Adaption-Innovation as a Measure of Cognitive Style
Adaption-Innovation as a Measure of Cognitive Style
Although the possible educational significance of differences in cognitive styles has been widely discussed there has been little substantial research in recent years. The problem seems to be that many of the constructs have been inadequately conceptualised and most of the test instruments display poor reliability. It is suggested that the use of the Kirton measure of adaption-innovation, which has an established reputation and research record in the fields of personnel selection and management training, might help meet the need of researchers in education.
Although there has been widespread agreement that cognitive styles might be important variables in influencing learning (e.g. Kogan 1971, Witkin et al 1977, Guilford 1980 and Messick 1984) there has been little solid research evidence in this field establishing the effect of such styles.
The main problems have been that neither the theoretical constructs nor the test instruments have been adequately refined. One need is to distinguish between style and ability. An ability can be seen to be unipolar, starting from the zero position of not possessing the ability at all, and then moving in a positive direction with increasing ability. By contrast, styles are bipolar, with each pole representing a possible preference for dealing with the test situation. We can reasonably attach values to measures of ability as there is no advantage in having low ability, but cognitive styles are neutral in the sense that any point on the preferred style continuum might be more productive according to the nature of the task. In this event we would expect measures of cognitive styles to be orthogonal to measures of ability. A further difference is that a person cannot by definition score on a measure of ability at a higher level than they are at, but a person can display coping behaviour in working for a short time with a style other than that they prefer. A distinction must be made, therefore, between the preferred style of an individual and the actual style manifested in a specific context. The adoption of coping behaviour is likely to be uncomfortable for the individual who will therefore revert to the preferred style as soon as possible. The underlying cause of individual preference is assumed to lie in the personality, so overall cognitive styles should change no more than the personality changes.
The failure to devise constructs and measures which meet these criteria can be demonstrated in the instances of two of the better known cognitive styles, field dependence-independence and convergency-divergency.
The concept of field dependence-independence has a long history going back to Asch and Witkin (1948). Originally conceived as a measure of body orientation, of being able to identify true horizontal and vertical when sitting in a tilted chair, it became elaborated to being a measure of distraction by the field around a task situation (Witkin et al 1962). For most research purposes a paper-and-pencil measure, The Group Embedded Figures Test (GEFT) (Oltman et al, 1971) is employed, The task is for the subject to pick out a simple shape which is concealed within an elaborate drawing.
A number of problems arise in the research application. Although the original construct is that of a style, the test results show that the GEFT scores are closer to those of an ability, as field independence seems to confer advantage in the resolution of almost any task (McKenna, 1984). It has been suggested by Vernon (1972) and by Linn and Kyllonen (1981) that the GEFT scores contain two factors, so that one may be of ability and one of style. Furthermore, Collings (1985) has shown that coaching seems to bring about a permanent change with subjects becoming more field independen. As indicated before, cognitive styles being a manifestation of personality characteristics should not be responsive to coaching, other than the short-term and contextualised case of coping behaviour. Certainly, many research reports describing the use of GEFT indicate a degree of uncertainty about its application (e.g. Howe, 1989), while theorists such as Guilford (1980) have offered alternative interpretations to what field dependence-independence is really about.
The concept of convergency and divergency attracted research attention for only a few years, starting with Getzels and Jackson (1962) who compared scores obtained on conventional closed tests, i.e. those whose items yielded only one acceptable answer, and open ended tests in which a variety of answers are acceptable, and value is attached to providing as wide a range as possible of responses. They found that the mean correlation between the open and closed tests was +0.26, whereas the correlation between the five open ended tests was +0.30. On the basis of this modest distinction they postulated the existence of two cognitive styles, convergency when the subject did relatively well with the closed tests, and divergency with the reverse bias. Viewed against rigorous research standards Getzels and Jackson had failed to establish their case. The open ended tests were supposed to measure a single creativity factor yet only about 0.1 of the variance could be attributed to a single factor. If a research student came up with such findings today we would probably say that it was an interesting concept, but it lacked empirical support, and further psychometric refinement was needed.
There were other difficulties. The original Getzels and Jackson study used an unrepresentative population of high ability pupils from one school. Another problem found by Hudson (1968) was that pupils who usually seemed convergent often showed a big change in their response if the test conditions were modified, making test reliability suspect. It was not clear when it was that they were showing coping behaviour. Within a few years the notion of convergency-divergency became widely discredited. This reaction is unfortunate because there may well be a valid difference in cognitive style embedded in these studies. Support for this possibility comes from Messick (1977), who in listing over twenty possible descriptors of cognitive styles, provides extensive references to convergent, serial or analytical thinking in contrast to divergent, holistic or global thinking. With such weight of evidence it is reasonable to assume that there is some valid underlying construct.
Meanwhile, experiences, such as those outlined above, have left many educational researchers frustrated by the belief that cognitive styles maybe important, but they lack the instruments to carry out empirical study effectively. In these circumstances consideration should be given to work outside mainstream educational research and the use of the Kirton measure of adaption-innovation is proposed as a possibility.
THE KIRTON ADATION-INNOVATION INVENTORY (KAI)
For more than twenty years Michael Kirton, of the University of Hertfordshire, has been researching the cognitive style of adaption and innovation. In essence the style refers to the behaviour of a person who is forced by external circumstances to make some change in their way of working. An Adaptor may accept or even welcome change, but would try to work within the prevailing paradigms, policies or agreed frames of reference, and modify existing practices just enough to cope with the changed circumstances. An Innovator tends to welcome the opportunity to have a radical rethink to see what novel practices can now be introduced.
Kirton’s cognitive style measure (KAI) is a thirty-two item paper-and-pencil inventory scoring on a five point Likert scale. The possible range for scores, therefore, runs from 32 to 160. The overall mean for the British and American, plus eight other national (language) samples (N=2744) having population means about 95 (range 43-149), Kirton, 2003, Adaption-Innovation in the Context of Diversity and Change, Appendix 6, Table A. However, as expected by Kirton’s problem solving theory, this mean neatly varies between occupational groups from about 84 to 114 (Table J). Within the general population there is a gender difference with males scoring between four and seven points more towards the innovative pole, but the bias is (as expected by the theory) reversed within some specific occupational areas, e.g., when women enter jobs hitherto considered to be “male roles” – these women (Table F) score slightly higher than the male average! Standard deviations for the various populations are in the range of 15 to 20 points. The scoring procedures are such that a high score is characteristic of the Innovator. There is no value attached to the scores, though, as in different contexts either adaption or innovation might be the most effective style. To be a works manager one may need to be an adaptor, in order to maintain business using the given workforce and plant. However, in some research fields, or in the organising of publicity, an innovative approach may be more successful.
Factor analysis reveals that there are three sub-factors contributing to the adaption-innovation measure so it is possible to record these subscales as well as the overall score. The three sub-scales are labelled O, for originality by finding it easy to generate novel ideas, sub-scale E refers to effort or work habits, and sub-scale R to whether the person tends to be rule-obeying or willing to break rules.
An introductory article on this construct and measure may be found in Kirton (1976). Fuller details can be found in a textbook (Kirton, 1989; 2003) and the test manual (Kirton, 1987), with other studies under way.
There are a number of features to give cause for confidence in using the KAI. It is a product of some psychometric refinement, shown for example by its Cronbrach and reliability value found in several studies to range between 0.84 and 0.88, extraordinarily high values for a non-cognitive test. Correlations with tests of ability are low, generally less than 0.1. Correlations with personality factors, e.g. by Carne and Kirton (1982) with those of the Myers-Briggs Type Inventory, can go up to about 0.5. This association with personality measures, but not with measures of ability, corresponds to what a test of style should produce. The final point is that the instrument has been extensively used in many parts of the world, principally in personnel selection and management training, and hundreds of journal articles and academic theses have been published about its use.
THE USE OF THE KAI IN EDUCATIONAL RESEARCH
With only a few exceptions, e.g. Kirton, Bailey and Glendinning (1991), the KAI has not been used in educational research. Recently, though, both some staff at the Centre of Educational Studies at King’s College and Michael Kirton and his associates have been exploring the possibilities. Four such studies are reported here. It is not claimed that these provide a full description of the potential of the KAI but merely give an indication of the possibilities. It is very likely many more will soon (next century!) be published.
1. To Study Management of Schools and School Departments
KAI has been mainly used to see how teams of managers work in industry and obviously it can be used in studying the management of schools. We at King’s have been particularly interested in how different departments in schools have responded to the changes enforced by the introduction of the National Curriculum. To take one example, the new Design and Technology Departments may contain staff drawn from former Craft, Design and Technology Departments alongside those drawn from Home Economics plus those trained in Information Technology. One can envisage some staff would welcome the opportunities the new curriculum provides for new ways of working while others will try to keep on as before. The main value of the measure is that it helps us to distinguish between the personal qualities of the staff involved and the institutional factors of the school management system.
From the use of KAI 1n industry we know of certain factors which affect team work. If two people have a difference in their KAI scores of more than about twenty points then they are likely to find it difficult to communicate effectively with each other. Knowing the KAI scores for people within the group allows this potential for misunderstanding and distrust to be anticipated. It is possible to organise a group so as to minimise these adverse effects, for example, by using people with intermediate KAI scores to act as bridges, or interpreters, between those on the opposite ends of the continuum. We have also encountered the situation in which all the staff in a department have nearly identical KAI scores and department meetings are characterised by such consensus that difficult issues do not get adequately debated.
2. To Study Why Some Pupils Persist in Holding on to Their Prior Beliefs
In the past two decades constructivist psychologists, particularly those working within science education, has come to dominate much of educational psychology (e.g. West and Pines, 1985). Learning is seen not simply as a process of filling a void in the mind but often involves changing the prior beliefs which pupils hold about the topic. The possession of such beliefs can cause them to resist learning and pupils may distinguish in their minds between school knowledge, which has to be reproduced in examinations, but is judged to be otherwise useless, and the knowledge which has been gained in the outside world and is believed to be useful in that world.
The ubiquity of these prior beliefs has been well established, witness the hundreds of research reports listed in the bibliography of Pfundt and Duit (1988). The pedagogic implications are not clear, it is usually suggested that one should aim to create cognitive dissonance by providing evidence which challenges the pupil’s prior beliefs. The weakness of this argument is that dissonance can produce many outcomes, and it may cause pupils to despair of ever understanding the subject (Head, 1986).
KAI allows us to see whether this pupil resistance to changing the mind is linked to the individual personally. If no such link is found, in other words the pupil is tending towards the innovative pole, then the implication must be that there is something in the classroom climate which is inhibiting learning. McNeil (1988) describes the process of ‘defensive teaching’ which may have this effect. Changing one’s mind can be a difficult process, as it involves abandoning the certainty provided by the former belief and entering a period of uncertainty before the new concepts are fully internalised. If pupils are afraid of making a mistake, or a fool of themselves, they may not have the confidence to undertake the enterprise. The KAI data should help a teacher understand what the situation might be with an individual pupil. Knowing whether the resistance to learning comes from personal or contextual causes allows us to plan appropriate pedagogic tactics.
KAI was originally developed for use with adults and we might question its use with school populations. With pupils in their mid-teens there seems to be some loss in overall reliability, although it is still about 0.7 for the overall test, which is reasonable. The main problem seems to be that we can no longer rely on the sub-scales, particularly the E scale. Michael Kirton is currently (2003) undertaking a detailed study of the responses from adolescents in order to develop a junior version of the KAI.
3. To study Sub-Groups of Pupils
The author has long been interested in personality factors linked to students choosing to study science (Head, 1979 and 1985). One difficulty has been that there is very little information about girls who opt for science.
We can generate two opposing hypotheses relating to KAI scores. Within the general population females tend to score lower, i.e. more towards the adaptive pole, than males. We also have evidence that scientists tend to be more adaptive than the general population. Adding these two effects together would suggest that girls pursuing science should be strongly biased towards adaption. The counter hypothesis is to point to the evidence that girls who choose science tend to have unusual personality qualities (Head and Ramsden, 1990) which can be explained by the fact that they have made a non-conventional career choice. In this event we might expect them to display, as has already been noted, innovative qualities. A study of eighty sixth form pupils in English schools has supported the second hypothesis more, with the mean score for female scientists being 97.1 compared to 92.6 for the male scientists.
Obviously KAI data might help throw light on other sub-groups of students, e.g. those of high ability but who are underachieving.
4. To Study Teaching Style and Learning Style
Whereas the previous three examples described studies being undertaken at King’s College this fourth example 1s drawn from the current work by Kirton.
It has already been noted that people who have very different KAI scores are likely to find communication difficult. This fact might explain the situation in which some pupils find it difficult to learn from a given teacher, even though the pupils are able and well motivated, and the teacher is usually judged to be effective. If such mismatch does occur and leads to learning problems then it opens up the crucial question whether there are pedagogic tactics available to alleviate the problem, possibly by the teacher being able to employ coping behaviour. Part of this study has to been to develop measures of teaching styles and learning styles to go along with KAI to see how these all relate to classroom interactions and outcomes.
The sorting out of cognitive style constructs and test instruments is a major task and it is not suggested that the use of the KAI will map all the territory. There is still the need to sort out what are the key variables from the total of more than twenty recorded in the literature and then develop appropriate tests. What the KAI does give us is a psychometrically refined instrument with which we can probe part of the territory. From the examples quoted in this article it does seem to give us a powerful addition to our range of research tools and one which is relevant to a number of educational contexts.
Asch, S.E. and Witkin, H.A. (1948) Studies in space orientation: II perception of the upright with displaced visual fields and body tilted, Journal of Experimental Psychology, 38, Pp. 455-477.’
Carne, G.C. and Kirton, M.J. (1982) Sty1es of creativity: test-score correlations between Kirton adaption-innovation inventory and Myers-Briggs type indicator, Psychological Reports, 50, pp. 31-36.
Collins, J.N. (1985) Scientific thinking through the development of formal operations: training in the cognitive restructuring aspect of field-dependence, Research in Science and Technological Education, 3, pp 145-152.
Getzels, F.W. and Jackson, P. W. (1962) Creativity and Intelligence (New York, Wiley).
Guilford, J.P. (1980) Cognitive styles; what are they? Educational and Psychological Measurement, 40, pp. 715- 735.
Head, J. (1979) Personality and the pursuit of science, Studies in Science Education, 6, pp. 23-44.
Head, J. (1985) The Personal Response to Science (Cambridge, Cambridge University Press).
Head, J, (1986) Research into “alternative frameworks”: promise and problems, Research in Science and Technological Education, 4, pp. 203-211.
Head, J. and Ramsden, J. (1990) Gender, psychological type and science, International Journal of Science Education, 12, pp. 115-121.
Howe, A. C. (1989) Imagery, cognition and spatial ability, in: P. Adey, J. Bliss, J. Head and M. Shayer (Eds) Adolescent Development and School Science (Lewes, Falmer Press),
Hudson, L. (1968) Frames of Mind (London, Methuen).
Kirton, M. J. (1976) Adaptors and innovators: a description and measure, Journal of Applied Psychology, 61, pp. 622-629.
Kirton, M.J. (1987) Kirton Adaption-Innovation Inventory Manual, (Hatfield, Occupational Research Centre).
Kirton, M.J. (Ed) (1989) Adaptors and Innovators (London, Routledge).
Kirton, M.J. (2003) Adaption-Innovation – in the Context of Diversity and Change (London, Routledge).
Kirton, M.J., Bailey, A. and Glendinning, W. (1991) Adaptors and innovators: preference for educational procedures, The Journal of Psychology, 125, pp. 445-455.
Kogan, N. (1971) Educational implications of cognitive styles, in: G.S. Lesser (Ed) Psychology and Educational Practice (Glenview, Illinois, Scott, Foreman and Co.)
Linn, M.C. and Kyllonen (1981) The field dependence-independence construct: some, one or none, Journal of Educational Psychology, 73, pp. 261-273
McKenna, F.P, (1984) Measures of field dependence; cognitive style or cognitive ability?, Journal of Personality and Social Psychology, 47, pp. 593-603.
McNei1, L.M. (1988) Contradictions of Control (London, Routledge).
Messick, S. (Ed) (1977) Individuality and Learning (San Francisco, Jossey-Bass).
Messlck, S. (1984) The nature of cognitive styles: problems and promise in educational practice, Educational Psychologist, 19, pp. 59-74.
Oltman, P.K. Raskin, E. and Witkin, H.A. (1971) Group Embedded Figures Test (New York, Consultlng Psychologlsts Press).
Pfundt, H. and Dult, R. (1988) Bibliography: Alternative Frameworks and Science Educatlon (Kiel, IPN).
Vernon, R.E. (1972) The distinctiveness of field independence, Journal of Personality, 40, pp. 366-391.
West, L.H.T. and Pines, A.L. (Ed) (1985) Cognitive Structure and Conceptual Change (Orlando, Academic Press).
Witkin, H.A., Dyk, R.B., Faterson, H.F., Goodenough, D.R. and Karp, S.A. (1962) Psychological Differentiation (New York, Wiley).
Witkin, H.A., Moore, C.C., Goodenough, D.R. and Cox, P.W. (977) Field-dependent and field-independent cognitive styles and their educational implications, Review of Educational Research, 47, pp. 1-64