Attaining transfer of learning through workplace learning
Authors: Ian Cornford
Affiliations: University of Technology, Sydney
Abstract:Workplace learning remains an important policy issue in many countries. No issue in workplace learning is perhaps more important than transfer of learning since the intention of all practical skill learning is that the original learning be applied beyond the initial learning occasion and setting. This paper explores the important issue of transfer of learning and the problems involved in attaining effective transfer of learning in the workplace, drawing upon more recent research and publications. The wide lack of understanding of the complex nature of transfer of learning and the lack of recognition of the importance of training and learning amongst employers are seen as creating particular difficulties in attaining effective positive transfer between different contexts.
Introduction
Over the fifteen years much attention has been given to increasing workplace productivity and national economic performance in an era of globalization and international competitiveness (Fuller, Munroe & Rainbird, 2004). Increased knowledge, highly skilled performance and problem solving are seen as central to enhancing workplace productivity. In achieving these goals the concept of transfer of learning is central. Transfer of learning involves the application of knowledge and skills gained in one setting to other settings that differ in terms of ‘subjects, settings, people, behaviors (sic) and/or time’ (Stokes & Baer, 1977, p.350). Transfer of learning is an essential element in problem solving (Haskell, 2001) while the current interest in generic skills revealed in vocational and school education amounts to a rediscovery of the importance of transfer of learning. However it is doubtful whether educators, policy makers and employers fully understand the complexity of the learning involved in either transfer of learning or generic skills (Cornford, 2005).
Transfer of learning involves very considerable complexity, and attaining positive transfer has long been recognized is one of the important and difficult problems in educational psychology (Cox; 1997; Haskell, 2001). It has been, and still is assumed incorrectly to be a passive, naturally occurring phenomenon (Stokes & Baer, 1977). In brief, transfer of learning is what all vocational educators either in the workplace or classrooms hope to achieve, and the credibility of all real-world education for the workplace is dependent upon achieving it. Yet the estimates are that only about ten percent of training is effectively transferred to the workplace (Detterman, 1993). The worrying factor in this is that much training is carried out in the workplace or in simulated settings, with such settings seen as most likely to encourage transfer of learning (Billett, 1996; Eraut, 2004).
Many of the current policies and practice of workplace or work-based learning are based on the assumption that the critical factors reside in social and cultural contexts (eg Billett, 1996; Eraut, 2004; Fuller et al., 2004)). Yet Detterman’s (1993) and others’ judgments (see Haskell, 2001) about transfer of learning indicate that there is a low rate of success of transfer despite in many cases similar social contexts in workplace and learning settings. Logically there appear to be significant other factors that are not being identified and captured in currently dominant socio-cultural paradigms and approaches to learning and transfer. There is also the problem that Fuller et al. (2004, p.1) have identified from the investment in human capital perspective, the increase in formal qualifications has not led to increased productivity. The conclusions that may be drawn from this in part are that formal qualifications are not including greater understanding of learning and adaptation processes, and specifically understanding of what transfer of learning is and how to achieve it.
This paper will examine problematic issues in socio-cultural approaches to transfer of learning before turning to analyzing the complexity of transfer of learning issues and what needs to be done to attain effective positive transfer in workplace settings. Aspects of workplace culture created by employers, managers and educators erect substantial barriers to effective transfer and these will be explored also.
Sociological and socio-cultural approaches to transfer and work-based learning
Socio-cultural paradigms have been seen as critical in the issues of transfer and learning by many theorists (eg Billett, 1996; Eraut, 2004). However the importance of these elements has been overstated where learning is concerned and thus, at best, what they offer is a half-truth. It is argued very strongly here that it is cognitive-constructivist views that hold the key to learning issues in transfer. Cornford (1996, p.5) stated almost a decade ago, when the infatuation with situated learning and socio-cultural paradigms was developing: “…it is the self-regulatory and control factors within the individual that determine whether a skill is learned or not, or to what degree social influence is accepted and whether there will be internalization or rejection of standards. Acquisition and maintenance of skills are dependent upon cognitive factors largely controlled by the individual..’. It is not claimed that cognitive psychologists hold all the answers, and it is recognized that social factors are often important (see Cornford, 1996, p.5). For example certainly power relationships in workplaces and workplace cultures (eg see Fuller et al., 2004) can be important influences on learning. What is argued here is that cognitive processes are central to understanding and attaining effective learning for workplaces but social factors will also influence acquisition and performance, hence there needs to be a distinct balance between the cognitive and social sides of the equation.
The main problem in essence is socio-cultural and sociological paradigms cannot explain learning processes, with the learning being taken for granted and the results of learning becoming the bases of these theories. What is more, when viewed objectively, sociology and derived disciplines have a poor track record over the past century in both predictive and explanatory value (eg see Sherden, 1998). Sherden (1998, Chapter 7) has argued that this is because of the complexity of social situations and foundations that these approaches are based upon. However, as Sherden indicates, through a citation of earlier writing by Karl Popper, mental processes are an important factor in understanding change in human society (Sherden, 1998, p.197).
Until now concentration upon socio-cultural factors in workplace learning literature, largely to the exclusion of cognitive elements, has resulted in far less attention to the learning part of the equation when the learning processes are absolutely central to the acquisition, storage and refinement of schemas or mental models in memory which are foundations of human understanding and behaviour (see Rumelhart & Norman, 1978). Learning per se is widely recognized as an important factor in the organizational development and management literatures (eg French and Bell, 1995), with learning being central to bringing about changes in culture, both in terms of the culture of a specific organization (French & Bell, 1995) but also of the wider cultures to be found in the wider communities of any nation (eg see Bandura, 1977, 1997). Garvan (1997), in analyzing issues in organizational learning and the learning organization, correctly argues that organization are inanimate and cannot learn; only individuals can learn and thus influence the actions of the organization and the quality and nature of the culture. Transfer of learning is certainly all about learning and adaptation of learning that leads to more complex understanding and knowledge. As will be argued further below, one of the main impediments to attaining effective positive transfer is hostile cultures. That is lack of understanding and positive attitudes within organizations, and specifically understanding of owners, manager and trainers (Broad & Newstrom, 1992; Yelon, 1992). Policy makers also are lacking understanding of both learning and transfer of learning (Cornford, 2005). For there to be enhanced workplace learning and performance there needs to be transfer of learning and for this to occur there needs to be changes in attitudes to teaching transfer of learning and changes in the workplace cultures where this learning is to be returned to and built upon (Eraut, 2004).
The most useful explanatory and predictive learning theory with application to both organizations and the wider society is that of Bandura, specifically his observational learning model involving modeling (Bandura, 1977) and later social-cognitive theory encompassing self-efficacy (Bandura, 1997). Central to Bandura’s social-cognitive theory is the concept of reciprocal determinism where person, behaviour (activity) and environment interact in complex ways (see Bandura, 1997). Bandura is an important, established theorist and researcher in social learning theory, yet references to his work, with its major body of underpinning experimental evidence, are conspicuous by their absence from the work of the sociological, socio-cultural theorists, situated learning theorists and activity theorists. This is a grave error since Bandura managed to successfully combine both social and cognitive aspects of learning as early as 1977. Unlike other theorists who have tried to reconcile social and cognitive elements (eg Billett, 1996), Bandura recognized that learning is a precursor to the establishment of behaviour and culture, and learning of individuals is the basis of cultural change. Bandura’s theories underpin the following analysis of transfer of learning that focuses upon acquisition, retention and performance processes and also motivation as the supremely powerful factor that is necessary for learning. These concepts apply equally to skill learning as to learning of more abstract cultural and social elements. In Bandura’s work social factors are seen as powerful in the selection and creation of models as well as allowing the performance of acquired knowledge and skill, especially practice and feedback, to refine the mental models or schemas, in different social settings.
Before looking at general principles in teaching and learning for transfer and the acquisition phase it is necessary to define and clarify the different types of transfer.
Types of transfer
Transfer of learning has fallen in and out of fashion in psychological research (Haskell, 2001), and thus it is not surprising that there is frequently lack of understanding even by educational psychologists of this often neglected phenomenon. There are a number of different types of transfer and understanding of this is important if real advances in workplace learning and training are to be made.
Earlier work identified a number of different types of transfer (see Cornford, 2002; Haskell, 2001). These include positive transfer where previous learning assists with later learning, negative transfer where earlier learning adversely affects later learning and zero transfer where there is no facilitative or inhibitory effect for earlier learning because the later material is so different from earlier learning. There is also the practically important differentiation between near and far transfer. Near transfer is essentially transfer that differs not too much from the original learning and specifically has similar or identical cues to indicate that application of previously learned knowledge and skills. Far transfer is transfer that involves fairly radical modification of previous learning for the new situation or where the cues are considerably different from those involved in the original application and learning. This radical, highly creative type of transfer is of course much desired in workplace and business settings to solve problems that even high status, highly paid managers and CEOs cannot solve (Cornford, 2002). There is also general transfer: this is essentially a notion that by leaning something like mathematics, aspects of learning are transferred to other subjects and areas. The notion of general transfer has been under sustained attack since Thorndike’s work (Cox, 1997: Haskell, 2001). Research into the development of expertise indicates that expertise, and to a degree learning more generally, is specific to the area of greatest familiarity (see Cornford, 1999; Haskell, 2001).
The above types of transfer are generally widely accepted by researchers. However Haskell (2001) has identified a wider range of types of transfer. The value of the wider classification has yet to be determined, but may have more value with language and liberal arts teaching since the above categorizations of long standing would appear to suffice for workplace learning. The important question is which types of transfer can be reasonably be expected to be achieved by training for workplaces to meet everyday expectations relating to problems for which already recognized solutions can be found.
Detterman (1993, p.21) has claimed that ‘significant transfer’, by which he means far transfer and also general transfer, are extremely rare yet Haskell (2001) has correctly argued that transfer is in fact very common depending on types and how you define it. However, Haskell agrees that highly creative and original solutions to major problems do not occur easily or very frequently. He also argues, like most researchers, that attaining transfer from training with any degree of confidence is difficult. Certainly it is unrealistic to believe that highly creative solutions to difficult problems can be readily attained by training. Yet, while even attaining near transfer remains a challenge with novice or inexperienced learners, there is certainly evidence that near transfer can be reliably attained if there is conscious awareness in learners and teachers and specific training for transfer is engaged in (Cornford, 2002; Haskell, 2001). Hence the position adopted here is that near transfer can reasonably be expected to occur if there is conscious, effective education and learning, and social conditions are receptive to applications of new learning. The following sections explore what can be considered effective education and training and what must be done to ensure receptivity of application of learning so that transfer can occur.
Effective education for transfer: some general principles
A major revolution in education thinking is required to ensure that transfer is no longer seen as a passive, naturally occurring phenomenon (Stokes and Baer, 1977), but becomes a conscious part of education for life beyond the initial learning setting, and specifically the workplace. This includes whether the learning/teaching occurs specifically in the workplace or whether it is off-site. Detterman’s (1993) analyses clearly imply that workplace training often fails when the task changes slightly from the original setting or where there are changes in technology or machinery. What seems central is meaningful education, as opposed to narrow training, and preparation of learners to anticipate problem solving and transfer. What is needed here are changes in the culture of teaching and training, and also employer expectations, with the notion of educating more broadly often in conflict with the realities of narrow training for workplace purposes.
Perkins and Salomon (1989) drew attention to the concept of high road transfer where openness to the need for adaptation and modified ways of thinking are central to achieving successful transfer. This aligns with Rumelhart and Norman’s (1978) concept of tuning as part of schema/mental model construction in memory, and adaptation of initial structures through later applications to make the schemas more effective. In turn, tuning involves generalization which seems to be a key to effective transfer (Stokes & Baer, 1977) and central too to the concept of generic skills (Cornford, 2005). If skill development can be established in the minds of educators and learners as further tuning, that is creating more effective models to guide thinking and behaviour after the initial acquisition of information, then more effective learning will be encouraged over the lifespan. Further support regarding the importance of tuning of schema comes implicitly from Fitts’ skill learning theory where the practice–fixation phase of skill learning, involving practice and feedback, leads to increasingly effective mental models guiding performance and also better actual performance in terms of speed and accuracy (Cornford, 1999).
Recognition of the need for problem solving perhaps needs to start early in educational processes but importantly there must be recognition of the limitations of the novice learner. If too much is demanded from those with a low knowledge base, or the initial learning too difficult, there will be ‘turning off’ and loss of motivation vital for other aspects involving maintenance of skills and application (Cornford, 1996). There also needs to be recognition that the novice should not be pushed into transfer and problem solving before there is solid understanding (see below). Too frequently transfer is expected without adequate understanding and learning having taken place (Cornford, 2002; Haskell, 2001). The classic example is with mathematics, where learners are expected too often to apply new principles to different examples before having grasped the basics. Employment of worked examples in teaching is currently seen as one way of overcoming these problems.
What is also central to effective learning for near transfer is guidance by more knowledgeable teachers who understand real world applications and the variety of such applications (Cornford, 2002; Haskell, 2001). It is necessary to reassign a substantial role for teaching, and even training, rather than hold the currently fashionable view that learning is the only thing that matters. Effective teaching and learning go hand-in-hand and this reality has been recognized over many centuries (see Bandura, 1977, 1997). We need to recognize this, even when confronted with the reality that the teacher cannot learn for the learner.
Initial learning
Cornford (2002) has advanced a model of sequential skill development, based on empirical evidence (see Cornford, 1991). This conceptualizes five steps in learning and schema development in the skill development process leading to effective transfer of learning. It also establishes a number of conceptually distinct assessment stages that provide feedback to both teacher and learner on progress in learning to successfully transfer initial learning. This model is developed for near transfer and conceives of a series of five steps in learning that moves from understanding and then consolidation of basic principles, to a stage of generalization from initial principles and then also consolidation of generalization, before moving still further to more generalization and transfer of learning to other settings. The model recommends that teaching and practice occur as much as possible in controlled situations where the buzzing complexity of the real workplace for does not distract and confuse the novice learner in early stages, and where understanding of the cues that signal the need to apply these particular skills can be learned in conjunction with the principles guiding the skills. Central to the skill development model also are the use of suitable performance models so that it is possible to see and learn from these how things need to be done. The skill model operates along the long recognized lines of modeling, practice, feedback and learning through application (Bandura, 1977, 1997), and seeks to establish sound schemas with solid understanding and performance before transfer to the complex, natural setting is attempted. (See Cornford, 2002 for more details.)