School Learning For Transfer Research Paper

1. Two Basic Approaches

The expectation for transfer has as long a history as institutional learning itself. Plato argued that the study of abstract reasoning assists the solution of daily problems. Similarly, the debates about Talmudic and Biblical texts in ancient times were argued to ‘sharpen minds.’ The scientific study of transfer has a long history as well; it dates back to the beginning of the twentieth century with the studies by Thorndike and Woodworth (1901) in which the idea that Latin and other taxing school subjects developed one’s general ‘faculties of mind’ was challenged. The expected transfer was not found; mastery of Latin, Greek, or geometry did not facilitate either performance or efficiency in learning other subjects. Thorndike’s findings led to his formulation of the theory of ‘identical elements’: transfer from A to B is likely to occur the more elements are common to both. It followed that transfer can take place only in the relatively rare case when there is clear and apparent similarity between the constituent elements of the learned topics. Thus, one would have to learn a host of independent issues and procedures as no transfer in the absence of identical elements should be expected.

Thorndike’s theory was countered by a more cognitively-oriented and Gestalt-flavored alternative that emphasized the meaningful conceptual understanding of underlying principles which then could be thought- fully applied to new situations. The more general the learned principle, the farther could it be transferred to new instances. Judd (1908) had his subjects learn to shoot at targets submerged in water after having learned principles of light refraction in water. Having learned these principles, subjects were much better at hitting the targets than those who have only practiced the activity.

As we shall see below, the two approaches—transfer as a matter of common elements and transfer as a matter of higher order cognitive processes of under- standing and abstraction—although having under- gone profound modifications and developments over the years, continue to this day to dominate the field.

2. Paucity Of Findings

Research on transfer as well as practitioners’ expectations for transfer are characterized by an ongoing pendulum-like oscillation between the two approaches. Such oscillation is not so much the result of evidence that supports both approaches as it is a function of the paucity of findings clearly supporting either one of them. Indeed, one of the hallmarks of the field is the great discrepancy between the expectation for transfer from school learning to the acquisition of new subjects or to the handling of daily events and the findings from controlled transfer studies. The latter often do not yield the transfer findings one would have expected. A typical case is the study by Pea and Kurland (1984) who have found that children learning LOGO programming showed no advantage in their ability to plan over children who did not master LOGO.

Transfer, when at all found in carefully carried out studies, appears to be highly specific, as when a particular procedure or principle is applied in a new situation which bears great and apparent (often analogical) similarity to the learning material, and when the process is accompanied by such facilitation as coaching, specific guidance and cueing (see Bransford and Schwartz (1999, Cox (1997, for detailed reviews). Such findings would seem to support Thorndike’s theory and pessimism about transfer, as summarized by Detterman (1993): ‘In short, from studies that claim to show transfer and that don’t show transfer, there is no evidence to contradict Thorndike’s general conclusions: Transfer is rare, and its likelihood of occurrence is directly related to the similarity between the two situations’ (p.15).

3. Transfer On The Rebound

Whereas Thorndike’s pessimism seemed to have won the day for a while, the research tradition of Judd (1908), with its emphasis on comprehension and general (nonspecific) transfer gradually showed its strength. As research of this tradition tended to show, when relevant principles or strategies are mindfully attended to, or better yet—either abstracted by learners and/or metacognitively monitored—transfer to a variety of instances can be obtained even in the absence of apparent common elements. At the same time, other research on the transfer of skills, ostensibly still in a tradition more similar to that originated by Thorndike (Thorndike and Woodworth 1901), has also yielded positive results. Thus, for example, Singley and Anderson (1989) have shown that when training to near automaticity of discrete skill components is carried out, transfer from one task (e.g., a line editor) to another (text editor) can be virtually perfect. Also more field-based research found support for impressive transfer from the study of universitytaught disciplines: Nisbett et al. (1993) have shown that the study of psychology, and also to a lesser extent medicine (but not law or chemistry), improve students’ abilities to apply statistical and methodological rules of inference to both scientific and daily problems.

4. The High And The Low Roads To Transfer

The renewed success of the two lines of research, often addressing higher order cognitions and information processing exercized by multiple activities, suggests that transfer may not be a unitary process as the two approaches differ in important ways. Observing these differences, Salomon and Perkins (1989) have developed a theory to account for the possibility that transfer takes either one of two routes (or a combination thereof), described as the high roAdvand the low road of transfer. The low road reflects to an extent the Thorndikian line, and more recently that of Anderson’s skill-acquisition and transfer theory. It is taken when skills, behaviors, or action tendencies are repeatedly practiced in a variety of situations until they are mastered to near-automaticity and are quite effortlessly applied to situations whose resemblance to the learning situations is apparent and easily perceived. Learning to drive and learning to read are two cases in point as is transfer from one text editor to another as studied by Singely and Anderson (1989). Other candidates for low road transfer are attitudes, cognitive styles, dispositions, and belief systems the application of which to new situations is rarely a mindful process.

In contrast, the high road to transfer is characterized by the process of mindful abstraction of knowledge elements that afford logical abstraction: principles, rules, concepts, procedures, and the like. It is this mindfully abstracted, decontextualized idea (‘ethnic oppression may lead to revolt’) that becomes a candidate for transfer from one particular instance (the Austro-Hungarian Empire) to another (Russia and Cechenia). Simon (1980) commented in this respect that ‘To secure substantial transfer of skill acquired in the environment of one task, learners need to be made explicitly aware of these skills, abstracted from their specific task content’ (p. 82, emphasis added).

There is ample evidence to support the role of the high road in obtaining transfer. One of the studies by Gick and Holyoak (1983) illustrates this point. Participants who were given two stories and were asked to write a summary of how the two stories resembled each other (that is, their common moral), showed a 91 percent transfer to the solution of another story, relative to 30 percent transfer of a non-summary group. Bassok (1990), showed that mastering algebraic abstractions (plus examples) allowed students to view physics problems as particular instances to which the more abstract algebraic operations could be applied. Physics, on the other hand, is too particular and thus students do not expect and do not recognize any possible relationship between it and algebraic operations. Research also directs attention to the role played by self-regulation and metacognitions in the process of mindful abstraction and transfer via the high road. Following the studies and review of Campione et al. (1982), Salomon et al. (1989) have shown that students interacting with a semi-intelligent computerized Reading Partner that provides metacognitive-like guidance, tend to internalize that guidance and transfer it to new reading as well as to writing situations.

The high road low road theory sheds light on the many failures of obtaining transfer in controlled studies. Close examination of such studies suggests that in many cases neither the low road of repeated practice nor the high road of mindful abstraction was taken. Not enough time is allocated for practice for the former, and not enough attention is given for mindful abstraction for the latter. As a consequence, neither near automatic transfer on the basis of easily recognized common elements can be attained, nor farther transfer on the basis of metacognitively guided mindful abstraction.

5. New Approaches

Research on transfer, until recently, did not challenge the basic paradigm and conception of transfer as a process involving change of one’s performance on a new task as a result of his or her prior performance on a preceding and different task. Such a conception of transfer has recently become challenged on the basis of new theories of learning, the role of activity, and the place of cognitions therein. These challenges can be arranged along a dimension ranging from the least to the more radical, relative to traditional notions of transfer.

Bransford and Schwartz (1999), coming from a mainstream tradition of cognitive science applied to instructional issues, argue that the current ways of demonstrating transfer are more appropriate for studying only full-blown expertise. According to them, ‘There are no opportunities for … [students] to demonstrate their abilities to learn to solve new problems by seeking help from other resources such as text or colleagues or by trying things out, receiving feedback, and getting opportunities to revise’ (p. 68). They recommend replacing the typical transfer task which they call ‘sequestered problem solving’ (SPS) by a conception of transfer as ‘preparation for future learning’ (PFL). Thus, one would measure transfer not by having students apply previously acquired knowledge or skill to new situations, demonstrating either knowledge how or knowledge that something, but rather by demonstrating knowledge with—thinking, perceiving, and judging with whatever knowledge and tools are available even if that knowledge is not consciously recalled. In other words, transfer would be expected to become demonstrated by way of ‘people’s abilities to learn new information and relate their learning to previous experiences’ (Bransford and Schwartz (1999, p. 69).

An illustration of the above can be seen in a study in which college and fifth grade students did not differ in their utilization of previous knowledge and educational experiences for the creation of a statewide plan to recover the bald eagle. This is a typical SPS way of measuring transfer, suggesting in this case that previous educational experiences did not transfer to the handling of the new problem. However, when the students were asked to generate questions that would need to be studied in preparation for the creation of a recovery plan (applying a PFL approach), striking differences were found between the two age groups in favor of the college students. Focusing on different issues, the two groups showed how, knowingly or not, previous learning has facilitated the generation of the list of questions to be asked. The use of tools and information resources would also show how previous learning of skills and knowledge becomes actively involved in the process of solving a new problem.

While Bransford and Schwartz (1999) continue to adhere to the traditional conception of learning as knowledge acquisition and to cognitions as general and transferable tools of the mind, others have developed an alternative, socio-cultural approach according to which neither the mind and its content (e.g., arithmetic algorithms) should be treated as a neutral toolbox ready for application wherever suitable, nor should knowledge-in-use and social context of activity be taken as two independent entities. Learning is a matter of participation in a community of learners, thus knowledge is not a noun denoting possession but rather a verb denoting the active process of knowing-as-construction within a social context of activity. One underlying assumption is that learner, activity, and content become united through culturally constructed tools. A second assumption is that learning is highly situated in a particular activity context of participation, and its outcomes are ‘becoming better attuned to constraints and affordances of activity systems so that the learner’s contribution to the interaction is more successful’ (Greeno 1997, p. 12). Seen in this light, the traditional cognitivist concern with the transferability of acquired knowledge across tasks becomes, from the newer situative perspective, an issue concerned with the task and the ‘consistency or inconsistency of patterns of participatory processes across situations’ (Greeno (1997, p. 12).

An important practical implication of this approach is that one would need to take into consideration the kinds of participatory activities to which school-based learning might be expected to transfer, rather than expect decontextualized school materials to transfer all on its own to other tasks, situations, and activities. Students are likely to restructure new situations to fit their previous practices and thus make the old available for transfer to the new. Illustrating this approach to learning and cognition, Saxe (1991), showed that Brazilian children’s mathematical experience as street vendors affected their math learning when later on they became enrolled in school, and vice versa. ‘[I]n recurring activities like practices of candy selling and schooling—contexts in which knowledge is well learned and in which similar problems emerge on a repeated basis—transfer is a protracted process. In their repeated efforts to solve practice-linked problems, individuals attempt to structure problem contexts in coherent and more adequate ways’ (p. 179).

The new approaches to transfer clearly deviate from the traditional views. They suggest new ways of measuring transfer, and more radically—challenge traditional assumptions by treating knowledge, skill, task, activity, and context as one participatory activity. By so doing, they emphasize the uniqueness of each activity setting. An important aspect of this approach is its ability to incorporate within the idea of participatory activity motivational and dispositional elements also. These elements have so far been neglected by much research on transfer, although they may be crucial for students’ choice to treat each situation as different, or as allowing students to exercise familiar patterns of participation. Still, it appears that despite these important novelties, the principles of transfer as a function of similarity between situations (or activities) and of depth of understanding general principles (e.g., of useful participation in a community of learners) have not yet been replaced but only redefined.

Bibliography:

1. Bassok M 1990 Transfer to domain-specific problem solving procedures. Journal of Experimental Psychology: Learning Memory and Cognition 6: 522–33
2. Bransford J, Schwartz D 1999 Rethinking transfer: A simple proposal with multiple implications. Review of Educational Research 24: 61–100
3. Campione J C, Brown A L, Ferrara R A 1982 Mental retardation and intelligence: Contributions from research with retarded children. Intelligence 2: 279–304
4. Cox B D 1997 A rediscovery of the active learner in adaptive contexts: A developmental-historical analysis of transfer of training. Educational Psychologist 32: 41–55
5. Detterman D R 1993 The case for prosecution: Transfer as an epiphenomenon. In: Detterman D K, Sternberg R J (eds.) Transfer on Trial: Intelligence, Cognition and Instruction. Ablex, Norwood, NJ
6. Gick M L, Holyoak K J 1983 Schema induction and analogical transfer. Cognitive Psychology 15: 1–38
7. Greeno J G 1997 Response: On claims that answer the wrong questions. Educational Researcher 26: 5–17
8. Judd C H 1908 The relation of special training to general intelligence. Educational Review 36: 28–42
9. Nisbett R E, Fong G T, Lehman D R, Cheng P W 1993 Teaching reasoning. In: Nisbett R E (ed.) Rules for Reasoning. Erlbaum, Hillsdale, NJ
10. Pea R D, Kurland D M 1984 On the cognitive effects of learning computer programming. New Ideas in Psychology 2: 37–68
11. Salomon G, Globerson T, Guterman E 1989 The computer as a zone of proximal development: Internalizing reading-related metacognitions from a Reading Partner. Journal of Educational Psychology 81: 620–7
12. Salomon G, Perkins D N 1989 Rocky roads to transfer: Rethinking mechanisms of a neglected phenomenon. Educational Psychology 24: 113–42
13. Saxe G B 1991 Culture and Cognitive Development: Studies in Mathematical Understanding. Erlbaum, Hillsdale, NJ
14. Simon H A 1980 Problem solving and education. In: Tuma D T, Reif R (eds.) Problem Solving and Education: Issues in Teaching and Research. Erlbaum, Hillsdale, NJ
15. Singley M K, Anderson J R 1989 The Transfer of Cognitive Skill. Harvard University Press, Cambridge, MA
16. Thorndike E L, Woodworth R S 1901 The influence of improvement in one mental function upon the efficacy of other functions. Psychological Review 8: 247–61