggs & Wagner, 1992; Dick & Cary, 1990). The taxonomy is a classification scheme aimed to reflect the developers' educational philosophies as well as their actual manifestations, by revealing how different functionalities are configured, the knowledge is structured and represented, and communication features are implemented. (for a more detailed description of related work and the background for the development of the taxonomy see Nachmias, Mioduser, Oren, & Lahav, in press).

Our taxonomy characterizes an educational Website by about 100 variables in four main dimensions: basic descriptive information, pedagogical and educational considerations, knowledge attributes, and communication features.


The descriptive dimension

This dimension includes basic information regarding the location, creators, target population and relevant technical data of a site. The information is organized in six categories: Site identification (e.g., name, URL, authors, affiliation -academic institution, public organization, government authority, private company, school, teacher, student); site evolution (e.g., creation date, last updating, sections under development); language or languages used in the site; target population and size , indicated by the number of html pages; and subject matter .


The pedagogical dimension

The variables in this dimension unveil the developers' stance regarding the type of instruction elicited by their site (e.g., target learning processes, instructional configuration and means, collaborative work, feedback, assessment). The variables in this dimension are organized in ten categories: instructional configuration , (e.g., Web-only resources or link to additional external resources); instructional model (e.g., directed and hierarchically organized, inquiry-oriented, open-ended); instructional means (e.g., hypermedia data-bases, virtual 3D environments, on-line student-modeling and adaptive mechanisms); interaction type (e.g., browsing, answering questions, performing simple or complex activities, using on-line tools, interacting with experts or peers); cognitive process elicited (e.g., plain information retrieval, complex processing of varied types of information, problem-solving, creative activity or invention); locus of control over the learning process; feedback (e.g., automatic evaluation answers, human expert's response either synchronic or asynchronic); help functions offered in the site; learning resources either embedded in the site's design or external physical and human resources; and evaluation (e.g., from standardized tests to alternative evaluation).


The knowledge dimension

This dimension relates to qualitative and structural issues concerning the site's knowledge and the knowledge-navigation support included in the site. The dimension comprises four categories of variables: representational structure , aiming to identify the organizational template underlying the knowledge stratum, (e.g., linear, branching, or Web structure); representational-means (e.g., text, still image, dynamic image, interactive image, sound, and the frequencies of their respective uses in a site); type of knowledge (e.g., declarative, procedural, dynamic/systemic models of phenomena or systems, continuously updated); navigation tools (e.g., thematic indexes, image maps, time-lines, iconic directional-pointers, search facilities, location maps).


The communication dimension

Networking by definition implies communication, or people's interaction with knowledge and/or with other people. The fourth dimension of the taxonomy relates to communication features in four categories: types of telelearning , focusing on the different shapes which distance learning via technology may take (e.g., tele-information-handling, tele-interaction, tele-manipulation, tele-creation); types of communication (e.g., synchronous, asynchronous; links structure of the site, e.g., hypertext links within the site, links to other sites, to other sites' databases, to non-Web tools and activities, to virtual reality environments or to humans); communication means (e.g., electronic mail, discussion group with or without moderators, chat facilities, video conference capabilities, or moo/mud features).


Sample selection

Since the defining properties of educational Websites, such as their educational aims and features, structural attributes or boundaries and size, are still loosely defined and open to many interpretations, a special process of selection of a random sample of Websites was takenÆ Five evaluators were selected as research assistants for this project. All five were students in the graduate program of Communication and Computers in Education at Tel-Aviv Universityís School of Education. All students had a scientific background: one had a B.Sc. in Physics, two in Computer Science, one in Biology and one in Mathematics. All were at the time science educators, and had studied graduate courses on educational usage of computers and communication systems.

Each of the evaluators was assigned to find about 100 educational Websites. They were instructed to select Websites that met the following criteria: (a) the site has been deliberately developed for educational purposes; and (b) it must be clearly focused and identifiable as a specific instructional unit (e.g., by its focus on a specific topic, or on a specific learning task). The first criterion means that, although any site in the Web can be used as resource for learning, only sites explicitely defined by their developers as pursuing educational goals were selected for this study. The second criterion was defined to avoid the selection of ìmega-sitesî, i.e., Websites that are in fact "umbrella-sites", or general-access-sites to conglomerates of educational projects or web pages. InÜaddition, the content area for the selection of the sites for the present study was circumscribed to mathematics, science and technology education. The language of the Website, or the age level of the target population, were not criteria for the selection

Each of the evaluators was instructed to look for about 100 Websites either by browsing or using search engines. All evaluatorsí sets were integrated into a common list, which was screened for duplicates and approved by the senior research team. Of the 524 sites that were initially selected, 436 were included in our final sample. Sample selection took place in March 1998 (see list of Websites in


Characterization of the selected Websites

At this stage each of the five evaluators received about 90 Websites, randomly selected from our list, to be characterized according to the WBLE Taxonomy previously described (Nachmias, Mioduser, Oren & Lahav, in press). To maximize the common framework of analysis, several meetings took place with all researchers during which the taxonomy was discussed and experimentally applied on a number of Websites. The evaluation process was carried out during April and May 1998. By June 1998, the database comprising the characteristics of 436 Websites was completed.


Data-base validation

To assess the validity of the database, a sample of about 25% of the Websites was analyzed once again by a different evaluator. As a result five variables (out of the 110 variables of the taxonomy) were excluded from subsequent data analysis. All other variables met the criterion of an at least 90% match between evaluators, and were therefore regarded as valid. This stage was completed by the end of June 1998.


The descriptive dimension

Table 1 presents basic descriptive information about the Websites in our study. Academic institutions and museums are the main contributors of educational Websites in our sample (about one third of the sites each). Public organizations, private companies, and varied educational agents are among the generators of the remaining third. It should be noted that these figures do not necessarily represent the actual distribution of educational sites by their originators, but the biased distribution resulting from our sampling procedure.

Table 1: Basic description of Websites (N=436 sites)



N of sites (%)

Site Creators Academic institution 152 (34.9%)
Public organization 40 (9.2%)
Private company 73 (16.7%)
Museum 145 (33.3%)
Education factors 26 (6.0%)
Language(s) English 388 (89.0%)
Hebrew 53 (12.2%)
Arabic 2 (0.5%)
Target Population Kindergarten 4 (0.9%)
Elementary school 92 (21.2%)
High school 270 (61.9%)
College / University 61 (14.0%)
Further education 9 (2.1%)
Site size (html Pages) 1-3 135 (31.0%)
4-10 66 (15.1%)
11-30 91 (20.9%)
31-70 64 (14.7%)
+70 80 (18.3%)


About 90% of sites in our sample were in the English language, i.e. the dominant language of the Web, with about 12% in Hebrew, and some in Arabic representing Israelís languages.

Sixty two percent of the sites were aimed at the upper elementary level, 22% at the elementary level, and the remaining 16% were developed for higher education.

Data about number of Html pages per site show that about a third of the Websites were short (1-3 pages), a third were median-sized (4-30 pages) and the remaining third were large (over 30 pages) sites. Given that one of the selection criteria of Websites in this study was the siteís topical focus, the size of the Website might be taken as an indicator of the extension of the treatment of the specific topic.

The evolution of the Websites is described in Table 2. Only for about 60% of the sites date of creation was available. A clear trend can be observed in pace of growth from 1993 to date. From year to year the number of sites increased by a factor of 2 to 3. Thus, most sites in ourÜsample were created within the last two years. Data about last update show that about half of the Websites were updated by their authors within six months prior to our review, and about 42% had not been updated for over one year. Only a few sites (8.5%)Üwere still under development as indicated in the sites (ìunder constructionî), while the majority was completed.

Table 2: Sites evolution in term of creation, updating and completion time

N of sites (%)
Year of creation 1993 3 (0.7%)
1994 11 (2.5%)
1995 22 (5.0%)
1996 62 (14.2%)
1997 151 (34.6%)
1998 8 (1.8%)
Not available 179 (41.1%)
Last update Over a year 183 (42.0%)
8 months 27 (6.2%)
6 months 39 (8.9%)
4 months 82 (18.8%)
2 months 105 (24.1%)
Completion Completed 399 (91.5%)
Under construction 37 (8.5%)


Table 3 shows the distribution of sites by subject matter issues. Given that our sample refers only to mathematics, science and technology sites, the table shows that all major science disciplines are represented, and that most Websites deal with biology physics or mathematics topics. Three quarters of the sites deal with one subject matter only; about one quarter of the sites were multi-disciplinary.

Table 3: Distribution of Websites by Subject Matter

(N of sites %)
Biology 110 (25.2%)
Physics 146 (33.5%)
Mathematics 98 (24.6%)
Chemistry ÜÜ 28 (6.4%)
Technology 54 (12.4%)
Earth Science ÜÜ9.2 40 (9.2%)
Computer Science ÜÜ 10 (2.3%)
Astronomy ÜÜ 30 (6.9%)
Other 66 (15.1%)
Multi Disciplinary
One discipline 331 (75.9%)
Two disciplines 63 (14.4%)
Three disciplines 19 (4.4%)
More than three 15 (3.5%)


The pedagogical dimension

Table 4 focuses on the pedagogical features of the Websites. More than 93% of the sites are aimed to individual work. Less than 3% support online collaborative work and 12% include learning activities which suggest classroom collaborative work as supplement to the online work.

Analysis of the instructional model embedded in the sites shows that a traditional, hierarchical, highly structured, and directed instruction mode still prevails. Only 28% of the sites support inquiry-based learning.

Web technology offers a wide range of possibilities regarding instructional means. Data shows that the most frequent means implemented are information-bases (65%) and structured activities (48%). Open-ended activities, tools, and virtual environments are included in about 7-13% of the sites. Only very few sites include online adaptive mechanisms.

Interactivity could be considered one of the major potential contributions of digital technology to instruction. The data shows that the lowest level of interaction according to our scale, namely browsing, is also the most frequent (76%). In about a third of the sites question-answer tasks were included; simple interactions, in which clicking or dragging objects on the screen activates a predetermined script, appear in about 42% of the sites; more complex interactions (e.g., manipulation of a number of variables), or use of online tools, are included in 3-7% Websites only; interaction with other people (e.g., expert, peers), mainly asynchronic, was found only in 13% of the Websites. Feedback features (either automatic or human) are included only in few sites, far below their presence in pre-Web digital learning materials (e.g., Azevedo & Bernard, 1995; Cohen, 1985; Cyboran, 1995). Help features of the sites were found at three levels: technical help features (e.g., installation of required fonts or plug-ins) were found in 21% of the sites; contextualized content-help (e.g., glossary, translation) in 25%; and didactic help (e.g., explanations, examples) in 17% of the Websites.

The most frequent cognitive processes elicited by the activities were information retrieval (52%) and memorizing (42%). Information analysis and inferencing were supported by activities in about one third of the sites. Only few sites support higher level processes such as problem solving or creation and invention.

Studentsí control over their work is supported in 86% of the sites. However, this figure should be considered in relation to the finding indicating that the most frequent interaction mode in the sites is browsing, meaning that studentís locus of control consists primarily in wandering through the web pages.

Educational Websites can be considered as a bundle of varied representational and pedagogical resources. In this category of the taxonomy we looked for the different types of resources in the sites, and whether these were constrained to Web resources or they were complemented with classroom resources. We found that 83% of the sites rely on within-the-site resources; only 31% provide links to other Web resources; 22% refer the learners to additional external resources, i.e. that the Website is only one component of the learning environment. Very few sites refer to experts and peers, either on or off-line, as learning resources.

Evaluation means, either standardized or alternative, are rare in educational Websites.


Table 4: Number of sites including instructional and learning variables


N of sites (%)


Instructional configuration

Individualized instruction

407 (93.3%)

Classroom collaborative learning

54 (12.4%)

Web collaborative learning

12 (2.8%)

Instructional model Directed

330 (75.7%)


123 (28.2%)





283 (64.9%)


56 (12.8%)

Structured activity

211 (48.4%)

Open-ended activity

43 (9.9%)

Virtual environment

30 (6.9%)

Student modeling/adaptive mechanism

9 (2.1%)





333 (76.4%)

Multiple choice question

137 (31.4%)

Simple activity

185 (42.4%)

Complex activity

13 (3.0%)

On-line tool

28 (6.4%)

Expert consultation

58 (13.3%)




Information retrieval

229 (52.5%)


183 (42.0%)

Information analysis and inferencing

142 (32.6%)

Problem solving and decision making

22 (5.0%)

Creation and invention

20 (4.6%)


Locus of


Student controlled

377 (86.5%)

Software environment controlled

77 (17.7%)

Mixed initiative

26 (6.0%)




71 (16.3%)

Human asynchronous

17 (3.9%)

Human synchronous

7 (1.6%)




Technical help

91 (20.9%)

Contextualized content-help

152 (34.9%)

Didactic help

73 (16.7%)




Within Website resources

363 (83.3%)

Linked WWW resources

135 (31.0%)

Additional external resources

93 (21.3%)

External resources only

4 (0.9%)

Real time data collection

6 (1.4%)

Ask an expert

38 (8.7%)

Ask a peer

17 (3.9%)

Evaluation Standardized tests

29 (6.7%)

Alternative evaluation

7 (1.6%)


The knowledge dimension

Websites are above all repositories of knowledge. In our analysis of the knowledge dimension of educational sites we looked at features such as representational structure, representational means or navigation tools. The Representational structure of the knowledge in the Websites can be of various types: isolated unit (23% of one-page sites), linear sequence of Web-pages (16%), branching structure (31%), or network structure (30%). These figures show that in only less than a third of the sites knowledge is represented in the network-like structure, claimed to be the quintessential representational template of the Web.

The great majority of the sites presented declarative knowledge (92%). Procedural knowledge appeared in 20% of the sites, and qualitative knowledge only in 7%. Only a few sites (4%) offered support for the knowledge to be expanded by the learners.

Table 5 shows the frequency distribution of representational means in the Websites. As could be expected, text is the dominant information conveyor in the Web. Visual means (e.g., images, photos, and illustrations) are less frequent, but were still included in about 60% of the Websites at least once per page. About 15% of the sites do not include any visual information. Interactive images, sound effects and real-time data update were rarely found (1 to 4% of the sites). Animation, mainly adornments in the form of visual-images-loops, appeared at least once in almost 20% of the sites.

Frequency of appearance of navigation tools is presented in Table 6. The most frequent support means for navigation within a Website is indexing (about 70% in the homepage as links to other pages in the site; 20% as within-the-page links; and 30% as visual indexes or content-bars), resembling the traditional orientation means of the print technology. Surprisingly, sophisticated navigation tools more suitable to the Web environment (e.g., image maps, search engines) appeared only in relatively a small number of sites.


Table 5: Distribution of sites by representational means (N of sites and %)

Not at


once in the site

50% of pages in the site One per page More than one
Text 2 (0.5%) 0 (0%) 3 (0.7%) 24 (5.5%) 407(93.3%)
Image 63 (14.5%) 64(14.7%) 55(12.6%) 117(26.8%) 137(31.4%)
Interactive image 419(96.1%) 10 (2.3%) 1 (0.2%) 1 (0.2%) 5 (1.2%)
Animation 357(81.9%) 35 (8.0%) 18 (4.1%) 18 (4.1%) 8 (1.9%)
Sound 426(97.7%) 6 (1.4%) 1 (0.2%) 2 (0.5%) 1 (0.2%)
Real-time updating 431(98.9%) 0 (0%) 3 (0.7%) 1 (0.2%) 1 (0.2ï)


Table 6: Presence of various navigation tools in the Websites


N of sites (%)

Index in home page

307 (70.4%)

Local-page indexing

82 (18.8%)

Content bar

131 (30.0%)

Tool bar

9 (2.1%)


5 (1.1%)

Alphabetical bar

15 (3.4%)

Image map

28 (6.4%)

Permanent frame-index

34 (7.8%)

Internal search engine

38 (8.7%)

Knowledge map

29 (6.7%)


The Communication Dimension

Unique features of Web technology which may contribute the most to educational processes are these related to communication with distant knowledge and people (e.g., peers, experts). Table 7 shows the frequency of inclusion of varied communication means. By far the most frequent tool is electronic mail (65% of the sites). Only part of the means supporting group interactions were included, and these were found in a few sites (e.g., discussion groups, chat -about 2 to 4%), while other are not supported at all in our sampleís sites (moo/mud environments, video-conferencing). Distance work (e.g., tele-operation, tele-creation) is supported in less than 2% of the sites.

Due to the popularity of chat environments and online multi-user games among school age Web users, we expected to find a great deal of support for synchronic activities in our sample of educational Websites. However, only about 4% of the sites provided any kind of synchronic activities.

Another defining feature of the Web is the network of links enabling the user to navigate both the site's pages as well as other sites (Table 8). About two thirds of the sites included linkage among pages in at least half of the pages. Only about a quarter of the Websites included links to other Websites in at least half of their pages, while 58% of the sites did not have any links to other Websites at all.

An analysis of the kind of resources towards which the users are directed, which are external to a site, shows that about 17% refer to external data-bases, about 10% to external activities pages, and a few more to various other resources (e.g., virtual reality environments, human resources).


Table 7: Presence of various communication means in the Websites

  N of sites (%)
Communication type  
Synchronous activities 17 (3.9%)
Communication means
e-mail 283 (64.9%)
Discussion group without mediator 15 (3.4%)
Discussion group with mediator 10 (2.3%)
Chat 8 (1.8%)
Moo/mud 0 (0%)
Video conference 0 (0%)
Tele-manipulation 1 (0.2%)
Tele-creation 7 (1.6%)



Table 8: Configuration of links in the Websites

Not at



in the site

50% of

pages in

the site

One per





Within the site 116(26.6%) 32(7.3%) 20(4.6%) 50(11.5%) 218(50.0%)
Links to external sites 253(58.0%) 68(15.6%) 37(8.5%) 32(7.3%) 4(10.6%)
LINKS TO: N of sites (%)
external databases 74 (17.0%)
external tools 12 (2.8%)
external activities 42 (9.6%)
virtual reality devices 8 (1.8%)
human communication 25 (5.7%)

At an additional level of analysis we looked for the differential presence of selected properties (or variables) in sites aimed at different age levels and created by different developers.

Table 9 shows the results of the cross-tabulation of selected variables by target population. The population levels considered were elementary (K-6; 96 sites in our data set); upper-elementary, comprising junior-high and high-school (270 sites), and higher education (70 sites). This indicates that for the sampled sites most development efforts in the k-12 range are directed towards the upper-elementary grade-levels.

Of the small percentage of Websites fostering any form of collaborative learning, most are targeted to the high school (the few found instances of Web-based collaboration were only for this level). Similarly, the largest number of inquiry-focused sites aim at the same age level. And the largest number of sites including evaluation means, mainly in the form of standardized tests, belonged to the higher education level. Conversely, a very low (or even null) level of collaborative work, discussion groups, or human-based feedback was found at this level of Websites.

Of similar interest are the results related to variables for which no difference was found among sites aimed at different population groups. No difference was found regarding most instructional modes (e.g., open-ended, problem-solving or creation activities); instructional means (e.g., online tools, technical, contextualized or didactic help, expert consultation); navigation aids (e.g., index, content bar, image maps); or configuration of the information (e.g., links within the site or to external resources). These findings are particularly interesting considering our expectation that different needs and requirements at each age-level would demand different instructional solutions.


Table 9: Cross-tabulation of selected variables by target population


K - 6










Classroom collaborative learning 10.4% 15.9% 1.4% 12.4% 11.20**
Web collaborative learning 0% 4.4% 0% 2.8% 7.58*
Inquiry-based 0.8% 34.4% 14.3% 28.2% 14.46**
Open-ended activity 6.25% 11.1% 10% 9.9% 1.88
Complex activity 2.0% 4.0% 0% 3.0% 3.53
On-line tool 3.1% 6.2% 11.4% 6.4% 4.66
Expert consultation 8.3% 14.0% 17.1% 13.3% 3.08
Problem solving and decision making 1.0% 5.5% 8.5% 5.0% 5.17
Creation and invention 3.1% 4.4% 7.1% 4.6% 1.52
Automatic Feedback 9.3% 16.6% 24.2% 16.3% 6.67*
Human asynchronous Feedback 0% 5.9% 1.43% 3.9% 7.99*
Human synchronous Feedback 0% 2.2% 1.43% 1.6% 2.23
Technical help 16.6% 20% 30% 20.9% 4.68
Contextualized content-help 34.3% 35.5% 31.4% 34.9% 0.50
Didactic help 15.6% 18.5% 11.4% 16.7% 2.11
Evaluation - Standardized tests 3.1% 6.3% 12.8% 6.7% 6.31*
Evaluation - Alternative evaluation 1% 1.8% 1.4% 1.6% 0.31
Index in home page 66.6% 69.2% 80% 70.4% 3.91
Content bar 30.2% 29.6% 31.4% 30% 0.09
Image map 5.2% 7.4% 4.3% 6.4% 1.20
Synchronous activities 2% 3.3% 8.6% 3.9% 5.15
e-mail 61.4% 65.5% 67.1% 64.9% 0.70
Discussion group without mediator 1% 5.2% 0% 3.4% 6.63*
Discussion group with mediator 0% 3.7% 0% 2.3% 6.29*
Links within the site 63.5% 65.2% 72.9% 66.1% 1.80
Links to external sites 25% 24.8% 34.3% 26.4% 2.68

(*) p<.05 (**) p<.01


Table 10 shows the crosstabulation of the same set of variables by Website source or developers. The vast majority of sites in our sample were generated by academic institutions and museums. Different types of institutions act according to different goals and beliefs regarding their educational roles, and it could be expected that these differences would find their way into the design of the Websites. The picture resulting from our data is not clearly consistent. In 11 out of 26 variables significant difference among the developers was found. For ten of theseÜvariables the largest number of sites holding the given feature was generated by academic institutions (automatic feedback, contextual help, index-type navigation aid, moderated discussion groups), and by public organizations (Web-based collaborative learning, open-ended activities, human asynchronous feedback, standardized tests, discussion-groups without moderator, and wide linkage to external sites).

The features mostly included in sites generated by private companies were different types of help (e.g., expert consultation, technical and contextual help), proven navigation aids (e.g., content page, content bar), and a fairly rich linkage structure. The features mostly included in museum -generated sites seem to be in line with the expected characteristics of educational activities in museum environments (e.g., collaborative, inquiry-based, and problem-solving activities, contextual help, within-the-site links).


Table 10: Cross-tabulation of selected variables by site developer


Acade-mic org.


Public org.


Private com.











Class collab. learning







Web collab. learning














Open-ended activity







Complex activity







On-line tool







Expert consultation







Prob/solv and decis/mak







Creation and invention







Automatic Feedback







Human asynch. feedback







Human synch. feedback







Technical help







Context content-help







Didactic help







Standardized tests







Alternative evaluation







Index in home page







Content bar







Image map







Synchronous activities














Disc. Group - no mediator







Discussion with mediator







Links within the site







Links to external sites









(*) p<.05 (**) p<.01



Web technologyís transition from its early rudimentary stages to the current ìeveryone-can-do-itî stage, generated high expectations among educators. These expectations relate to the Webís potential impact on educational processes in three main domains, by fostering (a) new pedagogical forms emerging out of unique features of the technology (a ìWebagogyî?); (b) improved information organization, representation and handling capabilities; and (c) enhanced communication processes among students and teachers and support for collaborative learning. Our aim in this study was to assess the extent to which educational Websites, i.e. sites deliberately developed for educational purposes, realize the potential and fulfill the expectations. Accordingly, in the following sections of the discussion we will elaborate on the study results in these three domains.

Pedagogical characteristics of WBLE's

Current pedagogical approaches support learning processes that require the studentsí active involvement in the construction of knowledge (Kafai & Resnick, 1996), their interaction with peers and experts (e.g., collaborative learning, distributed cognition, Perkins, 1993; scaffolding óÜVigotzky, 1978), the adaptation of instruction to individual needs (Reuser, 1996), and new ways to assess studentsí knowledge and learning (Mioduser, Venezky, & Gong, 1998). Our expectation was that the development process of educational Websites would beÜbased on these approaches. Moreover, given the innovative character of the technology, it could be expected that it would give raise to new pedagogical forms and instructional strategies.

Notwithstanding, the results indicate that this is not the case. Only 28.2% of the sites include inquiry-based activities, and more than three-quarters were highly structured, placing control over the learning process mainly with the computer. Most sites elicit cognitive processes such as information retrieval (52.5%) or rote learning (42%), fewer focus on analysis and inference processes (32.6%) and even less on problem-solving and decision-making (5%). Only a small number of sites include student-modeling and adaptation mechanisms. Considering the fact that network technologies appear to be an ideal milieu for the implementation of collaborative work, it is highly disappointing to find that only 2.8% of the sites support any form of collaborative learning. These results conclusively show that the pedagogical approaches favored by educators and researchers for the development of valuable learning environments are still far from being implemented in most educational Websites.

Another perspective on our findings refers to the configuration of instructional modes and means offered by educational Websites. Once again expectations exceed the actual state of affairs. The gap is most evident if we consider the accomplishments reached within the field of digital-technology-in-education in its previous form, namely (non-networked) instructional software. Regarding interaction types, we found that most sites include browsing (76.4%) or simple forms of interaction (42.4%), and few sites offer complex (3%) or even on-line (6.4%) activities. Few sites include any form of feedback, either automatic (16.3%) or human (5.5%). Most sites offer resources and means related to information handling (65%). Only few offer the student online tools (12.8%) or resources external to the site itself such as resources in other sites (31%) or experts (8.7%).

Pre-Web (digital) educational materials offer fascinating examples of the multiple ways by which educators succeeded in harnessing the new technologies to educational needs and goals (e.g., constructivist environments, intelligent tutoring systems, sophisticated multimedia learning environments). Against this rich background, and looking at pedagogical qualities and resources, the vast majority of educational Websites prove to be the unripe fruits of the promising but still immature Web technology.

Information representation and handling

Web technology is doubtlessly at its best concerning all functions related to the representation, organization and manipulation of information. The process initiated several decades ago with the massive irruption of radio, TV and cinema technologies is reaching impressive achievements with current digital imaging and integrated-media technologies. In cultural terms, the visual world (e.g., still images, icons, video-clips, animated graphics. movies) in its linguistic status has reentered the scene, in stronger presence and meaning than in its pre-Gutenberg incarnation. High-level and sophisticated integrated media is perhaps one of the defining characteristics of Websites today.

But again our results show that regarding their information representation and handling qualities, educational Websites lag behind state-of-the-art sites (it should be noted that we are not interested in educatorsí technical óand even pyrotechnical- use of imaging technology, but only in their use of visual languages). The vast majority of sites are still heavily based on text (93% of the sites include more than one text field in all their pages). About 58% of the sites include at least one image per page; most sites do not include interactive images (96.1%), animated images (81.9%), or sound.

Regarding structure and organization of knowledge, the Web is the realization of a well defined model: the hypertext (or hypermedia) model (Negroponte, 1995). Non-linear structure, complex linkage within and between information units, and appropriate navigation and search tools are defining features of this model. Our results reveal only a shallow presence of these features in the evaluated Websites. Only about half of the sites included withinthe-site linkage to a reasonable extent (more than one link per page), and about 11% of the sites refer to other sites (external linkage), to a similar extent.

A great deal of theoretical and empirical work has been done regarding curricular issues within the context of the print technology (e.g., West, Farmer, & Wolff, 1991). Do these curricular models fit the features, possibilities and constraints of the new technology? (For example, should curricular principles related to scope and sequence of content, and activities developed for linear technology be adapted to the hyperlinked nature of Web technology?) The intuitive answer is that curricular theory should be revised and perhaps expanded in light of these new features. Perhaps a convincing illustration can be found in the results regarding navigation tools in the evaluated Websites. Most sites include indexes (70%), several more contain content-bars (30%), and only a few contain more sophisticated tools such as image-maps, knowledge-maps, or internal search-engines. Both structure and orientation aids in most sites are reminiscent of curricular solutions which were devised for the previous (print) technology.


The Internet is not only an information environment. Perhaps its foremost function is to serve as a multifaceted environment for communication (Mitchell, 1995). A considerable number of communication modes and tools are offered to the users (e.g., e-mail, data-transfer, discussion groups, synchronous video and audio support for collaborative work). This technological infrastructure has great potential for the development of unique learning transactions and modes, e.g., collaborative learning among students in distant locations, or implementation of the apprenticeship model by involving students and experts in network-supported group work.

The results of this study show that most Websites used only limited communication resources. The most (and almost sole) present resource in the sites is electronic mail (about 65% of the sites). Other tools such as discussion groups, chat, or any form of distant work (e.g., tele-manipulation, tele-creation) were found only in a few sites. Moreover, features aimed to support working groups or learning communities were not found in any of the evaluated sites. The gap between expectations and actual implementation is even more evident in the communications domain than in the previously discussed domains: the technological resources do exist and are being successfully implemented in other areas of peopleís life (e.g., work, professional training, banking, shopping). In addition, human transactions and transactions between humans and information resources are quintessential to education, and it is not hard to conceive endless forms of support that communication technology could offer for these processes. Today this support is not yet a function in most educational Websites.



One step ahead for the technology, two steps back for the pedagogy. One can depict that way the usual loops affecting educatorsí assimilation of new technologies for the last decades. This phenomenon is not unique to education. The effects of the interplay between technological innovations and humans has been a matter of study both at the individual (e.g., cognitive processes, attitudes) and the social (e.g., cultural transformations) levels (e.g., Ihde,1990; McLuhan, 1964; Negroponte, 1995; Olson, 1976; Salomon, Perkins, & Globerson, 1991). In the early 20ís W. Ogburn (1964) coined the term ìcultural lagî to referÜto the differential reaction of distinct groups or parts of a culture to innovations and changes. Facing the advent of a technological innovation, different populations react at different paces and adopt different attitudes. The assimilation process of the innovation then proceeds through various transition stages. Succesful assimilation will depend on the individualís increase of awareness to the relevance of the esential features of the new technology to their life (e.g., work, learning, leisure time)® as well as on the development of appropriate social constructs (e.g., at the economical or cultural levels). History of technology textbooks frequently quote the examples of the first cars conceived as carriages-without-horses, or the first movies as filmed theater plays. Practitioners had to go through a complex maturation process until new languages and unique qualities were developed in the car making or film making fields.

As experienced educators we hold substantial models regarding the varied facets of our practice (e.g., how to build a lesson plan, to assess a learnerís performance or behavior, to develop a learning unit). These models are usually tied to the (technological) resources at hand, and they affect each other mutually. It seems reasonable to assume that when facing the assimilation of a new technology we use these models as input to the process. The result is usually a transition period during which the known models are replicated by means of the new technology. When first assimilating the computer technology, developers replicated the programmed instruction paradigm (Garner, 1966) by means of the new technology, initially in the form of electronic worksheets and booklets, which evolved in time into sophisticated drill and practice and structured tutoring software (Venezky & Osin, 1991). Our claim is that this study reveals a similar transitional phenomenon regarding the vast majority of educational Websites. Most sitesí main component is the information-base, built upon the hypermedia-CD model (even the linkage to external sites is a feature currently included in many hybrid CD-Web products). As for interactivity features based on the implementation of new technological resources (e.g., forms, Java applets, Shockwave), most online activities resemble the automatic-feedback (behaviorist-like) transactions of classic CAI (e.g., multiple-choice, select-correct-part, assemble-correct-configuration).

In light of these results, one can adopt the skepticsí perspective and argue that Web technology has little to offer to education. But, adopting a more thoughtful perspective, one may alternatively reflect on the character of this transition stage and generate new possible models and trends based on substantial educational needs.

The authors are members of the community of educators who deal with the problematic involved in the assimilation of Web technology to education (Mioduser, & Oren, 1998; Nachmias, Mioduser, Oren & Lahav, in press). Although in this study we have referred to the educational Websites population in quantitative terms, we are aware that a number of fascinating high pedagogical-quality sites do exist in the Web. But our main purpose in this study, rather than to focus on the exceptions, was to map and learn what existing sites and current trends, as they are launched into cyberspace, have to offer to educators and learners. Based on this and similar studies, the next steps should focus on the research and development of novel Web-based educational models (Windschite, 1998©, and on the implementation a revised configuration of the technology-assimilation evolutionary loop: two steps ahead for the pedagogy/technology, one step back for reflection and mindful planning of subsequent steps.