Modern brain research verifies that learning involves modifying neurons. In particular, neurons are interconnected by biological entities called synapses. Stimuli (perceptions) create electrical signals that move through neural tissue such as brain tissue by bringing about temporal changes in synapses. If, during some rather brief interval, many synapses that touch a neuron fire and release neurotransmitters, the cumulative effect is to cause a drastic change in the affected neuron. It fires. The result is that synapses emanating from the recently fired neuron pour chemical neurotransmitters into their synaptic clefts and, should sufficient numbers behave similarly and more-or-less synchronously, the neuron upon which they impinge will fire.

The massive biological circuitry involves many complex and often poorly understood features. In the end, however, neurons connected to muscles fire causing muscles to move in some fashion. This may create an observable effect which, when evaluated in the context of many muscles moving over a brief span of time, is called a behavior. Learning involves neuron modification. It involves changing the strengths of the synaptic connections between neurons. Over time, as the result of learning, different muscles react to neural inputs, and we say that different behaviors have resulted – that behavior has changed.

Books concerning cognitive psychology nearly always have at least one chapter devoted to neurons [for example, see Pressley and McCormick, 1995]. Pacesetters even have dared to suggest biological bases for how humans think [Crick, 1994]. The links between this theoretical base and classroom practice are very weak, however. Much current learning is described using a jargon developed under the name, constructivism. A key notion of constructivism is that learners construct their own knowledge. We interpret this to mean that the process of neuron modification is such that a teacher cannot transfer, whole and intact, either copies of his/her neural connections or some other idealized set of such connections to learners. Indeed, neither near copies nor even reasonable facsimiles appear in learners.

Understanding neural models for learning is a complex business. It is far beyond the purpose of this book. It is a far too common practice in studying learning to reduce complex systems and situations to simple cases described by trivial metaphors. At the risk of being guilty of such poor practice ourselves, we call your attention to some experiments conducted about learning in environments pervaded by controlled odors (Schab, 1990). In one study, a large student group was divided. One half of the group received instruction in the presence of the odor of chocolate, and the other half with the odor of mothballs. For testing of material learned, each group was further halved. Half of the chocolate treatment group was tested in the presence of chocolate odor, and the other half in the presence of mothball odor. The mothball group was treated in a parallel fashion. The result? Those taught in the presence of chocolate odor tested better in chocolate; those taught in the presence of mothball odor tested better in mothball odor. If you are a person that believes learning is disembedded from its contexts, this experiment will give you nightmares. If you believe that contexts affect learning, then these results will validate your beliefs.

Although most learning theories fail to explain this outcome, the neuro-logical model has little difficulty with it. The odors are perceptually detected, integrated into a holistic neurological mass of information, and increase the perceptual similarities in the presence of stimuli (test questions) so as to enhance performance. Whether pleasant (chocolate) or unpleasant (mothballs), the presence of the same odor during testing stimulates the same neural paths present during learning and thereby enhances the likelihood of a correct response.

There is a great deal of evidence supporting a neurological basis for learning. Near weekly revelations appear regarding details of neural mechanisms. However, these are of little use to the teacher. Successful teaching still relies upon information far removed from the neurological phenomena underpinning learning at molecular and cellular levels. Rethinking Innateness [Elman et al., 1996] gives an extremely powerful account of the neurological basis for learning.

In educational research, numerous situations arise where face validity is all that supports the choice of a strategy or methodology.

Educators often search for the holy grail of teaching, a goal well summarized in a paper by Bloom (1984). Numerous research studies suggest that some teaching strategies lead to substantially more positive learning outcomes than others.

Demanding that students learn material to a minimum standard – often called mastery learning – is extremely effective. Mastery strategies, discussed further in Chapter 15, involve frequent testing, with content re-tested until standards are met.

Cooperative learning, where learning tasks are subdivided within a small group of learners responsible for teaching one another different aspects of a learning problem, is moderately effective. Strategies for supporting cooperative learning are mentioned further in Chapter 5.

The Web is a place where one can read and read, and then read some more – without being forced to respond in ways that demonstrate learning. Much of what is on the Web is ideal for learners who are successful as passive learners. A highly motivated learner, for example, may thrive on vast amounts of Web information.

Few of us have students with such deep motivation. Curriculum materials that force students to respond, to make choices, to perform, to organize, and to think deeply about the material have better outcomes, generally, than ones in which they just read or listen. The former behaviors often are labeled under the heading of active learning, and much research indicates that active learning is more effective than passive learning. With curriculum materials that encourage active learning, the teacher is mainly a silent facilitator.

Research usually does not support the teacher-centered model as a preferred model, especially when pitted against active learning models. As a psychometric construct, however, active learning leaves much to be desired. The term active learning echoes in the halls of curriculum and instruction departments, but usually remains absent from works on teaching written by members of educational psychology departments. If one looks at materials written by learning experts, active learning usually doesn't seem to appear in either the tables of contents or the indexes.

We advocate active learning strategies because, when well developed, their effects can be substantial. Several chapters of Web-Teaching have been devoted to active learning strategies. Just clicking around the Web (surfing) is not an especially effective learning strategy. Purposeless surfing of the Web is not likely to bring about learning gains. Indeed, purposeful surfing may be very limited in its impact.

Significant confusion arises when teachers use the word technology in the context of teaching. In today's world, technology generally is interpreted to mean computer when applied to teaching. Technology may be broken down into two areas: the professional tools which incorporate small (or not so small) computers, and the communication systems that make use of computers.

Several studies have been reported in which graphing calculators (small computers) or powerful software (Mathematica, Maple, molecular structure software) were incorporated into teaching. College-level students using graphing calculators came to understand the concept of function substantially better than did students in traditionally taught control classes [Pressley and McCormick, 1995, p. 434]. When the software is merely demonstrated in lecture classes, learning gains are small or nonexistent [Klein, 1993]. In chemistry, students sometimes lost ground in classes with technology-based software presentations but little or no learner practice or feedback [Cassanova and Cassanova,1991; Cassanova, 1996]. When the technology tools are put in the students' hands, and the instruction modified to include activities that illustrate the power of the software, learning gains are often substantial [Hembree & Dessart, 1986; Cooley, 1995; Park, 1993; Porzio, 1994]. Again, when teaching newer technology tools (calculators, software), strategies for active learning consistently seem to give better results.

Teaching with technology tools requires some accommodations, however, and faculty still struggle to achieve a balance between the demands of the disciplines and the details of the technology interfaces (Runge et al., 1999).

The Web is a communications tool that all professionals use in similar ways. The most powerful learning experiences are those that engage students deeply in meaningful ways. They force active learning, and they provide realistic environments that have a way of nurturing motivation. Real-world activities are limited and still relatively unusual as far as being included within routine school-based instruction. Teachers struggle to create effective active learning environments in classrooms and laboratories. The challenge is to incorporate the principles of active learning into Web-based learning. Because the instructional delivery system is at the same time the worldwide scientific communication system, exciting new opportunities for involving students are emerging. A consortium of physical chemistry teachers reports good results with a small number of very challenging activities taught using the Web (Sauder et al., 2000). Poë describes the use of problem based learning in large lecture classes (Chapter 5).

Multimedia instruction passes the most important test that instructional materials must pass before they are used in classrooms, namely, face validity in the eyes of the teacher. When committed teachers see an excellent multimedia piece developed to teach some aspect of their discipline, usually their eyes widen, and broad smiles spread across their faces. They are impressed. The potential for creating media to portray phenomena visually and aurally is very great. We are no longer constrained by our students' ability to conjure up images in their minds to fit our verbal descriptions. Indeed, we may even introduce biases far and wide about the nature of phenomena through the teaching images we can create today.

There is every reason to believe, however, that multimedia materials in and of themselves do not lead to enhanced learning. Indeed, they often lead to slightly lowered learning when compared with comparable text materials. When adjusted to account for differences in active learning components, research studies aimed at studying media often show no significant learning differences [Clark, 1983; Clark and Salomon, 1986; Heinich et al., 1996]. The active learning is the key, not the multimedia.

While transforming content to a multimedia format may be a "cool" and popular thing to do, it by no means ensures learning gains. Salomon (1986) reported that perceptions of the difficulty of the medium lead to differences in mental effort by students. As a result of the greatly increased power of computers and lowered costs for creating and delivering multimedia, there is an emerging feeling that teaching can be improved by extensive use of multimedia in large lecture courses. Students favor multimediated instruction over conventional "chalk and talk" instruction [Ansorge and Wilhite, 1994; Pence, 1993]. They rate mediated courses more highly, and the instructors therein more highly than they do traditional courses. The results of Salomon appear to be quite replicable when the media are traditional chalk talk versus highly mediated lecture. Thus far, however, demonstrated learning gains from multimedia instruction have been particularly disappointing.

We do not challenge the notion that attitude is important, but rather assert that learning is not related in simple ways to attitude. In research about learning from multiple forms of media, Clark concluded that students often prefer the medium from which they learn the least [Clark, 1982].

Some teachers may take this to mean that, if we make students a bit less happy, they may do better. That’s not what this means at all! Quite the opposite is true. Let’s say you have a course where the content is well defined by a book and some other media materials. You can present the material, or you can engage in activities to motivate the students to work with the material, especially with mastery learning as a goal. Our experience suggests that you’ll get better results if you motivate first, and then cover the material. If you already have strongly motivated students who are good self-regulators, just cover the material.

Meta-analysis is often used in education. In this research technique, quantitative criteria for research are set forth. The literature is then searched for studies meeting these criteria, and the results of all such studies are considered together. Fletcher-Flinn and Gravatt (1995) reported a mean effect size of 0.24 from a meta-analysis that included 120 studies on the efficacy of computer assisted instruction. Liao (1998) reports a grand mean effect size of 0.48 from a meta-analysis of 35 studies involving hypermedia instruction. However, most of the individual studies that were used to generate these meta-analyses used no control for active learning components. Very often, the multimedia materials included interactivity that is not possible with print materials. Early research results do not support the notion that special gains are attributable to hypermedia [Dillon & Gabbard, 1998].

It is certain that the careful design of instruction is of primary importance. Some media enable certain types of instruction. Issues two and three of the 1995 Educational Technology Research and Development contain several excellent articles on media and instructional design [Clark, 1995a, b; Jonassen et al., 1995; Kozma, 1995a, b; Morrison, 1995; Reiser, 1995; Ross, 1995 a, b; Shrock, 1995; Tennyson, 1995].

We advocate the use of multimedia, especially when it can capture phenomena and portray them in ways heretofore impossible. But remember to incorporate active learning strategies.

Smith describes learning systems, ones that demand active participation on the part of the learner. To judge the importance of multimedia to his results, multimedia issues need to be separated from active learning issues. However, this separation is difficult since there would be no realistic way to try to create a learning system like Smith's without using multimedia. What would be the outcome if the systems Smith describes were used as a lecturer’s tool in slick multimedia classes during which the lecturer made the choices and spoke aloud while the students listened? We believe that the student evaluations would be good, but little if any learning gain would be demonstrated.

Pence reports very favorable student responses to multimedia, and implies that learning improvements are likely [Pence, 1993]. His use of multimedia, however, involves brief presentations followed by cooperative learning activities between pairs of students in the class. Any significant learning gains in this environment may be more related to the active learning strategy carefully integrated with the multimedia rather than just the multimedia alone.

There are times when intuition leads a teacher to suspect strongly that a multimedia approach will be superior. This again is a face validity matter.

With the advent of powerful computer animation tools, trying to get students to be able to think of chemical phenomena in atomic and molecular terms seems ideally suited for extensive use of multimedia. In a specific test of one aspect of multimedia learning, Williamson and Abraham [1995] report substantial learning gains when animations and visualizations (created by Gelder [1994]) were used to exemplify phenomena at the atomic–molecular level. The durations of the animations were brief, as was the number of exposures to them. Using an instrument designed specifically to assess learning in this concept realm, Williamson and Abraham found significant effect size of the positive impact of using the multimedia. Just seeing the animations in lecture led to large gains in scores on the measuring instrument; no additional benefit accrued from additional access during computer lab time. However, gains did not show up on overall course exam scores, individual items, or attitude assessments. Despite the lack of improvement of overall scores, this work is cited informally by chemistry educators as support for the multimedia effort. If a dozen studies like this showed similar results, perhaps the enthusiasm of this support would be justifiable. Abraham is surprised by the large size of the effect on the measuring scale used given the brief duration of the intervention. To one who holds views that learning is neurologically based, the outcome implies either that the learning is trivial (in conflict with our personal senses of face validity on this issue) or that the assessment is somehow trivial and is missing the mark. Abraham points to similarly vexing gains in the area of creativity subsequent to brief interventions [Abraham, 1996].

The materials tested by Williamson and Abraham were developed for an AP Chemistry course taught via satellite. In spite of the quality of the materials and the demonstrable excellence of the teacher, student success rates as judged by AP scores were "average." Although geared toward AP, many students chose not to take the AP test. When used throughout a well-designed and well-delivered course, evidence of special learning gains from use of the multimedia materials disappeared [Williamson and Abraham, 1995].

You might think from the preceding work that we are media bashers. In fact, we've spent big chunks of our careers developing multimedia materials of all sorts. Media let you present conceptual materials that are difficult, perhaps nearly impossible, to present with text alone.

For example, animations afford an excellent means for teachers to convey concepts. Increasingly available, packaged animations can provide a visually stimulating and enlightening view of confusing concepts. Without animations, a teacher can talk about these concepts at the podium, fumbling with chalk or overhead projector pens to try to explain complex ideas.

Gelder’s animations convey to students the concepts of how one might view the atomic and molecular world. ChemAnimations [Gelder, 1994], and the Saunders Interactive General Chemistry CD-ROM developed by Kotz and Vining [1996, Figure 2.03] are examples of commercial materials that afford chemistry teachers the easiest way to present complex ideas with minimal teacher preparation or media materials.

Animations also may serve the purpose of presenting data from an entirely different perspective. That is, the animation may not present information to answer traditional questions in a topic, but rather to generate new questions. Students may see patterns in the data as represented by the animation, causing them to reanalyze the data along new conceptual lines:

Face validity supports still another notion, namely, that certain kinds of teaching and interaction require face-to-face meetings and cannot be conducted electronically. Evidence in support of this notion is intuitive rather than based on the literature. Learning Networks provides some (understandably biased) commentary.

Instructors experienced in this technique spend considerable effort developing discussion questions. Views about when to join in the conversation remain under study. We revisit this issue in Chapter 5.

You and your students no longer need to be time bound. A student can log on at essentially any time of the day or night. The student need not be place bound; access can be from wherever there is Internet access.

Questions about online discussions for undergraduates remain. There is one quite remarkable report suggesting that synchronous chat sessions may be especially attractive to female students [Kimbrough, 1999].

Face validity might suggest that multimedia instruction will lead to superior learning, and that replacing face-to-face discussion with asynchronous discussion will lower the effectiveness of discussions. Consistent early data suggest quite the opposite in both cases.

Learning from multimediated instruction, all other things beings equal, is similar to or possibly a bit lower than from conventional instruction. Electronic discussions, on the other hand, hold up very well when compared to traditional, in-classroom discussions, and even may have some advantages.

It is clear that the Web is a low-cost delivery system for multimedia! The Web can be nearly as passive as television in the early days of instructional TV. If one didn’t have to click now and again, it would be every bit as passive. In some ways, "clicking" the remote on a cable TV may be more interactive; one decides after each click whether to stay on the current channel, or to click again.

Unlike traditional TV, however, the Web can be very interactive. The Web’s greatest intrinsic power for teaching is that it encourages branched, nonlinear instruction. Not only can students jump around among the materials that you have created for them, they also can access materials created by others. Indeed, they can create useful materials!

Be warned, it is quite possible to access misinformation on the Web, even more easily than in your daily newspaper. Web-based misinformation is a problem. In McLuhan’s sense of the term, everyone is a publisher [McLuhan, 1964]. McLuhan's reference was to photocopiers facilitating self-publication. It can be extended to the use of the Web, where personal Web pages almost have become a given. The Web is, in large part, a non-edited, non-juried publication medium. Web page writers have few, if any, restrictions on what they may publish. Where are the Web’s editors? Who are the Web’s editors?

In our first edition, we included the sentence: "We believe that some material requires face-to-face, press-of-the-flesh instruction." Frankly, we've been surprised at some apparent successes of Web instruction, especially in areas that we perceive to be value laden.

Opinions don't influence the mass (in grams) of a sodium atom, so this content can be handled at a distance. Technical content, and especially content that is convergent (i.e., has a single answer), is readily taught on the Web. Without ever leaving his or her home, a teacher could handle 50 or 100 graduate students effectively in a course with technical content that now has just 10 or 20 enrolled students.

The Web (Internet or intranet) is going to be used for instruction regardless of what teachers think, feel, or do. It will be used even in the absence of demonstrable research support. As a communications system, especially relative to conventional teaching, it is less expensive. At the college level, many costs including hardware, Internet linkup, cleaning, and air conditioning are transferred from the school to the student. Web-based instruction has happened, and it's growing.

As the designer of Internet multimedia materials, there is a model that can serve you well. Think of the teacher as the server: what will the server do in response to a client (student) question or request? What should the student see (and hear)? What options should the student have? Should you empower them in a particular instance, or should you insist that they come to you for information and service? You need to decide where to put your programming efforts, if any, on the server side, or the client side.

A Web server is shown in Figure 2.04. Notice that the device is no different from other typical desktop computers. In fact, a server need not even have a dedicated monitor and keyboard!

Most of what users see on the Internet is client-side material. Branching from one page to another involves returning to the server. But many leaps are within pages. Most browsers support client-side maps that empower the user to make choices at the client side without returning to the server.

Interactive teaching requires that students create and transmit information, ask open-ended questions, and so forth. In the best of teaching situations, your learners will have browsers that use powerful tools you’ve created for them to accomplish these tasks.

One feature of the Web is that, given free access, students can range far and wide. Judah Schwartz [1995] has discussed this issue. In conventional curricular materials, teachers had an excellent idea of where students had been, and what they knew – at least as far as a particular body of content is concerned.

If we don’t know where they have been on the Web, how can the teacher have an idea about what the students know? Schwartz speaks of "the voice of the author," and suggests ways in which we might design software so as to provide very considerable freedom for readers (students) without giving them completely free range. Providing learners with choice is a tricky issue for teachers. Most learners, especially new learners, often become lost when provided with free choice. However, providing choice under controlled circumstances, or creating the illusion of choice, often has positive learning outcomes. The overall effects tend to be small, with results often reported as approaching significance. In a review of several quantitative hypermedia studies, Dillon & Gabbard [1998] suggest that the nonlinearity of hypermedia learning environments has negative learning outcomes for weak students. In our view, the generalizations suggested by Hannafin [1984] still provide reasonable advice with respect to learner control versus lesson control. Jacobson & Spiro [1995] make the observation that learning for conceptual transfer is better-served by hypertext instruction, while learning for memory is better-served by linear instruction.

A major advantage of the Internet is that teachers do not have to worry about software and hardware issues as much as in other situations. Netscape Navigator and Internet Explorer are powerful browser programs available for several platforms. There are dissemination issues, but cross-platform compati-bility is reasonably easy over the Web.

If you want your students to download a spreadsheet file, or a formatted word processing file, you must decide what application/format to use. Teachers and students will be sending materials like wordprocessor files and spreadsheets back and forth. Forms, data, and reports must be transferable to and from the students. The net result is a need for platform independence. Since many software developers now write for cross-platform compatibility, a careful choice can often eliminate the need for picking a particular platform. Cross-platform compatibility is an important issue in Web teaching. Thankfully, the Web browser and other software developers are aiming at its accomplishment.

Web-based instruction is coming whether we think it is a good idea or not. No vote will be taken. Many costs actually are lower. Except for a few cases – such as the United States Medical Licensure Examination – there are no quality controls in education.

With a few word changes, perhaps the same sentiment noted by Hayes et al. in reference to reading materials could be applied to Web teaching. Times are certain to change. Imagine renting one electronic line to your house and using that for nearly all of your communication and video entertainment. That is one emerging change; costs are acceptable.

In the face of rising construction costs, legislators seem to be far less favorably disposed to build new schools than they were 30 years ago. But communications costs are decreasing. As a result, legislators seem anxious to incur the savings likely to be created as a result of using the Internet as an educational medium.

From our vantage point, Web teaching is here to stay – at least for a while!