Download Honey & Muntford Learning Style Questionnaire Correlation with Kolb's Theory: Pilot Study and more Slides Learning processes in PDF only on Docsity!
ISSN 1101-
The Honey and Muntford learning style
questionnaire: a pilot study
J.T. Moelwyn-Hughes & Y. Sayed
INTRODUCTION
L
earning style as a way of considering how students approach their studying has received considerable attention in the literature in recent years. Claxton and Murrell (1987) have comprehensively, but rather uncritically, reviewed nearly all papers on learning styles. Two conferences held in the United Kingdom (Richardson et a/ 1987), and Australia (Bowden 1986) summarise the major debates in this area. Schmeck (1988) has edited an authoritative monograph in which all the major theorists explain their research methodologies. As the definition of learning style depends on the theoretical perspective of the research school (Richardson 1987), there is no consensus in the literature (Moelwyn-Hughes). For the purpose of this article, Ford's (1985) definitions of strategy and style will be followed: "the particular way that an individual approaches a learning task, on one occasion.. .. A tendency to use one type of strategy constantly is a learning style".
Alex Main (1985) drew our attention to the work of Honey and Mumford in the field of learning styles in management education. Their learning style question- naire (LSQ) was based on Kolb's theory of experien- tial learning. The purpose of all three studies was to ascertain whether the inventories were appropriate to South African higher education; to establish whether particular disciplines have particular learning styles; and to inform academic staff of their students' profiles as learners as a means of improving teaching. Kolb's theory of experiential learning is based on a four-step cycle (see Figure 1). This cycle involves the steps of concrete experience, (C E), reflective observation (RO), abstract conceptualisation (AC) and active experimentation (AE). Depending on how persons respond to the four dimensions, they are characterised as divergers (high CE and RO), assimilators (high RO and AC), convergers (high AC and AE) or accom- modators (high AE and CE). There are two important
characteristics in Kolb's theory of learning: first. CE and AC and AC and RO are in dialectical tension with each other and, second, all four modes should be experienced in a learning cycle.
Kolb (1976) used his nine-item learning style inventory (LSI) to determine students' preferred learning styles. Students are asked to rank four words into those which best characterise them as learners (scoring 4), those which next best character- ise them as learners (scoring 3) down to a score of 1 for the word which is least characteristic of the student as a learner. For example, the following set of words involves all four learning modes: -feeling (CE)-watching (RO)-thinking (AC)-doing (AE). Only six scores are added for the four learning modes, but in every case the scores are paired according to the CE-AC or AE-RO links. Kolb's theory incorporates three basic assumptions: LSI scores are directly correlated with a student's choice of undergraduate majors; the choice of a profession (or subject) further enhances the student's preferred learning style; and only after graduation at master's level will students learn how to integrate the four learning styles (Kolb 1984).
HONEY AND MUMFORD
Honey and Mumford (1982, 1983) devised a Learn- ing Style Questionnaire (LSQ) for their work in management education. They developed it in order to promote effective learning by the understanding and use of individual learning styles. They maintain that learning styles must be understood in the total learning context and are influenced by the environ- ment in which learning takes place. The theoretical background from which they drew their work is based on Kolb (1984). Honey and Mumford accepted Kolb's hypothesis that people develop preferences for different learning styles in the same way that they
ISSN 1101-
J -H
ill ill
characteristics in Kolb's theory of learning: first CE and AC and AC and RO are in dialectical tension with INTRODUCTION each other and, second, all four modes should be experienced in a learning cycle.
L
earning style as a way of considering how students approach their studying has received considerable attention in the literature in recent years. Claxton and Murrell (1987) have comprehensively, but rather uncritically. reviewed nearly all papers on learning styles. Two conferences held in the United Kingdom (Richardson et al 1987), and Australia (Bowden 1986) summarise the major debates in this area. Schmeck (1988) has edited an authoritative monograph in which all the major theorists explain their research methodologies. As the definition of learning style depends on the theoretical perspective of the research school (Richardson 1987). there is no consensus in the literature (Moelwyn-Hughes) For the purpose of this article. Ford's (1985) definitions of strategy and style will be followed: "the particular way that an individual approaches a learning task, on one occasion. A tendency to use one type of strategy constantly is a learning style"
Alex Main (1985) drew our attention to the work of Honey and Mumford in the field of learning styles in management education. Their learning style question- naire (LSQ) was based on Kolb's theory of experien- tial learning. The purpose of all three studies was to ascertain whether the inventories were appropriate to South African higher education; to establish whether particular disciplines have particular learning styles; and to inform academic staff of their students' profiles as learners as a means of improving teaching. Kolb's theory of experiential learning is based on a four-step cycle (see Figure 1) This cycle involves the steps of concrete experience. (C E). reflective observation (RO). abstract conceptualisation (AC) and active experimentation (AE). Depending on how persons respond to the four dimensions, are characterised as divergers (high CE and ROJ, assimilators (high RO and , convergers (high AC and AE) or accom- modators (high AE and CE). There are two important
Kolb (1976) used his nine-item learning style inventory (LSI) to determine students' preferred learning styles. Students are asked to rank four words into those which best characterise them as learners (scoring 4). those which next best character- ise them as learners (scoring 3) down to a score of 1 for the word which is least characteristic of the student as a learner. For example, the following set of words involves all four learning modes: -feeling (CE)-watching (RO)-thinking (AC)-doing (AE). Only six scores are added for the four learning modes. but in every case the scores are paired according to the CE-AC or AE-RO links. Kolb's theory incorporates three basic assumptions: LSI scores are directly correlated with a student's choice of undergraduate majors; the choice of a profession (or subject) further enhances the student's preferred learning style; and only after graduation at master's level will students learn how to integrate the four learning styles (Kolb 1984).
HONEY AND MUMFORD
Honey and Mumford (1982, 1983) devised a Learn- ing Style Questionnaire (LSQ) for their work in management education. They developed it in order to promote effective learning by the understanding and use of individual learning styles. They maintain that learning styles must be understood in the total learning context and are influenced by the environ- ment in which learning takes The theoretical background from which they drew their work is based on Kolb (1984). Honey and Mumford Kolb's hypothesis that people develop preferences for different learning styles in the same way that they
develop any other kinds of style. While they accepted Kolb's model of experiential learning, they encounter- ed problems with the procedure and face validity of his LSI. As Honey and Mumford (1982) expressed it: "The LSI is based on 36 words (not sentences) which do not describe management activities; as a basis for the attribution of styles we found them less persuasive. both to us and managers. Nor did we find the description of the four styles ." either congruent with our own experiences. or meaningful to many of the managers with whom we dealt" Their LSQ differs from Kolb's in two ways: they base their 80 items on recognisable statements of managerial behaviour; and they believe the answers scored are a starting and not a finishing point They maintain that their preference is to focus on observable behaviour rather than the psychological basis for their beha- viour. Their LSQ has a total of 20 items for each of the four learning styles. Respondents are asked to indicate whether they agree or disagree with each item. Honey and Mumford (see Table 1) maintain that the majority of the items are behavioural in nature and that the LSQ covers general trends or tendencies running through a person's behaviour.
Alex Main (1985) has summarised some of the characteristics of the four preferred styles in Table 1. Honey and Mumford postulate that four basic learning styles may occur in variable combinations in any individual. They are Activist Reflector, Theorist and Pragmatist styles. They further suggest that individuals develop a preference for one of these. Alex Main reports that an individual will have developed a unique combination of skills, attitudes and approaches to learning and that a knowledge of these will assist the learner in building on his or her strengths.
The avowed aim of Honey and Mumford (1982) was "not to produce something that is academically respectable. but to produce something which' will give detailed practical guidance to those who are trying to develop their abilities, and to those who are trying to help them". They tested the reliability of the LSQ by getting a group of 50 people to fill in the LSQ twice. with a two-week gap. Pearson's product- moment coefficient of correlation between the two sets of results was 0.89. The validity was harder to determine as there were few established question- naires with which to draw comparisons. They believe that the face validity is not in doubt because it was rare for them to encounter anyone who disputed the findings of the LSQ results. Honey and Mumford correlated their LSQ with Kolb's LSI. They found the correlations listed in Table 2 from a sample of 29 respondents.
SAJHE/SATHO VOL. 7 NO.1 1993
RESULTS
The full 80-item inventory was administered to three first-year classes in 1987 and a further two in 1988 (see Table 3). Originally the results were presented according to the number of students in each learning category depending on the highest score recorded; if two or more scores were identical, the results appear under combination. The data was presented pictorial- ly in the form of pie-charts or histograms (see Figure
The present authors prefer the alternative method used by the researchers at Alverno College and Table 3 summarises all the data collected in 1987 and 1988 using the original 80-item LSQ. With the exception of the Chemistry I (1987) class, in which all mean scores are very close to each other, Reflector is the predominant style. Using Kolb's LSI we found that abstract conceptualisation (AC) was the highest score in the same classes. Based on Kolb's underly- ing theory we would have expected to note different styles for the three different classes studied and we would have predicted that Theorist should have been the predominant learning style or type.
Table 4 compares the results obtained in this study with those originally reported by Honey and Mumford (1982, 1983) and those recorded by Lucas (1988) at the Western Australian College of Advanced Educa- tion. Perhaps, not unsurprisingly. the profiles of the two student groups are very similar, with the Australian students scoring lower on the Activist scale and Reflector as the predominant type in both cases. The scores of both student groups are not too dissimilar from managers in training to cause alarm. and would indicate similar kinds of responses on the average. Pragmatist wins by a short head over Reflector for the managers but this could reflect the position they are in. More important is the closeness of the different mean scores to each other. There is very little difference between the mean scores and. when allocating students according to style, a small number ( < 5%) had a large predominance of one style over others (as shown by a difference in raw scores of 5 or more).
When students involved in this study were handed the LSQ they were also asked to respond to two open-ended questions concerning the inventory, and anything they might like to add about teaching and learning in general. The sumll!1ary of these responses is given in Scheme 1. The fact that many of the items caused the students difficulties, coupled with the practical realisation that the questionnaire was too long (it took the best part of a 45-minute lecture
15
develop any other kinds of While Kolb's model of experiential learning, they encounter- ed problems with the and face validity of his LSI. As Honey and Mumford (1982) it: "The LSI is based on 36 words (not which do not describe management activities; as a basis for the attribution of we found them less persuasive, both to us and managers, Nor did we find the description of the four either congruent with our own or meaningful to many of the managers with whom we dealt", Their LSQ differs from Kolb's in two ways: they base their 80 items on statements of managerial behaviour; and they believe the answers scored are a starting and not a finishing maintain that their preference is to focus on observable behaviour rather than the psychological basis for their beha- viour, Their LSQ has a total of 20 items for each of the four learning styles. indicate whether they agree or with each item, Honey and Mumford Table 1) maintain that the majority of the items are behavioural in nature and that the LSQ covers trends or tendencies running through a person's behaviour.
Alex Main (1985) has summarised some of the characteristics of the four preferred in Table 1, Honey and Mumford that four basic styles may occur in variable combinations in any individual. They are Activist. Reflector. Theorist and Pragmatist They further suggest that individuals develop a for one of these, Alex Main reports that an individual will have developed a combination of skills, attitudes and to and that a knowledge of these will assist the learner in on his or her strengths.
The avowed aim of Honey and Mumford ( "not to produce something that is respectable, but to give detailed practical to those who are trying to develop their abilities, and to those who are trying to help them" They tested the of the LSQ by getting a group of 50 to fill in the LSQ twice. with a two-week gap, Pearson's moment coefficient of correlation between the two sets of results was 0,89, The validity was harder to determine as there were few established question naires with which to draw comparisons They believe that the face is not in doubt because it was rare for them to encounter anyone who disputed the of the LSQ results. Honey and Mumford correlated their LSQ with Kolb's LSI. They found the correlations listed in Table 2 from a of 29
SAJHE/SATHO VOL. 7 NO.1 1993
RESULTS
The full 80-item was administered to three classes in 1987 and a further two in 1988 Table 3) the results were
category on the highest score recorded; if two or more scores were identical. the results appear under combination, The data was pictorial- ly in the form of pie-charts or Figure
The present authors the alternative method used by the researchers at Alverno and Table 3 summarises all the data collected in 1987 and 1988 using the original 80-item LSQ With the of the Chemistry I (1987) ciass. in which all mean scores are very close to each other, Reflector is the predominant style, Using Kolb's LSI we found that abstract conceptualisation was the score in the same classes. Based on Kolb's ing theory we would have note different styles for the three different classes studied and we would have predicted that Theorist should have been the predominant style or type,
Table 4 compares the results obtained in this study with those by Honey and Mumford (1982, 1 and those recorded by Lucas (1988) at the Western Australian of Advanced Educa- tion. Perhaps, not the profiles of the two student groups are very similar, with the Australian students scoring lower on the Activist scale and Reflector as the predominant type in both cases. The scores of both student groups are not too dissimilar from managers in to cause alarm. and would indicate similar kinds of responses on the average Pragmatist wins a short head over Reflector for the managers but this could reflect the position they are in, More is the closeness of the different mean scores to each other. There very little difference between the mean scores and, when allocating students to a small number ( < 5%) had a large of one over others (as shown by a difference in raw scores of 50r
When students involved in this study were handed the LSQ they were also asked to to two concerning the inventory. and they might like to add about teaching and in general. The suml'\1ary of these responses in Scheme 1. The fact that many of the items the students difficulties. coupled with the realisation that the questionnaire was too long (it took the best part of a 45-minute lecture
1
the case of the LS I and Reflector using the LSQ. This is a confounding result as Kolb claims that the style or type is dependent on a student's choice of under- graduate major. Our finding is confirmed by the work of Mentowski and Strait (1983) who found that students' scores were independent of subject choices. They undertook a longitudinal study of undergraduate students at Alverno College. Milwaukee. One of the few measures which altered over the student's career was Kolb's LSI scores. Entering students all achieved high CE scores. which altered to AC by the time they graduated Third-year students in Accounting. Che- mistry and Sociology were surveyed using the adapted (40-item) Honey and Mumford question- naire in 1989 to see whether a similar shift occurred in our students (some of whom would have completed the inventories handed out in 1987). No such change was observed.
Our study of Kolb's LSI showed that it lacked
SAJHE(SATHO VOL. 7 NO.1 1993
construct validity and. because of its shortcomings. we have abandoned its use The LSQ avoids the use of an ipsative scale but although it might assist students in identifying their strengths and weaknesses as learners it is too management-bound to be of assistance in higher education. The inventory needs to be subjected to more stringent mathematical and statistical testing. It also suffers from a definitive link to a model of student learning If one has serious intentions to improve undergraduate teaching. researchers have to be convinced of the instrument's validity and reliability beyond question and to be able to relate the results directly to students' approaches to learning. We found that this connection failed to materialise in our attempts to validate the LSQ at university level. Because of this. it was decided to discontinue working with the LSQ and to abort our intention to conduct validity and reliability studies.
Table 1 Summary of learning styles as identified by Honey and Mumford
17
the case of the LS I and Reflector (^) the LSQ is a confounding result as Kolb claims that the style or type is dependent on a student's choice of under- major. Our finding is confirmed by the work of Mentowski and Strait (1 who found that students' scores were independent of They undertook a longitudinal study of students at Alverno College, Milwaukee, One of the few measures which altered over the student's career was Kolb's LSI scores. Entering students all achieved high CE scores, which altered to AC by the time they graduated Third-year students in Accounting, Che· mistry and Sociology were surveyed using the adapted (40-item) Honey and Mumford naire in 1989 to see whether a similar shift occurred in our students (some of whom would have the inventories handed out in 1 No such was observed.
Our study of Kolb's LSI showed that it lacked
SAJHE/SATHO VOL 7 NO.1 1993
construct validity and, because of its shortcomings, we have abandoned its use The LSQ avoids the use of an but although it might assist their strengths and weaknesses as learners it is too management - bou nd to be of education. The inventory needs '~j"'~ to more mathematical and It also suffers from a definitive link of student If one has serious intentions to undergraduate teaching, researchers have to be convinced of the instrument's and question and to be able to relate the results directly to students' to learning. We found that this connection failed to materialise in our attempts to validate the LSQ at university level. Because of this, it was decided to discontinue working with the LSQ and to abort our intention to conduct validity and studies.
Table 1 of as identified by Honey and Mumford
1
ISSN 1101-
Table 2 Correlation between Kolb's and Honey & Mumford's LSI
Table 3 Honey & Mumford's learning style questionnaire scores, mean scores for classes surveyed in 1987 and 1988 according to learning styles using the 80-item questionnaire
Table 4 Comparison of results obtained in this study with other studies
- Honey and Mumford (1982) 1302 people from a wide cross section of managerial/professional people working in UK industry. •• P. Lucas. 749 students at Australian Institute of Management. Murdoch University and West Australian College of Advanced Education (Mount Lawley) Av age 30,
••• This study,
18
ISSN 1101-
Honey & Mumford's
Table 2 Correlation between Kolb's and
Table 3
& Mumford's LSI
questionnaire scores, mean scores for classes surveyed in 1987 and 1988 to learning styles the SO-item questionnaire
Table 4 Comparison of results obtained in this with other studies
- Honey and Mumford (1982) 1302 people from a wide cross section of managerial/professional people working in UK industry. •• P. Lucas. 749 students at Australian Institute of Management. Murdoch University and West Australian College of Advanced Education (Mount Lawley) Av age 30.
••• This study.
1
Table 5
Honey ,!!t Mumford's learning style questionnaires (LSO) results for classes surveyed in 1987 and 1988 according to learning style using the 40-item LSQ
Table 6 Likert style responses used in the LSQ
Table 7 Shorter Honey and Mumford LSQ 1989
20
!It Mumford's learning
Table 5 results for classes "'""""'''''' in 1987 and 1988 using the 40-item LSQ
Table 6 Likert responses used in the LSQ
Table 7 Shorter Honey and Mumford LSQ 1989
SAJHE/SATHO VOL. 7 NO.1 1993
Table 8 Shorter Honey & Mumford L.S.Q. - 3rd year students - 1989
Figure 1
Kolb's model of Experiential learning and the four main types of learners. After Kolb (1984)
Active
Experimentation
(AE)
Accommodator
Converger
Concrete
Experience
(CE)
Abstract
Conceptualization
(AC)
Diverger
Assimilator
Reflective
Observation
(RO)
21
SAJHE/SATHO VOL. 7 NO.1 1993
Table 8 Shorter Honey & Mumford LS.Q. - 3rd year students - 1989
Figure 1
Kolb's model of Experiential learning and the four main types of learners. After Kolb (1984)
Active
Experimentation
(AE)
Accommodator
Converger
Concrete
Experience
(CE)
Abstract
Conceptualization
(AC)
Diverger
Assimilator
Reflective
Observation
(RO)
SAJHE/SATHO VOL. 7 NO, 1 1993
PARLETI, M, & D, Hamilton (1972) Illuminative Evaluation: A New Approach to the Study of Innovatory Programs University of Edinburgh Centre for Research in the Educational Sciences, Occasional Paper 9, RICHARDSON, JTE, (1987), Research in Education and Cognitive Psychology (Richardson, JTL M,E, Eysenck & D, Warren Piper, Eds) Student Learning Milton Keynes: Open University Press, RICHARDSON, T.E, M,W, Eysenck & D, Warren Piper (Eds) (1987) Student Learning Milton Keynes: Open University Press, SCHMECK, R,R, (Ed) (1988) Learning Strategies and Learning Styles New York: Plenum Press,
23
VOL. 7 NO 1 1993
PARLETT, M. & D. Hamilton (1972) Illuminative Evaluation: A New Approach to the of I-'rrlnr::;,m~ University of Edinburgh Centre for Research in the Educational Sciences, Occasional 9. RICHARDSON, J.T.E. (1987). Research in Education and Psychology J.T.E, M.E. .-\is... "r·" & D. Warren Piper, Student RICHARDSON, T.E, MW. & D. Press. SCHMECK, R.R. (1 New York: Plenum Press.
