Media Kelompok 5

1. Learning and instruction

1.1 Basic concepts

1.1.1 Communications and learning

Variousjheories or models for communication have been put forward. They areof • two main types: psychological models, which examinc the interaction betwcen the learner and his environment (who says what to whom, under what conditions and with what eflect); and engineering models, which cxplain the process in such terms as 'input', 'output' and 'mcssage' and use analogies lo communication in electronic circuiu or servo-mechanisms. r- An example of the engineering kind of model is represented diagrammatically in 'f Figure 1.1. This is the Shannon-Weaver model (1949). A message is generated at source and is transmitted by some medium to a receiver at the destination. Any other irrelevant or distraeting messages being received are referred to as 'noise' in the communication system. The objective of eflective communication is to maintain the maximum 'signal to noise ratio' in the system. So far so good; the question is how to do this. A complex mathematical theory of communication has been developed on the basis of this model and its derivatives. It purports to cxplain communication in all its forms, human, electrical and so on. It has led to the design and produetion of machines which learn and machines which teach. The new seience of cybernetics is largely concerned with the application ofsuch communication theory to the explanation and simulation of human thought processes. Practical results so far include computers that play chess and learn from their mistakes to play a better game, and complex training devices which adapt the course of training on the basis of the individual learner's error pattern. Apart from such devices, which are in the main stiil only used expcrimentally, the average teacher is likely to feel very little direct impaet of cybernetics in the classroom for some tirr.e yet. However, the simple concept of signal to noise ratio may prove useful in considering the effects of various teaching procedures.

Two important points emerge frotn this classification:

1. The 'sensory channel' classification is fairly precise. Most people would agree where to place a given medium in this classification. The 'teacher control' classification is much less precise. Certain films, for example, being short and in clearly defined sections, are more under the teacher's control than others (in the sense that the teacher may think of many ways of showing the film, aitering the sequence of parts, stopping for a discussion, etc).

2. The 'teacher control' factor is but one of a large numberofsimilarfactors that one might have used for.classifying the media. Examples ofother factors which are particularly important to consider are: large groups, small groups or individual study; fixed or variable rates of presentation; sui tability for stable subject matter or for subject matter which often changes; learner-control etc.

This classification is not therefore particularly useful as it stands. One would require a multi-dimensional classification and even if one-could devise such a comprehensive classification, different people would no doubt wish to place particular media in dilFerent positions.

The sole purpose of presenting Figure 3.2 here is to illustrate that the choice of media is a comp!ex decision, influenced by a variety of factors, and therefore no quick and easy all-embracing rule is likely to be developed. Also it illustrates that the search for a classification or taxonomy of media is a counter-productive exercise, if approached in terms of the normal 'labels' attached to media (ie film, TV, concept loop). If we need further evidence of this, we need only look at the results of half a century of media research.

Whenever media are compared we usually get 'no significant difference' results. Peggie Campeau (1972) suggests that this may well be due to the 'blanket' nature of the research. One medium favours the achievement of some of the course objectives while the other favours the achievement of others. Thus existing differenc'feTKrSM out in the overall experiment. Another reason may be that the subject matter chosen for experiments may pre-judge the result. Naturally you compare two media on a topic where both have a reasonable chance of success. No one would set up a comparison of a printed book and a tape recording for a course on bird-song recdgnition: one medium is obviously inappropriate. So you choose an experimental topic which does not seem to favour either medium particularly, and are then surprised when no significant differences are found in the experimental results.

The 'mythology' of media must be stripped away. A particular medium, be it reality, book or BBC broadcast, is no more than a particular form of 'packaging' for a set of instructional stimuli. The quality of the packaging should not be neglected (ask any ad-man), but what really matters is the functional aspect of the package. Can the package hold the contents - is the medium capable of presenting the instructional stimuli required for learning? Can it arrange for the students to engage in the required learning activity? Very often, no one medium is capable of presenting all the required stimuli, so we are led to prescribe a 'multi-media package'. However, the only scientific justification for a multi-media approach is that one medium either is not capable of presenting all the required learning stimuli, or is not capable of eliciting all the required student responses.

The thoery that presenting the same information in several media enhances the learning process, though widely held, is little supported by experimental results. Similarly, experiments on the quality of a presentation (eg should we use colouror black and white TV) give inconclusive results when performed under controlled conditions./The trend of media research so far appears to lead to the conclusion that (as far as cognitive objectives are concerned) learning is influenced by the quality of the presentation only to the extent that the quality influences theclarity of the message.

We shall go more deeply into specific research findings related to specific media in relevant Iater chapters. We will see that most of the findings that seem to be be done with it) or to questions of the efficiency with which speahc media can be used to perform specific instructional tasks. In this chapter, we shall consider the general research on media as promoters of more eflective learning. We shall see that, as our purpose is merely to summarize the useful findings, we shall not take up much space at this point

Since Peggie Campeau, many other researchers have devoted themselves to reviewing and summarizing the thousands of researdi studies that the media efTectiveness question has spawned over the last 70 years or so. Early reviews tended to perform a sort of'head count' of all reported studies, weed out those that were obviously too unreliable in terms of their experimental design and then classify the others into three categories - medium A more effective, medium B (usually the mythical 'traditional instruction') more elfective, and thirdly 'no significant diflerence'. After several such studies consistently gave between 70% and 80% in the 'nsd' category with the remainder divided pretty close to equally between the other two categories, researchers stopped to think.

What they came up with was the 'meta-analysis', which involves the statistical combination of the results from various studies into one, more thorough analysis-if one solitary study shows little then perhaps a hundred similar studies, combined, will indicate a significant trend. In this field, perhaps the mostprolific meta-analyst has been James Kulik, who, together with his collaborators, has meta-analysed just about evcry significant media/technology trend ofthe last decade, including the Keller Plan (1979), the Audiotutorial (1979), programmed instruction (1980), computer-based college teaching (1980), visual-based instruction (Cohen, Ebling and Kulik, 1981), computer-based education in secondary schools (1983), ditto in elementary schools (1985), and adult education (1986). Unfortunately, despite all this efibrt and statistical sophistication, the conclusions persist — the majority of studies individually show no significant difTerences and when 'pooled', yield, at best, only slight advantages for the innovative technologies. For example, the study on visual-based instruction anajysed 320 research studies on all manner of audiovisual-versus-teacher comparisons, selected 72 that were soundly designed and comparable, and found that 74% of these showed the dreaded 'nsd' result. On the other hand, 75% ofthe remainder registered significant difTerences in favour ofthe mediated instruction. But hang on, that is only 19% ofthe total. And how significant is significant? Kulik reports that the size of these 'statistically significant' gains is of the order of 1.5 percentage points on a final exam. There is a world of diflerence between statistical and practical significance!

The upshot of such studies is that researchers at last agree that they have been asking the wrong questions all these years. But they do not yet agree on what are the right questions. For some it is the study of visual literacy, or semiotics. For others (eg Salomon, 1979) it is the pursuit of media 'attributes' defined by the , symbol systems they operate with. Yet others, for example Richard Clark* (1983, 1985), have come to the conclusion that no media, in their own right, haveany influence on learning efTectiveness, but are mere 'vehicles' for more, or less, well designed instruction.

Having said that, however, who would deny that they get more pleasure (and perhaps spend more time) watching colour TV than black and white TV? The completely unexpected speed with which colour television became established cannot be simply explained as 'keeping up with the Joneses'. Pressures built up which caused the BBC to go for colour in their schools broadcasting several years earlier than planned. Furthermore, who would deny that a multi-media presentation, ifwell put together, is more enjoyable than a presentation usingonly one medium throughout? It would seem therefore that design factors which influence the quality ofthe 'package' may have a very important role in helpingto ensure that learners participate in the presentation, and thus may help to achieve the course objectives in the open, non-experimental, situation.