Higher Education 24:291-316, 1992. 9 1992 Kluwer Academic Publishers. Printed in the Netherlands.
Impressions of disadvantage: I - school versus university study orchestration and consequences for academic support J.H.F. MEYER, T.T. DUNNE & A.R. SASS Teaching Methods Unit, University of Cape Town, Rondeosch 7700, Republic of South Africa Abstract. This study investigates the study orchestrations of engineering students who enter three universities from disadvantaged school backgrounds and are admitted to academic support programmes. The first part of this study examines the characteristics of the entry group as a whole and, on the basis of an analysis of the self-reported study orchestrations of the individuals involved, it is concluded that a significant subgroup of individuals enter university with manifestations of undesirable study behaviour that has serious consequences for academic support. The second part of this study examines the relationship between school study orchestration (as manifested on entry to university), subsequent study orchestration at two stages during the first year of study, and final outcome as evidenced in the end of year examination result. It is concluded that study orchestration is a relatively stable phenomenon for most individuals in the transition from school to university and that its early recognition can be a potential indicator of subsequent academic achievement depending on the nature of the assessment procedures employed in academic support programmes. The third part of this study investigates sources of variation in two subgroups of individuals whose study orchestrations change during the course of their first year. It is concluded that statistically significant, but different, sources of variation are associated with subgroups of individuals whose study orchestrations are either improving or deteriorating. The overall conclusions of this study are seen to be far reaching in terms of informing the selection procedures and the educational practice of academic support programmes as well as of undergraduate education in general.
Introduction The prediction of academic success or failure amongst entering first year university students has exercised the minds of educational researchers for decades. In an extensive review and synthesis of productivity research, Fraser, et al. (1987) discuss a generalised 'productivity model' that essentially consists of nine factors held to be 'powerful and consistent' predictors of learning outcomes. The three major direct 'causes' of learning that encompass the nine factors of the model are considered to be aptitude (ability, development and motivation), instruction (quality and quantity) and environment (home, classroom, peer group and mass media). The extensive synthesis of productivity studies carried out by Fraser, et al. (1987) and its interpretation in terms of the nine factor model, was intended to provide some guidance for improving the educational productivity of schools in so far as it identified factors influencing student learning. Literally thousands of studies attest to the extensive search for variables that can be used to predict learning outcomes. There remains, however, a pressing need to explore new approaches to the problem of educational productivity as new insights
292 into student learning become available. In the United States of America, for example, the Educational Testing Service has committed itself to an ambitious programme of research and development in order to produce a new generation of measurement techniques intended to improve the quality of educational assessment. Viewed against the perception that the overall academic achievement of students in the United States is deteriorating, there is a commitment to seek new educational measures that will reflect and promote the thinking skills of students in a wide variety of contexts (Educational Testing Service, Annual Report 1990). A further emphasis in seeking new measures of student learning is to improve the quality of teaching and of learning. To this, and other ends a wide ranging study of the American education research and development enterprise is being undertaken in order to harness the results of relevant research more effectively (Cross 1990). There is obviously not just a qualitative concern about what happens in schools the transition to university and the selection measures used to effect this transition are of equal importance. The quality of student learning therefore, and its likely consequences, is increasingly arousing the interest and concern of educational researchers from within a variety of research perspectives. In contrast to work carried out in the United States, research emanating from Sweden, the United Kingdom, Belgium, Holland, Australia, Hong Kong and South Africa has led to important conclusions that have persuasively and consistently linked learning outcomes to fundamental individual qualitative differences in the manner in which students perceive the content and context of learning, in the manner in which they engage learning tasks, and the manner in which they conceive learning itself. Much of this research has focussed on the learning behaviour of students in higher education, but it has also been demonstrated quite clearly that the same distinctive qualitative differences manifest themselves among secondary school pupils (Entwistle and Koz6ki 1985; Ramsden, Martin and Bowden 1989). On the increasingly popular subject of 'quality' then, it is somewhat surprising to observe that a significant and growing body of (non-American) research on student learning in secondary and higher education is not represented in the productivity model mentioned earlier. It is true that the model espoused by Fraser, et al. (1987) recognises some of the variables that contribute to individually distinctive study behaviour (notably, motivation and some general contextual factors) but there is, essentially, no recognition of the existence of qualitative differences in the manner in which learning tasks are engaged, or the concomitant conceptions of learning itself that are associated with such differences. It needs to be asked, therefore, whether research on student learning, and its emphasis on individual qualitative differences, can inform research on educational productivity. Contemporary research on student learning provides considerable impetus for asking such a question in so far as it has emphasised the important linkages between context, approach, and outcome among university students. At face value it seems reasonable to argue that a scholar, who adopts a theoretically desirable approach to studying in senior secondary school subjects, should do relatively well in such subjects at school, and in similar first year university
293 subjects. However, a scholar who similarly enters university with a history of theoretically undesirable approaches to studying should do relatively poorly. The implications of the above argument impinge directly on the question of whether the identification of academic potential and the prediction of academic success can be informed by a knowledge, in qualitative terms, of a student's study behaviour. This question is not only relevant to studies of educational productivity; it is of fundamental importance to some of the educational problems of many developing countries where a student place in higher education is an expensive and scarce resource that needs to be managed with great care. In order to answer this question, however, it needs to be recognised that there are a number of important assumptions in the argument that proposes it.
Clarification of terms
Before addressing these assumptions it needs to be recognised that the terms that have been used by various researchers to describe and qualify individual qualitative differences have evolved and become progressively more complex in terms of their conceptual interpretation. In order to avoid confusion, it is therefore necessary to locate some of these terms in the context of the present study. At the very simplest level, earlier studies essentially distinguished between qualitatively different forms of intention only, or congruent combinations of motive and intention (termed 'approaches') while later studies used the term 'orientations' to distinguish between more complex qualitatively different combinations of intention, motivation, learning styles, learning style pathologies and other additional study constructs. More recently, the term 'study orchestration' has been introduced to capture contextualised approaches to studying which have been shown to represent powerful sources of variation in the quality of study behaviour at an individual, and at a group level (Meyer 1991; Meyer and Dunne 1991). The concept of study orchestration thus captures individual (and group) study behaviour as reflected in a coalescence of constructs that cover both study approaches and perceptions of the learning context as manifested in a subject-specific response. The emphasis on contextualised study approaches is thus twofold; namely, perceptions formed about the context of learning, and the subject-specific nature of learning engagement. The one emphasis acknowledges the influence of the learning environment, and the other the influence of the subject being studied, in differential manifestations of learning behaviour (Meyer 1991).
Three assumptions related to study orchestrations
An exploration of the value of scholastic learning behaviour as an indicator of academic potential requires three basic assumptions. First, it has been emphasised that qualitatively different (for example, 'deep' and 'surface') approaches to
294 studying cannot be decontextualised; at an individual level it has been demonstrated that study orchestrations capture individual differences that are associated with learning outcome (Meyer 1991; Meyer, Parsons and Dunne 1990a). Thus, any argument that study orchestration might be transferable from school to university assumes relatively enduring contextual and other influences within the concept of study orchestration. The second assumption concerns the stability of an individual's study orchestration - in other words the extent to which individual study orchestration remains essentially stable during a course of study (that might constitute a relatively stable context) and might, for practical purposes, therefore be considered as a default behaviour of the individual concerned. Ramsden (1988) has marshalled what he regards as the 'persuasive evidence' and the conclusions from a wide variety of studies on student learning and argued the case supporting the general conclusion that study approaches are relatively enduring over time and across tasks. Theoretically, enduring patterns of study behaviour may be attributable, for example, to enduring motivational influences, preferred learning styles or habitual ways of engaging learning tasks. However, the most basic understanding of student learning engagement acknowledges sources of variation attributable to the context of learning as well, and this is held, by definition, to be a fundamental phenomenon for which there is also an impressive array of supporting argument (Entwistle and Ramsden 1983). Empirically, the evidence in support of this phenomenon is equally clear and its manifestation (within the concept of study orchestration) has been demonstrated at an individual, as well as a group, level of analysis (Meyer and Muller 1990; Meyer and Dunne 1991). There is, nonetheless, something of a paradox in asserting that approaches to learning can be both consistent and variable. The resolution of this apparent conflict appeals to a number of conceptual distinctions as to which aspects contributing to the overall study behaviour of students may be relatively variable or consistent. Thus, for example, it has been argued that variation in contextual perceptions that occurs within the same treatment group may be a function of previous experience, while characteristic modes of thinking, remembering and problem solving may be viewed as stable individual attributes (Ramsden 1988). It is the view of the present authors that further empirical evidence is required to address the question of stability more convincingly. The issue of sources of variation in individual study behaviour, in particular, has not yet been fully explored and such an analysis forms Part 3 of the present study. The third assumption concerns the transferability (within individuals) of study orchestration across, rather than within, any academic subject being studied. The ramifications of this assumption are more subtle than in the previous two assumptions in so far as the perceptions formed about the basic character of an academic discipline (for example, Mathematics or English) may shift in the transition from school to university. That is, notwithstanding the perception that the 'same' subject is being studied - even perhaps in apparently similar contexts - the mode of engaging it, and the perceptions that a student forms about it, may differ between school and university. In cases where there is no such apparent direct
295 semantic connection between subject content at school and at university, a further related and equally important assumption would be that study orchestration in, say, a global discipline called 'science' at school would be transferable to a cognate 'purer' speciality of the same global discipline (such as Physics, Chemistry or Applied Mathematics) at university. Where no such cognate analogues are perceived to exist, or where the analogues are more marginal (for example, Psychology, Librarianship, Social Work, Agriculture) any assumption regarding the transferability of study orchestration across subjects in the transition from school to university must represent, in theory at least, an act of faith in the light of evidence presented in two recent studies that individuals may manifest differential forms of study behaviour in different subject-specific contexts (Meyer and Watson 1991; Eley (1991)).
Background to the present study It is not known to what extent, or under what circumstances, the above assumptions may be generally valid. An earlier study carried out on academic support students in engineering investigated the stability of individual study orchestrations over time during a first year course in Applied Mathematics (Meyer, Parsons and Dunne 1990b). Some of the conclusions of this study are directly related to a number of the assumptions outlined in the previous section and were both surprising and unexpected. It had been anticipated that the study orchestrations of many individual students (especially those considered to be theoretically less conducive to efficient learning) would change during the first year course - primarily because of the influence of an academic support context implicitly intended to effect improvements in study behaviour. However, it was found that, while there was indeed some improvement (and some deterioration) in study orchestration amongst a minority of students, the majority of individuals manifested essentially stable patterns of subject-specific study orchestration that were associated with learning outcome in the manner expected. More specifically, it was concluded that students manifesting theoretically undesirable study orchestrations early on in the course were at risk of failing or performing poorly. This, in turn, raises the central question that the present study seeks to address, namely, whether such students enter universitywith an immediate history of theoretically undesirable study orchestration in a cognate school subject and, additionally, whether students entering university with desirable orchestrations preserve those orchestrations. It is thus also possible to explore the question of whether individual study orchestration in a school subject (Science) can be used to identify academic potential (manifested in terms of subsequent achievement) in a first year university subject (Mechanics).
The present study Three sets of students from three different universities formed the subjects of the
296 present study. These (mostly African Black) students are all carefully selected and admitted to study Engineering at three of South Africa's universities in three Bridging Programmes that will be referred to as Programmes 'A', 'B' and 'C' respectively. University 'A' in this study can be explicitly identified as the University of Cape Town but, for ethical reasons, the remaining two universities will be referred to anonymously. The students admitted to these programmes come from all parts of South Africa, the majority having done their schooling under the Department of Education and Training which caters predominantly for African Black pupils. They have matriculation exemption with Mathematics and Physical Science at Higher Grade and have received a recommendation from their schools that they are above average students. They are selected in the first instance by sponsoring companies and then by the co-ordinators of the Bridging Programmes and must be able to provide sufficient evidence of suitable personal qualities under interview conditions. Some sponsors also require an above average score on a battery of cognitive and reasoning tests which they administer. The students who are finally selected are provided with full cost bursaries by the sponsoring companies who also set the condition that the students attend a Bridging Programme. From past experience sponsors believe that this is the only way that these students will be able to compete on an equal basis with peers who have been fortunate enough to receive a better and more consistent education. All three programmes have the same aim which is to develop the students' academic attributes and personal qualities to the extent that they will succeed in their degree courses. What generally distinguishes these programmes from the regular first-year programmes is the commitment on the part of the teaching staff (all of whom are especially appointed) to the personal growth of the students as individuals. Particular emphasis is given to small-group tutorial activities which also assist the students (for whom English is generally a second language) to develop important communication skills. This commitment is further manifested in a system of academic mentoring by which students are supportively encouraged to become independent learners and assume responsibility for their intellectual growth. A mentoring system also operates during the vacations when students are placed in selected engineering industries and placed in the care of an appointed mentor. In addition, the programmes provide enrichment experiences and opportunities that aim to further motivate the students' interest in, and awareness of, engineering as a career. All three programmes thus provide a very supportive learning environment at a personal level. They are, however, dissimilar in the manner in which they relate their academic activities to their respective undergraduate engineering curricula. Programmes 'B' and 'C' are both based on one-year true bridging programmes; they offer a variety of non-credit introductory courses, the successful completion of which enables students to proceed to the regular first year engineering programme. Programme 'A', by contrast, is a reduced curriculum programme in which the credit-bearing courses of the first two years of the regular programme are spread over three years. In the first year the focus is on those areas which are deemed
297 critical indicators of future success in the study of engineering, namely, Mathematics, Applied Mathematics (or Mechanics) and Technical Communication (which includes the credit course of Engineering Drawing). Intensive support is given to students during the first year, while additional tutorial support on a reducing scale is offered in the second and third years.
General methodology The investigation of the manifestation of study orchestrations among students entering, and during, an academic support programme is a primary focus of the present study. The general methodology employed to solicit and interpret such manifestations at an individual level has been explained in some detail by Meyer (1991) and is not repeated here. In essence, the methodology used in the present study consisted of three administrations of a contextualised approach to studying inventory (see Appendix for a description of the variables that were employed and the symbols used to refer to them) to students admitted to the three programmes described earlier. These measures were used as the basis for establishing individual differences in entry study behaviour to the programmes as well as for investigating the subsequent stability of such behaviour during the course of the programmes. The academic year commences in February and the first administration, which was done prior to the commencement of the Programmes, was aimed at soliciting the manifestation of study orchestrations in Science at school. For practical reasons, this task could not be accomplished before the students arrived at the university campuses. Students were thus asked immediately on arrival to respond retrospectively to items contained in the inventory and to focus their responses very firmly in terms of studying Science during their final year at high school. These retrospective data are referred to further on as 'School' orchestrations. Data could not be obtained from the relatively few late arrivals at the three Universities and such individuals are excluded from the analyses presented. The inventory was further administered to the three groups of students in April and October. On these occasions students were asked to focus their responses in terms of studying Mechanics in their respective Programmes. These data are referred to further on as the 'April' and 'October' orchestrations respectively. For each individual, the data obtained from each administration of the inventory were categorised based on a synthesis of research findings on student learning applicable at an individual level. The initial conceptual basis for this categorisation has been described in Meyer, Parsons and Dunne (1990a) and subsequent refinements to it have been reported in Meyer (1991). In the present study, the contextualised study approach of each individual at the three points in time (on entry from school, and in April and October) was assigned to one of five categories intended to reflect conceptually distinct qualitative differences: the most theoretically desirable category, designated 'AA' (above average) captures a range of individual orchestrations that are conceptually faultless and which would be expected to be positively associated with learning outcomes that reflect the
298 acquisition of understanding. At the opposite extreme, the theoretically most undesirable category 'AR' (at risk) captures a range of orchestrations that predominantly includes disintegrated orchestrations, that is, those that are conceptually uninterpretable and which have been associated with academic failure or low-achievement. The third category, designated 'AV+' (better than average) is typically characterised by the intrusion of non-pathological influences into an otherwise theoretically faultless orchestration. In contrast, the category 'AV-' (worse than average) is characterised by orchestrations that are significantly pathological but which still retain some remnants of structure in terms of the meaning orchestration constructs. The fifth category, designated 'AV' (average) captures a range of orchestrations that are not amenable to a neat conceptual encapsulation but which fall between the characteristics of 'AV+' and 'AV-'. Typical examples of individual study orchestrations within each of these five categories can be found in Meyer (1991) and are not repeated here. When individuals with similar categorisations are combined into subgroups (and analysed statistically as is done further on), then the characteristics of such subgroups are, once again, intended to broadly reflect the same conceptually distinct qualitative differences and the extent to which they do, in fact, do so (and have done so in previous studies) may be taken as an empirical verification of the classification procedure itself. Since the analysis and conclusions of the present study are tied to the employment of this categorisation procedure, its reliability deserves further comment. In general, it needs to be emphasised that the broad conceptual basis for assigning individual study orchestrations to categories originally emanates from a synthesis of baseline studies on student learning that reached conclusions applicable at an individual level. It has been demonstrated, furthermore, that the association between such categorisations and learning outcome is consistent with the underlying model of student learning within which the baseline studies were carried out (Meyer, Parsons and Dunne 1990a, 1990b). The interpretation of a particular study orchestration is clearly potentially sensitive to subjective differences between independent judges although in practice it has been found that such differences are minimal - they typically arise in less than five percent of cases and usually represent borderline cases that fall between categories. Prior knowledge of an individual student can clearly also influence the objectivity of the process and for this reason all the classifications (performed by the first author) in the present study were done on anonymously presented data. In terms of further development, it is clearly important to determine the extent to which the relatively complex interpretation procedures and the application of associated conceptual 'rules' can be formalised in terms of programmable logic, or by the fitting of appropriate statistical models. Such investigations are presently underway and will be reported in due course. Presentation of results
In the analyses that follow, the fullest set of data appropriate to each analysis is
299 presented. While every attempt was made to obtain full longitudinal data for every individual, it will be observed that there are some minor fluctuations (due to missing data) in the sample sizes of the various data sets reported. The various analyses and their results are presented in three parts. Part 1 deals with an analysis of entry (School) study orchestration. The result of an unfolding analysis performed on the pooled group of students from all three programmes is first discussed and is followed by similar analyses of two subgroups conceptually constructed from the pooled sample on the basis of the individual study orchestration categories described earlier; one subgroup comprising individuals manifesting theoretically desirable and conceptually coherent study orchestrations (categories 'AA' and 'AV+'), and a second subgroup comprising individuals manifesting theoretically undesirable and conceptually incoherent study orchestrations (category 'AR'). The two resultant subgroup study orchestration characteristics are then compared in order to illustrate the inherent danger in drawing conclusions based on aggregated data that is assumed to be individually representative. The employment of unfolding analysis is relatively new to educational research of this nature. It has proved to be a versatile statistical procedure for exploring nonlinear data structures and it is, therefore, independent of any correlational assumptions. Its application does not depend on large sample sizes and it is, furthermore, statistically based on an individual difference model that does full justice to the existence and preservation of individual differences in any data structure manifestation. This means that any resultant data structure seeks to represent individuals singly and collectively in relation to any given set of observed responses to a set of variables. It equally well jointly represents associations between variables and, in several reported studies where such variables are study approach/contextual perception variables, it has been concluded that their associated empirical data structures bear a close conceptual resemblance to empirical factor structures associated with similar sets of variables. In one study (Entwistle, Meyer and Tait 1991) such a comparison is made that is directly based on the application of the two quite dissimilar statistical procedures to the same set of data. A basic difference between the two procedures is that in an unfolding analysis any association between variables is not manifested in terms of correlation-based coefficients, but by the proximity of the variables relative to each other in a kdimensional space. Thus, for example, associated variables (and/or individuals) will be characterised by relatively small inter-point distances and will typically form discernible clusters in the space that can be conceptually interpreted in a manner that is, once again, analagous to the extraction and interpretation of factor structures. (In an unfolding analysis all inter-point distances are uniformly interpretable irrespective of whether the points represent individuals or variables). In the present study, data structures are represented in a three-dimensional space - a choice that reflects the basic dimensionality of the model of student learning within which the data are obtained and for which there is also empirical support from other studies. For ease of graphical simplicity points representing individuals are not
300 plotted in this space in the present study but this does not affect the interpretation of the data structures involving the points that represent the variables. A fuller treatment of this procedure and further discussion of the analogies referred to may be found in Meyer and Muller (1990), Meyer (1991) and Meyer and Watson (1991). In Part 2, it is the categorisations of individual study orchestration that are used as the basis of the various statistical treatments employed to explore the existence of measures of association between these categorisations and leaming outcome. In order to do this, a categorisation of the available outcome measure (the raw final end of year mark) was constructed corresponding to <40%, 41%-50%, 51%-60%, and >60%. These categories were subsequently reduced to 'Fail' (<50%) and 'Pass' (~>50%), within programmes, in order to eliminate statistical artefacts that might exaggerate any associations. In Part 3, an exploratory analysis is carried out to determine statistically significant sources of variation in two subgroups of students whose study orchestration categories changed within the programmes. This analysis is based on the mean scores of the subscales imbedded within the inventory.
1. The entry School study orchestration for the group as a whole The pooled group of students (n = 154) entering the programmes constitute a viable sample that can be used to explore some of the presumed attributes of group identity. The assumptions on which such an exploration rest appear to be reasonable. By far the majority of the students are African Blacks, have been exposed to similar forms of inferior schooling, and have obtained their matriculation exemptions from the same educational authority. As a group they have also been selected by very similar processes to enter university - in the majority of cases by the same sponsoring bodies. Recent studies of student learning (using both the group and the individual, as a unit of analysis) have cautioned against the not uncommon practice of failing to discriminate between passing and failing students (Meyer, Parsons and Dunne 1990a; Entwistle, Meyer and Tait 1991; Meyer and Dunne 1991). Such discrimination is, by definition, not possible in the present study as all the students passed - and arguably passed the school examinations well, given their background circumstances and the selection criteria used to admit them to university study. They constitute, in comparison with their less successful school peers, a relatively small minority of students of the highest calibre. Given the fact, then, that all of the students have, at face value, achieved relatively well and/or been selected with great care to enter university, there is an understandable prevailing assumption that the end result of the selection process is therefore a group of individuals who are essentially homogeneous in terms of both disadvantaged school background and demonstrable potential. Academic support programmes, in fact, tacitly assume the existence of similarity within the groups entering their programmes and typically espouse the importance of meeting the
301 'real learning needs' of such students based on a presumed group identity. The first unfolding analysis that was carried out (not presented) was for the pooled entry group as a whole. In essence, the solution represented the group study orchestration in terms of the underlying structure of the data which, in this case, was quite seductive in appearance. At face value, this solution, although slightly conceptually disjointed in some respects, would not constitute a basis for expressing concern about the group as a whole given its disadvantaged educational background. There was evidence in this solution of an association between study approach constructs that conceptually constitute the basis of a meaning orchestration (in the form of a deep approach, intrinsic motivation, relating ideas and use of evidence). This combination of intention, motive and process was, however, not as well contextualised as it might have been in terms of the deep contextual perceptions with which it is normally associated. In similar vein, there was evidence of what could conceptually be interpreted as an incomplete strategic orchestration that was well contextualised (a strategic approach coupled with deep perceptions of methods of assessment, books and of human relationships as well as surface perceptions of course content), although achievement motivation with which such an orchestration is normally associated, was, in this case, relatively isolated. Evidence in support of a reproducing orchestration was primarily in terms of conceptually coherent, but weaker, linkages between the pathology of improvidence, a fragmented approach, a memorising approach and, perhaps, external motivation. A similar linkage associated the pathology of globetrotting with a perceived heavy workload. It could be inferred from such an analysis that the data structures represent the basis of individual differences and, that since the evidence supporting meaning/strategic structures was stronger than for reproducing ones, the pooled group as a whole has been well selected. Conclusions such as this one can be very misleading as will now be demonstrated. The solution presented in Figure 1 is based on a subgroup of individuals whose School orchestrations were categorised 'AA' or 'AV+'. In conceptual terms, this solution is almost faultless and it compares favourably with a theoretically ideal model of how a group of high achievers might manifest a patently successful school orchestration; the central cluster of 'hearts' defines an exceptionally robust meaning orchestration in terms of intrinsic motivation (IM), deep approach (DA), use of evidence (UE) relating ideas (RI), reflection (RE), operation learning (O1) deep perceptions of books (BD), of methods of assessment (AD) and of human relationships (RD). The influence of comprehension learning (CL), while clearly present, does not appear to be as strong as that of operation learning (O1) - an observation that has been made in other studies involving Science students. Furthermore, achievement motivation (Am), as well as a strategic approach (St), do not appear to be more than a positive or supportive influence (as opposed to forming the basis of a clearly defined strategic orchestration). The surface and deep perceptions of learning space (Is, LD) are associated with one another (but
302
Z 2.15
1.05
't St g.L
l
/.l
eW
-0.06
-1.16t 2.96
".88
Y
-2.80
Fig. 1. Joint space representing AA/AV+ sub-group (n=84); overall stress=0.135
303 Legend: Hearts
DA, deep approach; IM, intrinsic motivation; RI, relating ideas; UE, use of evidence; CL, comprehension learning; RE, reflection; BD, books (deep); AD, methods of assessment (deep); LD learning space (deep); RD, relationships (deep); St, strategic approach; O1, operation learning; Am, achievement motivation. Pyramids ma, memorising approach; fa, fragmented approach; sb, syllabus boundness; ff, fear of failure; ip, improvidence; cs, course content (surface); ls, learning space (surface); rs, relationships (surface); wl, workload; ds, disorganised study methods; gL, globetrotting; eM, extrinsic motivation. (See Appendix for explanation of above terms).
somewhat isolated from the central cluster). This is unusual, but may be an indication of the generally poorly perceived facilities found in the majority of African Black schools. There is a suggestion, too, of some influence on the meaning orchestration in the form of a memorising approach (ma). Again, this is not unexpected given the widely accepted view that the schools in question generally place undue emphasis on rote learning to pass examinations. The remainder of the constructs are dispersed in the space and, apart from the linkage between workload (wl) and the pathology of globetrotting (gL) that was also observed to exist in the initial solution, they represent little of further interest. If all successful entering students manifested an orchestration similar to the one in Figure 1 then there would, indeed, be a basis for assuming that the careful selection process had facilitated a recognisable (and desirable) form of group identity in terms of study behaviour. This, however, is unfortunately not the case. The unfolding solution presented in Figure 2 is based on the subgroup of individuals whose study orchestrations were categorised as 'at risk' ('AR'). This solution is fundamentally and disturbingly different from the one presented in Figure 1. There is no conceptually recognisable meaning orchestration. The 'hearts' are dispersed in the space while the 'pyramids' occur in conceptually clear associations that are disturbing in so far as they represent a characteristic of the learning behaviour for the subgroup as a whole. The association, for example, which is evident in the lower left hand side of Figure 2 represents a clear linkage between fear of failure (ff), a memorising approach (ma), perceptions of a heavy workload (wl), the pathology of globetrotfing (gL), disorganised study methods (ds) and, further up, extrinsic motivation (eM) and the pathology of improvidence (ip). On the fight hand side of Figure 2, the deep approach construct (DA) is isolated from all the supporting processes and motivational influences that are collectively considered to be essential for the acquisition of understanding. The only redeeming feature (if it can be called that) is the linkage between the (process of) use of evidence (UE), achievement motivation (Am), and operation learning (O1) that is weakly contextualised and somehow apparently associated with syllabus boundness (sb). It is dangerous to attempt to read too much into a solution such as the one
304
Z 0.99
o,t!
St
.k -0.06
-1.12
-2.17 3.41
14
2.18 0.95 -0.28 Fig. 2, Joint space representing AR sub-group (n=22); overall stress=0.211
305 Legend: Hearts
DA, deep approach; IM, intrinsic motivation; RI, relating ideas; UE, use of evidence; CL, comprehension learning; RE, reflection; BD, books (deep); AD, methods of assessment (deep); LD, learning space (deep); RD, relationships (deep); St, strategic approach; O1, operation learning; Am, achievement motivation. Pyramids ma, memorising approach; fa, fragmented approach; sb, syllabus boundness; if, fear of failure; ip, improvidence; cs, course content (surface); ls, learning space (surface); rs, relationships (surface); wl, workload; ds, disorganised study methods; gL, globetrotting; eM, extrinsic motivation. (See Appendix for explanation of above terms).
presented in Figure 2 because of the diversity of non-uniform orchestrations that can still occur within the 'AR' category. However, insofar as the solution presented for interpretation is an optimal one for this particular subgroup, there can be little doubt that it is attempting to capture a representation of study behaviour that, quite apart from its unexpected existence in a select group of otherwise academically successful students, is cause for serious concern in its own right as other studies have demonstrated (Meyer and Muller 1990; Meyer, Parsons and Dunne 1990a, 1990b). It should be noted that it was not the intention of the foregoing analysis to invoke a self-fulfilling argument about the subgroups. The point is that the data were used to conceptually classify individual students into subgroups whose overall characteristics, in terms of the same set of variables that informed the classification of the individual study orchestrations, are then compared by means of Figures 1 and 2. The foregoing analysis could equally well have proceeded from a purely empirical perspective (not presented) using a methodology described in detail in Meyer and Muller (1990). In this case the same conclusions would have been reached in respect of the subgroup differences since a majority of individuals in the 'at risk' subgroup would have been identified. However, the classification procedure was invoked to ensure that the selection of the 'at risk' subgroup was effected from a conceptual basis, rather than a purely empirical one. The concern about Figure 2 and its consequences for the students in that subgroup are thus based on the fact that, although chosen from a supposedly homogeneous group of pre-identified students with potential to perform well academically, they collectively display the alarming study orchestration depicted. There is a robust conceptual, as well as a firmly established empirical, basis for this disturbing manifestation. Its precise interpretation in this case is not as important as the recognition of the fact that it patently exists and that it can apparently coexist, unnoticed, in analyses based on larger aggregations of individuals. There is no basis for arguing that the individuals represented in Figure 1, in comparison with the individuals represented in Figure 2, are in any way similar. It is equally clear that the initial analysis carried out on the pooled group does not inform the existence of such contrasting and individually representative data structures.
306 There is an understandable belief among academic support practitioners in South Africa that 'the real learning needs' of disadvantaged students can all be met in essentially similar ways. It is the view of the present authors that the question of how to meet such 'real learning needs' in the case of individuals considered to be manifesting extreme 'at risk' orchestrations may, in fact, be conceptually meaningless. What, for example, in a university education context, might such needs be in the case of an individual manifesting, and confirmed to be manifesting (by means of interview data), a disintegrated, conceptually incoherent study orchestration and who has, at best, an unsophisticated conception of what 'learning' is? This is a provocative question but it needs to be asked. It is obvious from the foregoing analysis that students manifesting theoretically disturbing study orchestrations are able to gain access to university in spite of carefully administered selection procedures. Given the fact that such students can be identified on or even prior to admission, it is not nearly as obvious what can, or should, be done as a response to the problems that these students undoubtedly encounter and create for their sponsors as well as their host institutions. A range of possibilities suggest themselves and these include doing nothing (passive acceptance of the status quo), prevention as part of the selection process, exclusion in the absence of any improvement within the programme and intervention through specific treatment. There is an urgent need for further research to inform any such choice and the polemic debate that will no doubt surround it.
Associations between school orchestration and outcome
In order to examine the consistency of study orchestrations over time, a series of 5• contingency tables were constructed in which rows corresponded to categories of school orchestration and columns to April and October orchestrations respectively. These contingency tables (not presented) indicated a general consistency of individual orchestrations over time. In order to demonstrate statistically the consistencies within the contingency tables, the Gamma statistic (Goodman and Kruskal 1979) was employed. Additionally, the D (C:R) statistic (Somers 1962) was employed on the assumption that row categories are predictors of column categories. In either case, these1 statistics may be interpreted as the differences between probabilities of concordant and of dissonant pairs of orchestrations for two randomly chosen individuals. Absolute consistency yields a value of +1 for either statistic and inverse consistency yields a value of -1. Table 1 presents these two statistics and the corresponding values of the t-statistic associated with the null hypothesis test of no consistency. It can be seen from Table 1 that, for each of the three academic support programmes, individual study orchestration categories are remarkably stable. While it has been established from previous studies that orchestrations are relatively enduring within disciplines over time, it is surprising that the orchestration categories remain so stable over the school/university transition when it is an expectation of the academic support programmes that improvement will occur,
307 especially in the 'below average' and 'at risk' categories. The fact that this pattern occurs in all three of the academic support programmes studied suggests that, despite any perceived contextual or discipline differences in the school/university transition, the apparent general stability of the orchestrations is not an anomalous feature. Table 1, Associations between school orchestration a n d support p r o g r a m m e orchestrations in April a n d O c t o b e r P r o g r a m m e ' A ' (n = 37) School/April
School/October
G a m m a = 0.634
G a m m a = 0.674
D(A:S) = 0.527 t = 4.779
D(O:S) = 0.533 t = 5.084 P r o g r a m m e ' B ' (n = 36)
School/April
School/October
Gamma = 0.640 D(A:S) = 0 . 4 9 6 t = 3.872
G a m m a = 0.885 D(O:S) = 0.695 t = 10.171 P r o g r a m m e ' C ' (n = 44)
School/April
School/October
G a m m a = 0,521 D(A:S) = 0,405
G a m m a = 0.571 D(O:S) = 0,444
t = 3.811
t =4.195
It must be concluded that, on the one hand, either the orchestration itself is stable and impervious to external influences or that the categorisation process is insufficiently sensitive to shifts in orchestrations that actually occur, and that this process therefore imposes a spurious impression of enduring study behaviour, or, on the other hand, that the presumed inherent objectives of the programmes are not being accomplished. In view of previous research findings the authors are inclined towards the last alternative. The question then arises whether the apparently stable orchestrations are, as they are theoretically expected to be, associated with academic success within the programmes. Previous research informs an expectation that, in so far as the outcome measures reflect the acquisition of understanding rather than simply a quantitative increase in knowledge, such an association will be present. Stated differently, the measure of learning outcome (in this case, final examination results) should discriminate between students on the basis of qualitatively different study orchestration categories. It can be seen from Table 2a that, for Programme 'A', there is evidence of mild association between the school orchestration categorisations and the final end-ofyear examination results. This reassuring association is, as expected, more evident
308 Table 2a. Associations between orchestration a n d learning o u t c o m e (excluding dropouts) for p r o g r a m m e ' A ' n=38 School Full categories
Reduced categories
G a m m a = 0.134 D(E:S) = 0 . 1 1 0 t = 0.874
n=37 April Gamma = 0.089 D(E:A) = 0 . 0 6 6 t = 0.576
n=37 October G a m m a = 0.226 D(E:O) = 0 . 2 0 2 t = 1.632
School
April
October
G a m m a = 0.272 D(E:S) = 0.138 t = 0.993
G a m m a = 0.298 D(E:A) = 0 . 1 5 2 t = 1.142
Gamma = 0.479 D(E:O) = 0.251 t = 1.993
Table 2b. Associations between orchestration and learning o u t c o m e (excluding dropouts) for p r o g r a m m e ' B '
Reduced categories
n=36 School
n=36 April
n=36 October
Gamma = 0.390 D(E:S) = 0.191 t = 1.412
G a m m a = 0.544 D(E:A) = 0.275 t = 1.867
Gamma = 0.586 D(E:O) = 0 . 2 9 0 t = 2.280
Table 2c. Associations between orchestration a n d learning o u t c o m e (excluding dropouts) for p r o g r a m m e ' C '
Reduced categories
n=44 School
n=44 April
n=44 October
G a m m a = 0.063 D(E:S) = 0 . 0 2 2 t = 0.199
G a m m a = 0.184 D(E:A) = 0.065 t = 0.626
G a m m a = 0.167 D(E:O) = 0 . 0 6 0 t = 0.563
for the October orchestration categories. When the original subgroups are collapsed to form 'AA/AV+', 'AV', 'AV-/AR' and fail/pass groups, the pattern of association between categorisation and outcome is found to be sharper and to steadily improve through to October. Stemming from this observation, it may be inferred that the discrimination of risk categories from study orchestrations is more effective at grosser levels ('AA/AV+', 'AV', 'AV-/AR') than at the more nuanced levels ('AA', 'AV+', 'AV', 'AV-', 'AR')for groups of this size, but there is, as yet, insufficient statistical evidence to support this view fully. The data presented in Table 2a, which are based on observations for only those individuals who eventually wrote the final examination, although not statistically significant, are consistent with a priori expectations based on the conclusions of previous studies in a variety of educational settings and, in particular, with studies
309 based on an individual difference model. The same conclusions also hold true in respect of the data for Programme 'B' where a similar pattern in the data is evident (Table 2b), but not for Programme 'C' (Table 2c). Programme 'C' appears to be exhibiting a deviation from the expected patterns in that the association between orchestration and outcome becomes weaker after April. There is insufficient data to determine this change as being statistically significant, and it may well simply reflect an artefact of variation. Such a result is, however, in conflict with the previously stated expectation that the outcome measure should discriminate between students on the basis of qualitatively different orchestration categories. Tables 2a, 2b and 2c specifically address the issue of association between orchestration categories and learning outcome in those individuals who completed their corresponding programmes. When corresponding data for individuals who did not complete their corresponding programmes (that is, those individuals who left or dropped out of their programmes) are included in the analysis, the fact that only the School and April orchestrations are available for these individuals implies that only the School and April orchestration statistics are affected. The overall association between orchestration and outcome (with dropout now identified with failure) for Programmes 'A' and 'C' can now be compared (Table 3a and 3b respectively). Both programmes now exhibit similar measures of association for the School and April orchestrations, but not for the October orchestration which remains unchanged. It needs to be emphasised, then, that the inclusion of those individuals who did not succeed in completing their respective programmes fails to produce evidence to contradict the conclusions emanating from the analysis presented in Tables 2a, 2b and 2c. The authors are led to conclude, on the basis of the evidence available to them, Table 3a. Associations between orchestration a n d learning o u t c o m e (including dropouts) for p r o g r a m m e ' A '
Reduced categories
n=46 School
n=45 April
n=37 October
Gamma = 0.232 D(E:S) = 0.123 t =0.914
G a m m a = 0.407 D(E:A) = 0.205 t = 1.711
G a m m a = 0.479 D(E:O) = 0.251 t = 1.993
Table 3b. Associations b e t w e e n orchestration and learning o u t c o m e (including dropouts) for p r o g r a m m e ' C '
Reduced categories
n=57 School
n=57 April
n=44 October
G a m m a = 0.268 D(E:S) = 0.133 t = 1.175
G a m m a = 0.417 D(E:A) = 0.207 t = 2.010
G a m m a = 0.167 D(E:O) = 0 . 0 6 0 t = 0.563
310 that an apparent feature of Programme 'C' is to pass nearly all the students who have not dropped out by mid year. This conclusion is partially supported by the fact that the inclusion of intermediate marks (obtained during the course of the Programme) into an overall composite end of year mark (as is done in Programme 'C') diminishes the association between orchestration and outcome (based on the composite mark) from its previous levels. For Programmes 'A' and 'B', however, it can be concluded that School study orchestration does represent an indicator of 'potential'. Although this evidence is not statistically significant, it is sufficient to underline the need for future research into the patterns of admission and exclusion in academic support programmes. A failure of examinations to discriminate on the basis of qualitative differences in study behaviour may not, in fact, serve the best interests of any programme, its sponsors or the individuals in it. It would appear that caution needs to be exercised in using examination results to evaluate the success of first year academic support programmes. It should be noted, though, that there is nothing particularly remarkable in observing that university examinations can fail to identify students who do not fully grasp the meaning of basic concepts in the subject being examined. The folklore of higher education abounds with examples that testify to the more whimsical practices of assessment procedures. The thrust of the conclusions regarding the failure of assessment procedures to identify students who are manifesting conceptually incoherent study orchestrations is far more serious. Such students are, by definition, not marshalling and integrating the collective processes required for the acquisition of understanding. Any unsophisticated conceptions they may have concerning the nature of learning in higher education and its 'rewards' could, conceivably, be further reinforced by inappropriate assessment procedures rather than altered in the manner desired.
Sources of variation in study orchestrations By considering the categorisations of all the individual study orchestrations it is possible to separate out two further subgroups whose orchestrations either improved or deteriorated from April to October, that is, within the programmes themselves. The sizes of these two subgroups are too small (n=27 and n=26 respectively) to confidently allow a separate analysis to be carried out for each programme. However, based on the assumption that the general supportive atmosphere and its implicit influences within each of the three programmes is essentially similar, an exploratory analysis was attempted to determine the sources of variation in the orchestrations within these two changing subgroups. It was anticipated that such an exploratory analysis might also provide some empirical support for previously reported conclusions concerning sources of stability in study approaches such as enduring motivational states, preferred learning styles or habitual modes of task engagement as discussed earlier. For each of the two subgroups, the response differences (on each of the twenty five subscales contained in the Inventory) between April and October were used to
311 Table 4. Significant sources of variation between improving (I) and deteriorating (D) orchestration subgroups
Variable
Deep approach Use of evidence Strategic approach Books (deep) Course content (surface) Intrinsic motivation Globen'otting Improvidence Fragmented approach Memorising approach Disorganised study methods
t
p value
I-mean (s.d.) n=27
D-mean (s.d.) n=26
2.622 2.130 2.095 2.221 2.257 3.383 -3.897 -4.076 -3.602 -3.364
0.01 0.04 0.04 0.03 0.03 0.0014 0.0003 0.0002 0.0007 0.0015
0.074 0.44 0.250 0.77 0.389 0.57 0.200 0.48 0.210 0.63 0.306 0.60 -0.445 0.67 -0.407 0.87 -0.252 0.71 -0.526 0.92
-0.298 -0.144 0.029 -0.092 -0.128 -0.260 0.221 0.452 0.392 0.215
0.58 0.55 0.68 0.48 0.43 0.62 0.57 0.64 0.59 0.66
-4.332
0.0001
-0.393
0.569
0.74
0.87
Note: Relating ideas, achievement motivation and deep perceptions of methods of assessment have p values of 0.07, 0.07 and 0.08 respectively and are thus marginally close to being significant.
calculate mean scores. Hypothesis testing using the two-sample t-statistic was then performed in order to establish which of these mean changes were significantly different for the two subgroups. These t-tests incorporate adjustments for unequal variances where such adjustments are demanded by the data. Statistically significant results from this procedure are summarised in Table 4. Given that two subgroups of students who have been categorised as improving or deteriorating respectively are being examined, the variables which apparently distinguish between these two subgroups are given in Table 4. Viewed as a whole, the major sources of variation in the study orchestrations of the two sub-groups are readily apparent. Of particular interest is the inclusion of some contextual variables in Table 4 (and in its footnote) which underline, once again, their important influence on study behaviour. The apparently anomalous response to the course content (surface) perception subscale may simply reflect the fact that the improving subgroup is becoming more holistically aware of their learning environment (by adding surface perceptions to their already established deep perceptions) and that the deteriorating subgroup are becoming even more impoverished in terms of their contextual perceptions. The contrast between improving and deteriorating orchestration subgroups can also perhaps be interpreted as representing a growing commitment to the university environment as opposed to a decline in such a commitment coupled with a loss of interest and motivation. The conclusions of this exploratory analysis are thus of some theoretical interest in so far as they support the relative stability of some constructs but not of others. In terms of the motivational constructs, for example, intrinsic motivation appears to be one of several sources of significant variation while extrinsic motivation does not. This fact, together with the marginal status of achievement motivation (noted in the footnote of Table 4) calls into question any assumed 'enduring motivational
312 states' in individuals undergoing qualitative changes in study orchestration and is a potentially significant indicator of the implicit influences in academic support programmes that are intended to alter students' motivations. Similarly, a qualitative deterioration in study orchestration is significantly attributable to variation in pathological influences but not in their associated learning styles. It cannot be inferred from the data available that the significant shift in these pathologies and in the related shifts in fragmentation, memorisation and disorganised studying are merely an indication of a shift away from a meaning towards a reproducing orchestration. The manifestation of these shifts at an individual study orchestration level does not support such a simple uniform interpretation and requires further investigation which is presently underway. It needs to be established whether disintegrated orchestrations, in particular, are extreme variants of a reproducing study orchestration or whether they represent a conceptual entity in their own fight that appears, at present, to be absent in contemporary models of student learning.
Discussion The present study has highlighted fruitful areas for future productivity research, particularly in the context of academic support programmes. Much of the educational practice in these programmes is driven by commendable enthusiasm and is rooted historically in a 'study skills' intervention philosophy that is patently obsolescent in terms of modifying student approaches to learning (Entwistle 1992). Given the enormous importance of the task and the richness of the opportunities that it presents, it is surprising how little published research on student learning has emanated from academic support activity, notwithstanding the evolutionary intellectualisation of the debate that surrounds it and the frequent use of 'student learning' terminology in such debate. In South Africa, academic support programmes have grown considerably both in number and in size in recent years. As a growth industry, and an expensive one to maintain at that, it is imperative that it be informed, and be transformed, by research on student learning carried out within its own unique context and, perhaps, even more so by research carried out in schools. There is simply no point, for example, in maintaining a belief in the group identity or, indeed, in the group potential, of the educationally disadvantaged (that further serves as a basis for adopting particular forms of educational practice), when the evidence against such a uniform identity can so easily be illustrated. The conclusions of the present study thus impinge heavily on a number of prevailing assumptions underpinning the general educational practice of academic support programmes. There is, firstly, the inescapable conclusion that there is a significant proportion of highly selected students entering such programmes who are, in a very real sense, at a multiple disadvantage. They, unlike many of their peers, enter university not only from a disadvantaged school background, but with a presentation of study behaviour for which, at the moment, there appears to be provision for neither diagnosis nor treatment. Failure to acknowledge this fact,
313 however painful, is both irresponsible and pretentious. Many students manifesting study orchestrations categorised as being 'at risk' remain stable in terms of this categorisation, and the end result is, in theory at least, inevitable in terms of academic failure or low achievement. There is thus a serious need to identify such students and realistically engage the question of what can be done to assist them. If there are intervention mechanisms that are likely to benefit them then they need to be researched with vigour. A further conclusion is that the problem of identifying 'at risk' students in academic support programmes is not necessarily aided by the methods of assessment employed. There is not a sufficient volume of accumulated evidence from studies on student learning against which to question seriously the educational merits of assessment procedures which fail to differentiate between students manifesting f u n d a m e n t a l l y different forms of study behaviour. It can be argued that such a practice is not in the interests of any of the students involved. It must, finally, be observed that the findings of the present study emanate from a particular and important sector of undergraduate students. While some of the conclusions presented have particular implications for the enterprise of academic support, they also represent stimulating challenges in terms of general undergraduate education. They point to a need to recognise and respond to differential forms of entry study behaviour, to introduce new questions into undergraduate education that concern the manifestation of individual qualitative differences, their implications, and how to raise students' awareness and control of their own learning behaviour. The conclusions of the present study have formed the basis of a subsequent, follow-up study aimed at addressing these questions in the form of an intervention programme aimed specifically at students identified as being 'at risk' on arrival at university. The results and conclusions of this follow-up study will be reported as the second part of the present study.
Appendix
Study orchestration subscales and their meaning. (A sample item f r o m each subscale is given in italics).
1. Contextual perception subscales (See Note below) DEEP PERCEPTIONS OF BOOKS (BD): An awareness of the organisational attributes of books. Books are selected on this basis and used in relation to the value of the information they contain. When selecting books for study purposes, 1 often examine their 'search apparatus' (such as the index, list of contents, chapter headings, cross references). DEEP PERCEPTIONS OF METHODS OF ASSESSMENT (AD): An awareness of the content,
purpose, types and benefits of tests and exams, as well as the value of written feedback from teachers. The educational purpose of tests is usually clear to me. DEEP PERCEPTION OF LEARNING SPACE (LD): An appreciation of the importance of the
relational, rather than the functional, uses of chalkboards and the equipment in classrooms or laboratories as well as an awareness of where one sits in a classroom. 1 usually notice how the teacher uses the blackboards. DEEP PERCEPTIONS OF HUMAN RELATIONSHIPS (RD): An appreciation that one can be
314 helped and guided by others and that human interaction is affected by one's own attitudes. 1 am conscious of the way that my attitudes towards teaching and learning affect my relationships with others. SURFACE PERCEPTIONS OF COURSE CONTENT (cs): Attention specifically on the detail of the content in terms of its volume, structure and perceived relevance. The structure of the content in the subjects 1 am studying is usually clear to me. SURFACE PERCEPTIONS OF LEARNING SPACE (Is): A concentration on those aspects of the learning environment (noise, legibility, equipment) which affect the ease and accuracy of information transfer. I usually notice the legibility of what is written on the blackboard or on an overhead transparency. SURFACE PERCEPTIONS OF HUMAN RELATIONSHIPS (rs): An uncritical reliance on the words of the teacher or textbook while ignoring other aspects of the teaching/learning relationship. In class I usually write down what the teacher says or writes on the board. WORKLOAD (wl): A feeling that too much work is covered and expected, reflected in too many topics and too much written work, giving rise to a feeling of pressure. There seems to be too much work to get through in the course here. 2. Discrete study approach variables (See Note below) DEEP APPROACH (DA): A conscious intention to understand new material even if this requires considerable effort. 1 usually set out to understand thoroughly the meaning of what I am required to learn. INTRINSIC MOTIVATION (IM): A strong interest in, and even excitement about the subject being studied that extends beyond the demands made in class. My main reason for being here is so that 1 can learn more about the subjects which really interest me. RELATING IDEAS (RI): Relating ideas between, as well as within, subjects, as well as a conscious attempt to relate material to real life situations and integrate it within a personal framework. 1 try to relate ideas in this course to ideas in other subjects whenever possible. USE OF EVIDENCE (UE): The critical use of evidence in order to draw conclusions and an examination of evidence where this is used to support an argument. When I'm reading an article or research report, 1 generally examine the evidence carefully to decide whether the conclusion is justified. COMPREHENSION LEARNING (CL): Divergent thinking or 'mapping out' a subject as part of the comprehension of new ideas. I like to play around with ideas of my own even if they don't get me very far. REFLECTION (RE): The process of reflecting on past learning experiences or real life experiences and deriving fresh insights from them. I sometimes think about things 1 have previously learned and change my mind about their meaning. STRATEGIC APPROACH (St): A strategic manipulation of resources to meet perceived academic requirements. When I am doing a piece of work, I try to bear in mind exactly what that particular teacher seems to want. OPERATION LEARNING (O1): An engagement of problem solving that is reliant on factual detail and logical analysis. I generally prefer to tackle each part of a topic or problem in order, working out one step at a time. ACHIEVEMENT MOTIVATION (Am): A motivation to succeed, especially in competition with others. It is important to me to do things better than other people, if l possibly can. MEMORISING APPROACH (ma): A rote learning approach to studying in which important information to be "learned" (such as facts and definitions) is committed to memory by way of repeated rehearsal. I learn things by writing them over and over or by saying them to myself. FRAGMENTED APPROACH (fa): An inability to see the relationships between ideas or concepts. The "learning" of material that is perceived to be fragmented and poorly understood. Much of what l am studying seems to consist of unrelated bits and pieces. SYLLABUS-BOUNDNESS (sb): A narrow focus on the requirements of the task and a preference for clear guidelines and structure. I like to be told exactly what to do in essays, assignments or projects. FEAR OF FAILURE (ff): A general concern with failing, but linked to exam tension, speaking in class, and pressure of work. 1 am scared that 1 might fail this course this year.
315
IMPROVIDENCE (ip): A failure to integrate detail into an overall picture and an over cautious reliance on detail and procedure. Although I generally remember facts and details, l find it difficult to fit them together into an overall picture. DISORGANISED STUDY METHODS (ds): A general disorganisation reflected in poor time management (including putting off work), distractions and a backlog of important work. I find it difficult to organise my study time effectively. GLOBETROTTING (gL): An inability to back up a general picture with the necessary detail, leading to unsubstantiated conclusions and the use of irrelevant material. Although I have a fairly good general idea of things, my knowledge of the details is fairly weak. EXTRINSIC MOTIVATION (eM): Studying and subject choice is seen as specifically career-related and as a means to obtaining a good job. My main reason for being here is that it will help me to get a better job.
Notes 1. The term study orchestration refers to the contextualised study approach adopted by individual students or by groups of students. 'Contextualised' in this sense refers to (perceptions of) the learning context within which particular forms of study approach are adopted in a particular subject. 'Study approach' is used in a generic sense to represent conceptually coherent combinations of subscale constructs representing various forms of motivation, intention, learning style, and so on. 'Study approaches' are thus contextualised in various manifestations that are interpreted as being representative of the study behaviour of particular individuals or groups of individuals. 2. All of the contextual perception subscales except workload (wl), owe their conceptual origins to the work of Meyer (1988). The workload (wl) subscale derives from the Course Perception Questionnaire used in the study by Entwistle and Ramsden (1983). 3. The study approach subscales are substantively those of the Approaches to Studying Inventory (ASI) used by Entwisfle and Ramsden (1983). The original surface approach subscale (sa) of the ASI is split into two variables in the present study (ma and fa) based on a study by Meyer and Watson (1991), while the inclusion of the reflection subscale (RE), owes its conceptual origin to the work of Boud, Keogh and Walker (1985).
References Boud, D., Keogh, R. and Walker, D. (eds.) (1985). Reflection: Turning experience into learning. Kogan Page, London. Cross, C.T. (1990). 'National goals: four priorities for Educational Researchers', Educational Researcher 19(8): 21-24. Educational Testing Service Annual Report (1990). Helping America Raise Educational Standards for the 21st Century. Princeton, New Jersey. Eley, M.G. (1991). 'Differential adoption of study approaches within individual students' Higher Education 23, 231-254. Entwistle, N.J. and Ramsden, P. (1983). Understanding student learning, Croom Helm, London. Entwistle, N.J. and Kozrki, B. (1985). 'Relationships between School Motivation, Approaches to Studying, and Attainment, among British and Hungarian Adolescents', British Journal of Educational Psychology, 55: 124-137. Entwistle, N.J., Meyer, J.H.F. and Tait, H. (1991). 'Student failure: disintegrated patterns of study strategies and perceptions of the learning environment', Higher Education, 21 : 249-261. Entwistle, N.J. (1992). 'Student learning and study strategies', in The Encyclopedia of Higher Education, Pergamon.
316 Fraser, B.J., Walberg, H.J., Welch, W.W. and Hattie, J.A. (1987). 'Syntheses of Educational Productivity Research', InternationalJournal of Educational Research 11: 145-252. Goodman, L.A. and Kruskal, W.H. (1979). Measures of Association for Cross Classifications. New York: Springer-Verlag. Meyer, J.H.F. (1988). 'Student perceptions of learning context and approaches to studying', South African Journal of Higher Education 2: 73-82. Meyer, J.H.F. (1991). 'Study orchestration: the manifestation, interpretation and consequences of contextualised approaches to studying', Higher Education, 22:297-316. Meyer, J.H.F. and Dunne, T.T. (1991). 'The study approaches of Nursing students - Effects of an extended clinical context', Medical Education, 25:497-516. Meyer, J.H.F. and Muller, M.W. (1990). 'Evaluating the Quality of Student Learning. I - an unfolding analysis of the association between perceptions of learning context and approaches to studying at an individual level', Studies in Higher Education, 15(2): 131-154. Meyer, J.H.F., Parsons, P. and Dunne, T.T. (1990a). 'Individual study orchestrations and their association with learning outcome', Higher Education, 20(1 ): 67-89. Meyer, J.H.F., Parsons, P. and Dunne, T.T. (1990b). 'Study orchestration and learning outcome: Evidence of association over time among disadvantaged students', Higher Education, 20(3): 245-269. Meyer, J.H.F. and Watson, R. (1991). 'Evaluating the quality of student learning. II - Study orchestration and the curriculum', Studies in Higher Education 16:251-275. Ramsden, P. (1988). 'Context and strategy. Situational influences on learning', in R.R. Schmeck (ed.), Learning strategies and learning styles. New York: Plenum Press. Ramsden, P., Martin, E. and Bowden, J. (1989). 'School Environment and Sixth Form Pupils' Approaches to Learning', British Journal of Educational Psychology, 59: 129-142. Somers, R.H. (1962). 'A new asymmetric measure of association for ordinal variables', American Sociological Review, 27:799-811.