• Recent Posts

  • Constructivist Theory applied to Instruction: the continuing debate.
  • Learning Management in a Historical Perspective.
  • Instructional Design, Learning Objects and SCORM revisited
  • MOBILE PHONES IN AFRICA: TRANSFORMING SOCIETY AND (MAYBE) EDUCATION
  • Dare We Practice What We Preach?

• Archives

  • January 2008
  • December 2007
  • July 2007
  • May 2007
  • April 2007
  • November 2006
  • October 2006
  • July 2006

• Categories

  • Current Trends
  • Distance Education
  • Topics for debate

• TTS-Global

  • Technology-based Training Systems
• Search for:

First Principles of Instructional Design:
the recurring issue of knowledge vs. skill.
Alexander J. Romiszowski.
(Educational Technology 46/6, November-December 2006, pp67-71)

Forward into the Breach? Or Back to First Principles?
The first of this series of “Topics for Debate” columns focused on the issues surrounding the promotion and assurance of Quality in Distance, Flexible and ICT-based Education.  This topic was chosen as it was the theme of the then-forthcoming ICDE World Conference on Distance Education. The column raised some questions regarding such issues as whether technological or pedagogical advances were leading the quest for effectiveness, efficiency and quality improvements in the new e-learning forms of distance education. It questioned to what extent the quest for quality was different in the context of these new media, or whether there were some underlying basic principles that are equally valid in both distance and conventional learning contexts. It also questioned the long term sustainability of the trend towards collaborative group learning in online environments unless something is done to limit the extra workload this places on the teachers, the financial misconceptions that institutions tend to make about the real costs of effective e-learning systems, and some of the pedagogical misconceptions that are introduced by such catch-phrase terms as reusable learning objects, SCORM compliance, and blended learning.
Well, the ICDE22 World Conference is now over. All these issues were debated by some 1500 delegates from over 70 countries.  As could be expected, some papers presented enthusiastic visions of a new era in education, and some of the presenters did so with rallying cries that recalled those of Shakespearean knights, kings and crusaders: forward into the breach dear friends – the future is ours! However, I was interested to note the large, and increasing, number of voices crying: hold your horses – where are you going – why are you going there – let’s get back to first principles!  

Old Wine in New Bottles?
Nowhere was this more evident than in the after-effects of my own keynote debate presentation, in which I addressed the issues I had raised in my last column in this journal, related to the controversial paper on the failure of Minimally Guided Instruction, recently published by Paul Kirschner, John Sweller and Richard Clark in Educational Psychologist 41(2). I stressed that I was not necessarily in agreement with all of the conclusions of the authors, but was raising these issues as I felt it was important to seriously examine such apparently damning conclusions with respect to some of the most prominent recent paradigm shifts in our field. However, I inevitably stirred up the wrath of a few dyed-in-the-wool social constructivists, who told me what they thought of me and then refused to debate further. But for every such reaction, there were many – surprisingly many – conference delegates who declared that it is high time to reign in the gallop towards the adoption of each new educational philosophy without first examining the basic teaching-learning principles that it promotes, verifying whether the new approaches are indeed that much different from past practices, and if so, whether there is an adequate and sound research base to justify them.
The debate focused particularly on the statement in the aforementioned paper that, for many decades, each generation of educators has promoted new teaching-learning procedures under new names, but based on the same set of basic principles that had already failed to stand up to the scrutiny of experimental research. One participant labeled this as marketing old wine in new bottles. That led the debate to consider whether the wine was basically a good or a poor vintage, or whether the bottling process was at fault: whether the apparent failure to improve learning was due to faults in the basic principles underlying discovery, experiential and constructivist approaches to instruction, or to the inappropriate or inexpert manner in which such approaches are often implemented in practice. Not surprisingly, the debate did not reach any unanimous conclusions. Indeed, rather than provide clear answers to the initial questions posed, it raised yet more questions. Out oioned oned t familiar wtiongy research and development this topic. constructivist approaches to instruction, or to the inexpne question that was voiced several times through the ensuing days of the conference was: if we are to return to basic principles, what are they – where do we find a clear statement of the first principles of instructional design? This is the topic for our debate this month.    

Merrill’s First Principles.     
David Merrill has over the last few years published quite extensively on this topic, one paper published in 2002 in Educational Technology Research and Development 50(3) actually bearing the title “First Principles of Instruction”.  He has since published other papers that use the “first principles” identified in this paper to build some innovative instructional strategies, for example Levels of Instructional Strategy, that was published in the July-August issue of this magazine. However, we are interested here in the details of the “first principles” and how they were derived, so the following analysis is based on the 2002 seminal paper.   Merrill first defines his terminology, using Charles Reigeluth’s construct of “basic” and “variable” methods:
“A principle (basic method) is a relationship that is always true under appropriate conditions regardless of program or practice (variable method). A practice is a specific instructional activity.  A program is an approach consisting of a set of prescribed practices. Practices always implement or fail to implement underlying principles whether these principles are specified or not.”
The basic premise of Merrill’s paper is that a set of basic “first principles” can be found in most instructional design theories and models and that these design principles apply regardless of the instructional program or practices prescribed by a given theory or model.  He identifies such a set of basic “first principles” through the analysis of the “practices” proposed in some recent instructional design theories and models. These include the Star Legacy program of the Vanderbilt Learning Technology Center, Howard Gardner’s Multiple Approaches to Understanding, van Merriënboer’s 4-Component ID Model, David Jonassen’s descriptions of Constructivist Learning Environments, and the proposals of many other theorists published from the mid-1990’s onwards. To temper this selection of modern, or in some cases post-modern, viewpoints, Merrill also cites the writings of Johann Friedrich Herbart from over 200 years ago, demonstrating that the same basic principles may be identified in older as well as more recent work. So, what are these basic principles? Merrill identifies the following list of five that seem to be reflected in most of the instructional theories analyzed.
Problem-centered.  Learning is promoted when learners are engaged in solving real-world problems. Furthermore, learning is best promoted when learners are first shown the task that they will be able to do or the problem they will be able to solve as a result learning, when learners are engaged at the whole problem or task level, and when learners solve a progression of appropriately selected problems.
Activation. Learning is promoted when relevant previous experience is activated, by asking them to recall, relate, describe, or apply knowledge from relevant past experience that can be used as a foundation for the new knowledge. Furthermore, learning is promoted when learners are provided relevant experience that can be used as a foundation for the new knowledge, and are provided or encouraged to recall a structure that can be used to organize the new knowledge
Demonstration. Learning is promoted when the instruction demonstrates what is to be learned rather than merely telling information about what is to be learned, when the demonstration is consistent with the learning goal, when learners are provided appropriate learner guidance, and when the media used for the demonstration play a relevant instructional role and multiple forms of media do not compete for the attention of the learner.
Application.  Learning is promoted when learners are required to use their new knowledge or skill to solve problems and when the application (practice) during learning and the later posttest are consistent with the stated or implied objectives. Furthermore, learning is promoted when learners are guided in their problem solving by appropriate feedback and coaching which is then gradually withdrawn, and when learners are required to solve a sequence of varied problems.
Integration. Learning is promoted when learners are encouraged to integrate (transfer) the new knowledge or skill into their everyday life and are given an opportunity to publicly demonstrate their new knowledge or skill. Furthermore, learning is promoted when learners can reflect-on, discuss, and defend their new knowledge or skill, and when they are able to create, invent, and explore new and personal ways to use their new knowledge or skills.
Merrill quotes copious references to experimental research that supports the validity of each of these first principles in a variety of teaching-learning contexts, drawing on the work of researchers from a variety of educational philosophy “camps”. He concludes his analysis by stating that the many different theories and models reviewed in his paper do not seem to involve fundamentally different “first principles of instruction”. All the theories and models reviewed incorporate at least some, usually most, of the five cited principles, though none of them includes all five principles in all their relevant aspects. Also, some theories and models reviewed include additional principles or prescriptions that are not described in his paper.  However, no theory or model reviewed includes principles or prescriptions that are contrary to those described. Let us see where this leads us……

Is this the Egg of Columbus? Or of the Curate?

Merrill, in his paper, visualizes a basic “first principles” ID model in the form of an egg-shaped diagram, where the centrally placed egg represents the problem upon which the learning activities are to be centered, and placed around the egg are the four other first principles – I quote:  “(1) activation of prior experience, (2) demonstration of skills, (3) application of skills, and (4) integration of these skills into real-world activities”. Figure 1 illustrates the general idea. Note that the five principles are transformed into prescriptive “phases” of the instructional process, and the nature of the diagram suggests that these phases need not occur in a predetermined sequence, but may be executed in any appropriate sequence, or indeed concurrently.  The “egg” metaphor was probably not an intended part of the model’s diagrammatic representation, but I have used it here as it leads me into the next part of the discussion – introduced by the somewhat cryptic subtitle of this section. Maybe I should explain further…

Figure 1. The five Phases for Effective Instruction (Merrill, 2002)

The eggs of Columbus and of the Curate are well established idiomatic expressions of the English language. But as some readers may not have learnt English as their original mother tongue, and as at least one of the two idiomatic expressions is very common in Britain, but not so in the USA (we are said to be “two nations separated by a common tongue”), I shall quote the stories that gave origin to the expressions, using that wonder of modern information technology – Wikipedia – as the source for both.
The first story goes back, presumably, to the fifteenth century. Christopher Columbus attends a dinner given in his honor. Columbus asks all the gentlemen in attendance to make an egg stand on end. After all the men tried and failed, they stated that it was impossible. Columbus then placed the egg’s small end on the table, breaking the shell a bit, so that it could stand upright. Columbus then stated that it was “the simplest thing in the world - anybody can do it, after he has been shown how!”
The second expression “a curate’s egg” means something having a mix of good and bad qualities. The phrase originated in a cartoon in the humorous British magazine Punch on 9 November 1895. …. it pictured a curate taking breakfast in his bishop’s house (and attempting to eat a boiled egg of dubious quality). The Bishop says “I’m afraid you’ve got a bad egg, Mr. Jones” …the curate replies, “Oh, no, my Lord, I assure you that parts of it are excellent!”
Is Merrill’s “egg” similar to the egg of Columbus, or of the Curate, or maybe to both in different aspects? First let’s take the Columbus metaphor. In showing that there are underlying basic principles, tested over time, that reappear in just about any new formulation of instructional theory, Merrill draws our attention to the fact that, at least at a more general systemic level, instructional design theory is much simpler than we are sometimes led to believe.  “Good” instructional theory must be based on “valid” observations of what works and what does not. These observations may vary in specific cases due to differences in the context, content or conditions of instruction, but it is good to know that if these situational differences are allowed for and the principles are generalized, we see a basic pattern of generally applicable “first principles”.
However, in order to be able to fit so many apparently different instructional theories, based on different philosophical assumptions, into one “mould”, Merrill has to be very generous in his use of basic terminology. For example, with reference to problem-centered learning, he says: “The definition of problem varies among theorists… for some a problem is engaging in some form of simulation of a device or situation … for others it merely means being involved in some form of real world task.”
I cut the quote at this point in order to observe that the more you loosen your definition of “problem” the greater the range of different situations that will fit into the category. Any professional training program is, by definition, conceived to prepare people to perform certain tasks in the “real world” context of their jobs. Thus any such program is “problem-centered” according to the above definition. Now let us imagine the simplest and most basic professional training model/method I know – the TWI (Training Within Industry) method. This was developed during WWII as a stop-gap rapid approach to train the “unskilled” people (mainly housewives and retired senior citizens) who replaced the younger “skilled” labor that left the factories to go to war. One can see all five of the “first principles” reflected in this simple basic training approach. Similarly the most sophisticated and most (behaviorist) theory-based approach to instruction that I know – Tom Gilbert’s Mathetics – seems to pass the membership test with flying colors on all counts – especially with respect to the Demonstration, Application and Integration principles, that are embodied in the basic D-P-R (Demonstrate – Practice – Release) sequence that forms the “spinal chord” of the Mathetics model of programmed instruction.
So, is this eclectic “herding” of all existing theories and models, emanating from all imaginable theoretical and philosophical sources, into one common “stable” a good thing or a bad thing? Is it a “simple when you have seen it” but powerful insight like the egg of Columbus, or a problematic issue like the one faced by the Curate?  On one level, it is definitely good, as it shows that when it comes down to solving real instructional challenges in the real world, there are similarities as well as differences between alternative approaches – and maybe sometimes and in some senses the similarities outweigh the differences.  Therefore, let us put a stop once and for all to the partisan wars that tend to break out between different educational psychology and philosophy “camps”.  On another level, it is not all that good, as it still leaves us with the dilemma of selecting appropriate approaches for different practical situations. The same basic first principles may be reflected in alternative instructional approaches, but does that mean that any approach is as good as any other in a given specific context / for any given content / under any imaginable real-world conditions? Obviously that is not the case.  So, let us delve a little deeper.  The earlier quote, truncated midway, continues as follows: “The author’s use of the word problem includes a wide range of activities the most critical characteristic being that the activity is some whole task rather than only components of a task and that the task is representative of the tasks the learner will encounter in the world following instruction.”
It would seem that here we are beginning to see an aspect of restriction in the use of the term “problem-centered”.  Only “whole-task” instructional approaches qualify. Part-task approaches are excluded. Other exclusions appear as the above quote continues: “Problem-centered instruction is contrasted with topic-centered instruction where components of the task are taught in isolation (“you won’t understand this now but later it will really be important to you”) before introducing the real world task to the students.”  This would seem to exclude the topic-centered courses and curricula that form the bulk of current school programs in most countries. But it may also be excluding many of the small-step programmed instruction approaches I mentioned above, that are often based on a task analysis approach that breaks down a complex real-world task into a hierarchy of component sub-tasks and organizes the instruction on a step-by-step model, from the known to the unknown, or from the simple to the complex.
This raises a number of questions and issues. For example, what happens to the well developed research base on the relative merits of whole-task and part-task training for different skill categories? What happens to the accumulated knowledge-base on best practices for instruction of different categories of learning? And what about the two examples in the last paragraph: the latter is an example of a particular approach to instructional design that has a long pedigree of success and therefore, although it may not be best in all situations, should it be totally discarded; the former is, however, not an example of instructional design strategy at all, but of curriculum design strategy – is this not a very different case? It seems to me that the part of the above quote that Merrill included in brackets clearly excludes courses and programs based on topic-centered curricula, but it is not clear whether he also means to exclude task-analysis-centered approaches. Clearly there is a world of difference between the learning of some component sub-tasks a matter of hours, or even just minutes, before the mastery of the whole task (the case of most sequenced instructional programs) and learning a meaningless subtask months, or even years, before it is integrated into real-world authentic problem situations (the case of a poorly planned curriculum).
I feel that some of the uncertainty generated springs from over-generalizing the basic terminology – and therefore the concepts – of our field. Perhaps the problem(s) with understanding the issues raised by the definition of “problem” may be resolved by analyzing the definition(s) used for “knowledge” and “skill”.  Earlier, I quoted Merrill’s definitions of some of his first principles as the demonstration, application and integration of skills. Yet later in the paper, he repeats the principles, but replaces the word skills by knowledge. This is not some lapse of attention, as there is a footnote that states: “The author used the word knowledge in its broadest connotation to include both knowledge and skill, and to represent the knowledge and skill to be taught as well as the knowledge and skill acquired by the learner.”
So, the generalization not only seems to include knowledge and skill as two near-synonyms in our vocabulary – a point I will return to later – but it also seeks to include both the external connotation of knowledge as something to be found in books and other media, and the internal connotation as something constructed by the learner’s cognitive processes. This, in effect, herds a legion of different animals into the same stable – to include constructivist, cognitivist and behaviorist theories and philosophies – maybe also to include the soft technologies of instructional design and the hard technologies of knowledge management - all under the same roof. This is a fascinating topic to discuss and debate, that warrants its own column some time in the future. For this month, however, let us return to the issue of knowledge and skills.  

The importance of differentiating knowledge and skill in the ID process
It is very common to talk, both colloquially and technically, of instruction as the development of knowledge, skills and attitudes. This three-domain distinction goes back to well before Bloom, Krathwohl and others used the terms cognitive, psychomotor and affective as the basic domains of their taxonomies of educational objectives. There has since been a colloquial tendency to treat these two sets of three terms as synonymous. However, technically speaking, it is clear that they are not. Skills may be physical (psychomotor) or intellectual (cognitive), and many are a combination of both. Also, in my view, some components of the affective domain are skills rather than attitudes.
One should also mention that the original Bloom taxonomy of the cognitive domain includes both knowledge-objectives (knowledge, comprehension) and skill-objectives (application, analysis). I have been pointing out since the 1970’s that this makes for some difficulty in using the taxonomy. It is only in Anderson and Krathwohl’s recent (2001) updating of Bloom’s taxonomy that the knowledge and skill elements have been distinguished, by reorganizing the categories on a grid where one axis represents the different categories of knowledge and the other represents different categories of cognitive processing. And of course, cognitive processing is a task or activity that involves a set of skills that people may possess and develop to different degrees. 
The authors of the revised taxonomy acknowledge the contribution of my past work on knowledge and skills as a basis for some aspects of their revisions. Incidentally, they also acknowledge the earlier work of David Merrill, of about the same period, which distinguished knowledge and skill aspects of learning by means of a grid-representation of learning categories, as part of Component Display Theory (surprisingly, CDT is not directly mentioned and is indirectly almost negated in the paper here analyzed). So, I return to my writings of the 1970’s and 1980’s to pick up on a few points that might form the basis for further continuation of this debate.
First the basic definitions: knowledge is information that has been acquired, organized and stored in a reusable, and useful, manner in the mind (that is, of course the internal connotation – for the external one, just substitute the word media for mind); skill is the capacity to do something (i.e. perform) with a given degree of effectiveness, efficiency, etc (i.e. competence).If we accept these definitions, then some quite important implications follow on.
(a) Knowledge is a “go – no go” quantity. Either you have it or you do not. Or, you may have it but are unable to access and use it due to the way it is currently organized. Partial knowledge is the result of having some of the information elements or units, and not others. Learning of knowledge involves a combination of (1) the addition of new elements to the store and (2) the reorganizing of the existing, and the new, elements to form new knowledge structures. Furthermore, this learning process can often occur in a one-shot sort of manner – the sudden “eureka” phenomenon of, in an instant, reorganizing one’s ideas and acquiring a precious insight. Two people exposed to the same information (experiences) may incorporate different subsets of information elements into their knowledge structures, and even if they acquire exactly the same elements, may integrate them in different ways with previously existing knowledge. Thus people generally differ in the knowledge (both content and structure) that they possess. These internal differences are generally difficult to measure accurately and objectively, as they can only be inferred indirectly from tasks performed, or from the analysis of communication processes such as speech, writing, or artistic expression.
(b) Skill, on the other hand, is something that develops with experience and practice. The learning process for skills is seldom if ever one-shot. Repeated and appropriate practice is required to achieve higher degrees of competence. Two people may possess the same skills, but developed to different degrees. Later, after some learning process, they may still have exactly the same skills developed to higher, but again different, degrees. They may even exhibit the same skills to different degrees of competence in different contexts (tasks). As these degrees of competence are generally task-related, they may often be measured directly, accurately and objectively by analysis of the end results of specific tasks.
(c) Skilled performance, of course, requires the use of knowledge. The more sophisticated and complex the skilled activity, the more it depends on the existence of an adequate knowledge base and its appropriate utilization. The existence of an adequate knowledge base usually implies prior learning, although in some cases the learning of the knowledge may conveniently accompany the practice of the skill (these are the cases where just-in-time-training actually works). In other cases, part of the knowledge-base may be external, as a performance support system to be combined with internally stored knowledge during skilled performance, without ever being learnt.
(d) The appropriate utilization of the knowledge base (whether internal or external) may involve the use of other supplementary skills (e.g. critical thinking) and knowledge (e.g. problem-solving heuristics) that are necessary for the internal planning or control of the task. These meta-skills and meta-knowledge elements are internal and not directly observable, but their presence and power may be inferred from the competence with which the skilled activity is performed. They may also be observed and evaluated indirectly and subjectively through the process of conversation (live or online). Skilled activity can, and indeed should, also be self-evaluated by the performer. This process is called reflection. It depends on the use of specific meta-skills, sometimes characterized as the skills of conversing with oneself. The more knowledge-dependant the skill, the greater is the importance of reflection for its development.
(e) The driving forces, both economic and social, behind institutionalized education are generally derived from the value of the resultant skilled performances. Therefore, not only is it easier to do accurately and validly, but it is also more appropriate for final assessment of learning to be in terms of task performance criteria – and therefore skills rather than knowledge. The acquisition and internal organization of knowledge is then evaluated indirectly, through the measurement of the results of its application in skilled performance. Furthermore, as the acquisition of knowledge “for its own sake” seldom interests anyone other than maybe the students themselves – and even then only sometimes – we may conclude that the difficult-to-accomplish accurate measurement of knowledge-learning is not an essential part of the summative evaluation process.
(f) However, the measurement and evaluation of the internal process of acquisition, construction and organization of knowledge is more than an area of academic study for psychologists – it interests instructional designers and also teachers, as their understanding of this process and their ability to measure it, albeit indirectly and subjectively, is a key to the promotion of effective learning. It is therefore an essential part of the formative evaluation of instruction. Having said that, however, one should add that this focus during teaching on the internal processes of knowledge-learning, should not lead to the neglect of the ultimate goal of the exercise, which is to develop appropriate degrees of externally demonstrable and measurable skills-performance.  Regrettably, in some recent instructional theory literature, these two aspects apparently seem to be in conflict rather than being seen as complementary.

Comments