ITForum Paper #2

Instructional Transactions in CBI:

A Discussion and Application of

Merrill's Definition

Dr. Barbara Grabowski
Pennsylvania State University

Like some research papers(??!), computer-based instruction, while following a recommended instructional sequence, namely rule-example-practice, can be pretty deadly, but it doesn't need to be. So, what are the factors you can think of which would decrease "deadly" and increase the probability that someone would first attend to, and then interact with information on the screen? What comes to mind? Flashing bullets? Animation? Humor? Novel stories embedded to cause dissonance? Others? Most often, the response to this question is some kind of on-screen computer-generated action. In fact two big reasons why this response is given, in my view, is that too much time is spent worrying about what the screen is going to look like and not enough time on what the message says. Too much time is spent on trying to make the computer do fancy things, and not enough time on how to coax the learner into thinking and doing.

Besides many tenets which guide instructional design, there are three major assumptions which are foremost in guiding my thinking about how to engage learners in an instructional and learning dialogue:

The purpose of this discussion is to examine the model presented by one respected researcher in the field for advice on how to address the issue of creating engaging, motivational instruction which follows the above assumptions about learning--namely that of Transactions ala Merrill. My purpose in this essay is not to present an expository article on the points brought out by him, but rather to engage you in a discussion of those points as they are translated into non-deadly computer based instruction. After all, electronic discussions should be transactive and motivational, too?

The topic for this essay was stimulated by a current class discussion in an advanced design for multimedia technology course of the Merrill chapter in Jonassen, D.H. (Ed.). (1988). Instructional designs for microcomputer software. Hillsdale, NJ: Erlbaum. No one can deny that the examples in the chapter, are very dated and not so innovative. Nevertheless, the points made are very intriguing and valuable, indeed, and can stimulate innovative thought.

So, what's the point of Component Display Theory? If you come away from reading Merrill's chapter with the idea that the key to the chapter was applying Component Display Theory, in my view, you have missed the pearls. The pearls lie in the second half of the chapter in which he describes prescriptions for an authoring system. This half of the chapter was originally published as an article in the Journal of Computer-Based Instruction (1987) and then reprinted in Merrill's definitive work (1994), Instructional design theory.

Let me first summarize the pearls of wisdom, then offer interpretation, and application:

What type of interactions, then, comprise a transaction? Three important criteria appear to be determinants--the degree of mental activity, first and foremost, second is the type of response required by the learner, and last is the type of response provided by the computer. Each of these criteria can be viewed on a continuum of low to high interaction:

To be completely in line with this definition, a transaction must have some aspects of all three criteria present.

Therefore, in attempting to respond to the question of what types of interactions would constitute a transaction, consider this scenario--a student is asked to respond covertly (i.e., did you see it?) to some computer bell or whistle used to attract attention, with the computer responding simply by proceeding to the next linear output in the sequence. It seems this application of the criteria does contain some aspect of all three criteria, but would fall severely short of measuring up to the attributes specified by Merrill's definition, thereby violating the spirit of his intent. While this feels like "a necessary, but not sufficient condition" there are three important issues brought out by this example--(1) each criteria is characterized by the lowest level of interaction; (2) an uncertainty exists that, with covert responding, the learner actually mentally processes the information; and (3) this transaction contains only one interaction.

[What other issues can you see?]

Each Criteria Characterized by the Lowest Level or Type of Interaction:

Criteria 1: Level of Mental Activity

Is attracting attention a sufficient level of mental activity to be included in an instructional transaction? Anderson and Meyer (1988) would say that attracting attention would be sufficient to constitute a communication interaction. This does not mean that a high level of communication has occurred, just that there is evidence of communication. Having a student reflect on the issue at hand at any level--from focusing attention (initial curiosity), through coding, organization, integration, translation, or evaluation--constitutes an extremely necessary part of the transaction equation. Providing only low levels of cognitive activity, however, translates into only low level interaction, and therefore, would not be sufficient in a complete transaction, unless, perhaps, the cognitive level required by the task the student is attempting to learn is at the level of discrimination. A better strategy would be to stimulate higher levels of mental activity by asking the learner to think about an example, note similarities or differences, summarize, create an analogy, rearrange information, generate an application, etc.

Criteria 2: Type of Learner Response

Given that all levels of mental activity are appropriate, is covert responding equally as appropriate as a required overt response? While some I am sure will disagree, I believe very strongly that covert responding is appropriate in instruction (including computer based instruction) if the external conditions stimulate some level of mental activity (see the above section for a discussion of level of mental activity). Perhaps an appropriate strategy would be to attract attention through covert responses and maintain attention by requiring overt responses. However, I like to think beyond that. While it is impossible for the computer to respond to a learner's thoughts that are not directly converted into action (is that next year's computer?), I am not convinced that a clever designer cannot design instruction that takes advantage of covert thinking. Does this smack of subversion? Take for example, in computer-based instruction, a learner must respond overtly in some way--even with covert processing, one assumes that the learner will be pressing some key to continue. What I am suggesting is that the learner be asked to process information at a higher level, and then be asked only to verify their understanding through a lower level overt response. Will this work? Take for example, the learner is asked to compare and contrast two images on the screen. They are asked if they noted any differences—yes or no. If they say no, the computer responds by providing advice that others have noted differences, giving a hint as to the area to explore further, and asking them to think again. OR, they are then not asked to type in anything substantial, but rather asked to click on the areas of discrepancy. This is a think "off line," respond on-line strategy.

[What other strategies can you think of?]

Criteria 3: Type of Computer Response

There are three questions/issues to consider here:

(1) Does a computer response to learner input consisting of a linear progression to new information added to a screen constitute a sufficient condition for this criteria to qualify as one component of a transaction? This issue caused great discussion in our class. In my view, a computer can provide the appearance of responding to different individual inputs which vary from choosing different parts, processes via buttons to clicking and dragging images on the screen to free response and input. Sophisticated branching, while nice, is not a necessary condition in any part of the transaction for this criteria to be met. But, herein lies the challenge for creative thinkers, designers, and programmers. I believe that there are strategies that we can use to judge free response--such as providing a bank of expert responses and putting the onus of responsibility on the learner to evaluate his/her own response, highlighting key words in the learner's response to indicate areas of agreement with an "expert," or by having the learner use the graphical user interface to draw a graphic on which the "correct" answer could be overlaid. This is an area for rich discussion over the next week. One student in previous years suggested that by having the student's responses stored to be available to an instructor may provide enough motivation to the learner so he/she will not just respond with gibberish—one of the very real drawbacks of this approach.

(2) Is there any assurance that the learner actually mentally processes the information? The problem with covert responding is that the author cannot be certain that mental activity is occurring. The only way around this is to first create a situation in which it CAN occur, and then try to maximize the probability that it WILL occur. Insights on how to do this can be gleaned from Keller's ARCS model (1988) using familiarity, human interest language and graphics (such as using personal pronouns), and inquiry arousal. Merrill suggests rhetorical questions, attention focusing information, and experiential presentations.

(3) Can one interaction constitute the compete transaction? My view on this issue is that one interaction can constitute a complete transaction, but only if it includes higher levels of all three criteria. What is interesting about this issue, however, is that if a transaction contains more that one interaction, then the designer is freer to use all the levels of the three criteria--low and high--just as long as higher levels of interaction are included. In fact, I have been very impressed with some of the ideas that have been presented in my class (perhaps they will share those this week) which require only low level interaction, but was bothered by the fact that they didn't appear to follow Merrill's guiding principles to the letter--and therefore could not be considered a transaction. Taken alone, like the first example which stimulated the identification of the issues for this discussion, low levels of interaction are not sufficient, however, used in combination with other interactions, they would be.

To sum up this analysis, I would conclude that to meet the conditions of a transaction, all levels of each criteria can be present, and some inclusion of the activities from the extreme right side must be present. Therefore, evaluating the value of a transaction can only be done by evaluating the transaction as a whole, rather than by discrete segments of the transaction.

Some final thoughts:

There are three conditions required in a transaction:

(1) mental activity—-providing a focus on WHAT the learner is thinking about as they view each component of a transaction. It should be more than just reading;

(2) response level--If an overt response is not possible all the time, require covert responses; and

(3) the computer must respond to different learner input in different ways (broadly defined!). I like to advise my students to break out of old paradigms of thinking about the best way to design instruction. The key is that if you (the computer) is telling them, and not guiding them, it is not a transaction. A performance support system, therefore, would not be classified as a transaction. And I am sure many of you may disagree!

[In your view, what is a transaction, and what are some examples?]


Anderson, J.A. & Meyer, T.P. (1988). Mediated communication: A social action perspective. Newbury Park: Sage Publications.

Gagne, R., Briggs, L. & Wager, W.W. (1992). Principles of instructional design (4th ed.). Fort Worth: Harcourt, Brace, Jovanovich.

Keller, J. M. (1988). Use of the ARCS motivation model in courseware design. In D.H. Jonassen (Ed.), Instructional designs for microcomputer courseware. Hillsdale, NJ: Lawrence Erlbaum Associates.

Merrill, M.D. (1987). Prescriptions for an authoring system. Journal of Computer-Based Instruction, 14(1), 1-8.

Merrill, M.D. (1988). Applying component display theory to the design of courseware. In D.H. Jonassen (Ed.), Instructional designs for microcomputer courseware. Hillsdale, NJ: Lawrence Erlbaum Associates.

Merrill, M.D. (1994). Instructional design theory. Englewood Cliffs, NJ: Educational Technology Publications.

Wittrock, M. (1974). Learning as a generative activity. Educational Psychologist, 11, 87-95

Particular thanks are extended to my design class for their insights and challenges in this discussion--David Birdwell, Shelley Canright, Mark DiRocco, Al Folsom, Chih-Chiah Hsu, Roland Isnor, Glenn Johnson, Yi Jung, Ed Kleinman, Marilyn Lake-Del Angelo, Chih-ming Lee, Marina Samouilova.

Barbara Grabowski
Penn State University


The buttons that appear below will be found at the bottom of each page of the discussion. The first button will take you back to the previous page (in this case, to the beginning of paper #2). The middle button will take you to the ITForum home page. The last button takes you forward into the discussion as it progressed on-line.