Tuesday, 3 September 2013

The Conditions of Meaningful-Learning

One of the main researchers who has connection with the constructivist movement was David Ausubel. In 1968, he put forward the case that the most important thing for teachers to know at the outset of the teaching is what each student knows already. However, he held a different approach to how the teaching material should be presented in the classroom or by self-study than Bruner. He argued that students need guidance if they are to learn effectively and he advocated the direct instruction learning approach. Ausubel (1968) focussed on both the presentational methods of teaching and the acquisition of knowledge. He made a major contribution to learning by studying and describing the conditions that lead to ‘meaningful learning’. He attempted to find ‘the laws of meaningful classroom learning’.


According to Ausubel, meaningful learning presupposes:
• “That the learning material itself can be non arbitrarily (plausibly, sensibly, and non randomly) and substantively (non verbatimly) related to any appropriate cognitive structure (possesses “logical meaning”).
• That the particular learner’s cognitive structure contains relevant anchoring idea(s) to which the new material can be related.
• The interaction between potentially new meanings and relevant ideas in the learner’s cognitive structure gives rise to actual or psychological meanings. Because each learner’s cognitive structure is unique, all acquired new meanings are perforce themselves unique.”
                                                                                                                           Ausubel et.al, (1978)


The meaningful learning processes exist when the new concept can be linked to the pre-existing concept in the learners’ cognitive structure (for example, already existing relevant aspect of knowledge of an image, an already meaningful symbol, a known concept or a proposition). The new concept interacts in a non arbitrary (in the sense of plausibly, sensibly and non randomly), and substantive (non verbatimly) basis with established ideas in cognitive structure. Thus, meaning derives directly from associations that exist among ideas, events, or objects. As the new knowledge is subsumed into the existing knowledge, it interacts and modifies it and the entire new matrix now becomes more elaborate and new linkages form between concepts. Obviously, this theory can only become reality if the teacher finds out what the learner already knows. Orton (2004) argued that, if an attempt is made to force children to assimilate and accommodate new mathematical ideas that cannot link to knowledge which is already in an existing knowledge structure, then the ideas can only be learned by rote. In contrast with meaningful learning, rote learning results in arbitrary literal assimilation of new knowledge into cognitive structure. It occurs when no relevant concepts are accessible in the learner’s cognitive structure.


 ‘Rote-meaningful’ learning is a continuum, which relies on the learner and differs from one learner to another. The feature of the cognitive structure of the learner interconnects in a diverse degree from topic to topic in the ‘rotemeaningful’ continuum. The learner’s existing knowledge and the way that new knowledge is linked to existing knowledge involves subsumption. Ausubel postulates that cognitive structure is hierarchically organised which means the less inclusive sub-concepts and details of specific data are organised, under the more inclusive concepts. Therefore, good expository teaching should be given to the learner to ensure that a new concept is linked to relevant existing knowledge. According to Ausubel (1968) the advance organiser is “an advanced introduction of relevant subsuming concepts (organisers) which can facilitate the learning and retention of unfamiliar but meaningful verbal material”. The idea of advance organizer was introduced by Ausubel in the following two cases:
a) When the student does not process proper subsumers, e.g. when the material is totally new and the learner does not have relevant information to which they can relate the new material.
b) When the student obtains appropriate subsuming information, but the information is insufficiently developed and is not likely to be identified and linked to the new information.


The existing components of the knowledge structure to which new learning needed to be correlated; subsequently they become recognized also as ‘anchoring’ ideas or concepts. At this point Ausubel’s idea of subsumers is similar to Bruner’s view of readiness. So, if the subsumers are there the student is efficiently ready. Orton (2004) stated that meaningful learning implies an understanding of constraints. He also added that any theory of learning mathematics should take into account the hierarchical nature of the subject, and there should not be many occasions when new knowledge cannot be linked to existing knowledge. For instance, it is not possible to learn about integers and about rational numbers unless the natural numbers are understood meaningfully .

Discovery Learning

Discovery learning is a method of inquiry based on the learner rather than a teacher oriented view of the teaching and learning processes, and it is advocated in the work of Piaget (1972), Bruner (1966), and Papert (1980). It requires learners to proceed in the same way as scientists when investigating nature, using processes that are very similar to the processes of scientific discovery (Klahr & Dunbar, 1988; Jong & Njoo, 1993; Joolingen & Jong, 1997). The central purpose of discovery learning is that the learner obtains or constructs knowledge by performing experiments. 


Discovery learning is considered to be a promising way of learning for a number of reasons, the main being that the active involvement of the learner with the field would result in a better knowledge base (Joolingen, 1999). Unfortunately, many students are still taught largely by exposition and are given little opportunity to learn by discovering (Orton, 2004). Papert concentrated on the impact of new technology on learning, and he “gained his enthusiasm for active, discovery-type learning environments directly from Piaget, with whom he worked for five years” (Orton, 2004, p: 97). Papert expressed his belief that enriching the learning environment through the use of materials was more important than Piaget had suggested (ibid). He created Logo which is a computer programming language as a tool to improve the way that children may think and solve problems in mathematics. A mobile electronic ‘Logo turtle’ was developed and children were encouraged to solve problems and trace out shapes on a classroom floor. Papert (1980) argued that the usual mathematics curriculum was meaningless to most children, but the invention of Logo provided for them an opportunity to construct knowledge in meaningful way. However, Orton (2004) stated that there are some objections on the grounds that Logo is too difficult and it takes too much time, and that subsequently using Logo in ordinary classrooms has “convinced many teachers that pupils cannot work entirely on their own in the way Papert seems to suggest is both possible and desirable, and that skillful teacher mediation between the children and the software is needed”.


In teaching mathematics, words such as ‘discovery’, ‘investigation’, ‘activity’ and ‘problem-solving’ are famous and have became part of the mathematical language (Orton, 2004). At the present time, there is much awareness about the importance of the use of discovery learning in mathematics classes, especially at primary level. With older students, discovery learning might sometimes be a suitable method, but it is very rarely used. In practice, it is very hard to apply discovery learning in higher educational level and as Ausubel states what was created over the past four centuries cannot be rediscovered by our students in ten or fifteen years. Thus, meaningful verbal learning which provides expository teaching can be effective and in some ways better than other methods.

The van Hiele Theory of learning geometry

The van Hiele theory of learning geometry also drew heavily from the work of Piaget (1972) and Gestalt theory (Wertheimer, 1912). After studying Piaget’s work, Pierre van Hiele and Dina van Hiele-Geldof thought that students’ geometrical competence might well improve by progressing over a period of time through successive stages of thinking. There are five levels in van Hiele theory of learning geometry as follows (Orton, 2004):


Level (1) Visualization:

Students can recognize figures as whole entities (triangles, squares), but do not recognize properties of these shapes (right angles in a square); visual impression and appearance exert a storing influence, thus a square cannot also be a rectangle; drawings of shapes are based on holistic impressions and not on component parts; names may be invented for shapes according to their appearance, for example, ‘Slanty rectangle’ for parallelogram.

Level (2) Analysis:

Students can analyse figures’ components such as sides and angles but cannot relate between figures and properties logically; properties and rules of a class of shapes may be discovered empirically (for example, by folding, measuring, or by using a grid or diagram); a figure can be identified from its properties; generalizations become possible, for example, all squares have four sides, the angles of triangles total 180º.

Level (3) Informal Deduction:

Students can establish interrelationship of properties within shapes and can make simple deductions, though the intrinsic meaning of deduction is not understood; a shape may be used to establish that a square is a rectangle; a statement cannot be separated from its converse.

Level (4) Deduction:

At this level, students can appreciate the need for definitions and assumptions, and can present proofs within a postulational system; The interrelationship and role of undefined terms, axioms, definitions, theorems and formal proof can be understood; proof as the final authority is accepted; inter-relationships among networks of theorems can be established.

Level (5) Rigor:

Students at this level can work abstractly and can compare systems, can examine the consistency and independence of axioms and generalize a principle or theorem to find the broadest context. Geometry is seen in the abstract with a high degree of rigor, even without concrete examples.


All school geometry courses are taught at Level 3 . The van Hieles also recognized some features of their model, including the fact that a person must proceed through the levels in order. The most important features of van Hieles system as summarized by Fuys et.al (1988) are as follows: (a) the levels are consecutive; (b) each level has its own vocabulary, symbols and network of relations; (c) what is implicit at one level becomes explicit at the next level; (d) material taught to students above their level is dependent on reduction of level; (e) progress from one level to the next is more subject to instructional experience than on age or maturation; and (f) one goes through various ‘phase’ in arising from one level to the next.

Monday, 2 September 2013

Relations Between Conceptual and Procedural Knowledge

Throughout this century, instructional reform has oscillated between emphasizing mastering of facts and procedures on the one hand and emphasizing understanding of concepts on the other (Hiebert & Lefevre, 1986). Few today would argue that either type of mathematical knowledge should be taught to the exclusion of the other. Much less agreement exists, however, concerning the balance between the two that should be pursued or concerning how to design instruction that will inculcate both types of knowledge. Multidigit addition and subtraction has proved to be an especially fruitful domain for studying the relations between conceptual and procedural knowledge. Children spend several years learning multidigit arithmetic. They must learn the carrying procedure for addition and the borrowing procedure for subtraction. Understanding these procedures requires understanding of the concept of place, that each position in a multidigit number represents a successively higher power of ten. It also requires understanding that a multidigit number can be represented in different ways, for example, 23 can be represented as 1 "10" and 13 "1's".


Many children have difficulty understanding place value, and, as noted earlier, they frequently use buggy procedures that reflect this lack of understanding. For example, second-graders often do not correctly carry when adding multidigit numbers (Fuson & Briars, 1990). Instead, they either write the two-digit sums beneath each column of single-digit addends (e. g., 568+778=121316) or ignore the carried values (e. g.,
568+778=1236). Although there are exceptions, procedural skill and conceptual understanding usually are highly correlated. One source of evidence for this view is cross-national studies. For example, comparisons of Korean and American elementary school children have revealed parallel national differences in conceptual and procedural knowledge of multidigit addition and subtraction. Fuson and Kwon (1992) asked Korean second and third graders to solve two and three digit addition and subtraction problems that require carrying or regrouping. Then the children were presented several measures of conceptual understanding: ability to identify correctly and incorrectly worked out addition and subtraction problems, to explain the basis of the correct procedure, and to indicate the place value of digits within a number. Almost all of the Korean children used correct procedures to solve the problems and also succeeded on all of the measures of conceptual understanding. Stevenson and Stigler (1992) reported similar procedural and conceptual competence in first through fifth graders in Japan and China.


On the other hand, a number of studies reviewed in Fuson (1990) indicated that American children, ranging from second to fifth grade, frequently lack both conceptual and procedural knowledge of multidigit addition and subtraction. Lack of conceptual understanding was evident in findings that almost half of third graders incorrectly identified the place value of digits within multidigit numbers (Kouba, Carpenter & Swafford, 1989; Labinowicz, 1985), and in findings that most second through fifth graders could not demonstrate or explain ten-for-one trading with concrete representations (Ross, 1986). Lack of procedural knowledge was evident in findings that children of these ages frequently erred while using paper and pencil to solve multidigit addition and subtraction problems (Brown & Burton, 1978; Fuson & Briars, 1990; Kouba et al., 1989; Labinowicz, 1985; Stevenson and Stigler, 1992).


 Taken together, these results suggest that conceptual and procedural knowledge are related; Asian children have both, and American children lack both. Within the U. S., conceptual and procedural competence are also highly correlated. Second and third graders who correctly execute the subtraction borrowing procedure also are more accurate in detecting conceptual flaws in a puppet’s subtraction procedures than are children who do not execute the subtraction algorithm consistently corrrectly (Cauley, 1988). Conceptual understanding of multidigit addition and subtraction and the ability to invent effective computational procedures are also positively correlated in first through fourth graders (Hiebert & Wearne, 1996). This correlation leaves open the possibility that conceptual understanding could be causally related to children inventing adequate computational procedures, but also the possibility that knowing the correct procedure could be causally related to increased conceptual understanding (by allowing children to reflect on why the correct procedure is correct.) One relevant source of evidence is examination of the order in which individual children gain procedural and conceptual competence. It turns out that a substantial percentage of children first gain conceptual understanding and then procedural competence, but that another substantial percentage do the opposite (Hiebert & Wearne, 1996).


Studies aimed at improving teaching of multidigit addition and subtraction typically emphasize linking steps in the procedures to the concepts that support them. In general, these teaching techniques successfully increase both conceptual and procedural knowledge. Although not currently conclusive, they suggest that instruction that emphasizes conceptual understanding as well as procedural skill is more effective in building both kinds of competence than instruction that only focuses on procedural skill (Fuson & Briars, 1990; Hiebert & Wearne, 1996). A question that remains, however, is which type of knowledge should be emphasized first. Many opinions have been offered on this topic, but until recently, no directly-relevant experimental evidence was available. A recent study by Rittle- Johnson and Alibali (in press), however, provides such evidence. They examined fifth graders’ performance on mathematical equality problems of the form a+b+c=_____+c. Some randomly-selected children were presented conceptually-oriented instruction, other children were presented procedurally-oriented instruction, and yet others were presented neither. Then all children were presented practice solving problems, followed by a posttest that tested both conceptual and procedural knowledge. The conceptually-oriented instruction produced substantial gains in both kinds of knowledge; the procedurally-oriented instruction produced substantial gains in procedural knowledge and smaller gains in conceptual knowledge. To the degree that this result proves general, it suggests that conceptual instruction should be undertaken before instruction aimed at teaching procedures.


Cognitive Variability and Strategy Choice

Children's thinking has often been described as something like a staircase, in which children first use one approach to solve problems, then adopt a more advanced approach, and later adopt a yet more advanced approach. For example, students of  children's basic arithmetic (e.g., Ashcraft, 1987) have proposed that when children start school, they add by counting from one; sometime during first grade, they switch to adding by counting from the larger addend; and by third or fourth grade, they add by retrieving the answers to problems. More recent studies, however, have shown that children's thinking is far more variable than such staircase models suggest. Rather than adding by using the same strategy all of the time, children use a variety of strategies from early in learning, and continue to use both less and more advanced approaches for periods of many years. Thus, even early in first grade, the same child, given the same problem, will sometimes count from one, sometimes, count from the larger addend, and sometimes retrieve the answer. Even when children master strategies that are both faster and more accurate, they continue to use older strategies that are slower
and less accurate as well. This is true not just with young children, but with pre-adolescents, adolescents, and even adults (Kuhn, Garcia-Mila, Zohar, & Anderson, 1995; Schauble, 1996).


This cognitive variability is a spontaneous feature of children's thinking. Efforts to change it do not usually meet with much success. For example, in one study, first-to-third-grade teachers were interviewed regarding their beliefs about their students' arithmetic strategies and their evaluation of whether the students' use of multiple strategies was a good thing (Siegler, 1984). All of the teachers recognized that the children used multiple strategies, though most viewed this as a bad thing. One teacher said that she was constantly discouraging students from using strategies such as counting on their fingers. When asked how often she had done this with the pupil in her class who did it the most often, she asked, "How many days have there been in the school year so far?" This teacher and others recognized that even when they explicitly told students not to use their fingers, they did anyway, even if they had to do it by putting their fingers in their lap, under their leg, or behind their back. There is a certain logic that supports the teacher’s view. Older students and those who are better at math don't use their fingers, whereas younger and less-apt students do. One goal of  education is to make younger and less apt students more like older and more apt ones. Therefore, children who use their fingers should be discouraged from doing so.


However, children actually learn better when they are allowed to choose the strategy that they wish to use. Immature strategies generally drop out naturally when students have enough knowledge to answer accurately without them. Even basic strategies such as counting fingers allow students to generate correct answers when forbidding use of the strategies would lead to many errors. Further, students who use a greater variety of different strategies for solving problems also tend to learn better subsequently (Alibali & Goldin-Meadow, 1993; Chi et al., 1994; Siegler, 1995). This is in part because the greater variety leads the students to be able to cope with whatever kinds of problems they encounter, rather than just being able to cope with a narrow range. Allowing children to use the varied strategies that they generate, and helping them understand why superficially-different strategies converge on the right answer and why superficially reasonable strategies are incorrect seems likely to build deeper understanding (Siegler, forthcoming).


Children’s use of diverse strategies makes it essential that they choose appropriately among the strategies. To choose appropriately, they must adjust both to situational variables and to differences among problems. Situational variables include time limits, instructions, and the importance of the task. For example, in a magic-minute exercise, it is adaptive for children to state answers quickly, even if they aren't absolutely sure of them. Similarly, if it's very important to be correct in the particular situation, then checking the correctness of
answers becomes more worthwhile. At least from second grade onward, children shift their choices appropriately to adapt to such situational variations.


Adaptive choice also involves adjusting strategy use to the characteristics of particular problems. When children are faced with a simple problem, it often is ideal for them to use a strategy that can be executed quickly, because it will be sufficient to solve the problem. In contrast, when faced with a more difficult problem, they may need to adopt a more time consuming and effortful strategy to generate the correct answer. Adaptive choice involves using quick and easy strategies when they are sufficient, and using increasingly effortful ones when they are necessary to be correct. Research on strategy choice has also revealed some surprising similarities in the performance of children from different socio-economic groups. Children from low-income backgrounds, particularly low-income African-American backgrounds, are often depicted as choosing strategies unwisely. Suggestions have been made that instruction focus on improving their metacognition and their strategy selection. However, selection of appropriate strategies does not seem to be their main problem, at least in the context of arithmetic. Their strategy choices are just as systematic and just as sensitive to problem characteristics as those of children from middle income backgrounds (Kerkman & Siegler, 1993). 


Instead, their problem seems to be that they do not possess adequate factual knowledge. This in turn seems to be due to less practice in solving problems, and to less good execution of strategies, rather than to any high level deficiency in their thinking. The findings indicate that greater practice and instruction in how to execute strategies may be the most useful approach to improving their arithmetic skills.

Phonological Loop and Mathematics

Many studies have investigated the relationship between the multi-components of working memory and learning mathematics. Researchers have found significant correlations between mental arithmetic in children and the phonologic loop (Adams & Hitch, 1997; Towse & Houston, 2001; Jarvis & Gathercole, 2003). The role of the phonological loop also has been explored in adult counting (Logie & Baddeley, 1987) and arithmetic (Ashcraft et al., 1992; Logie et al., 1994; Lemaire et al., 1996). Holmes and Adams (2006) indicated the phonological loop is thought to be important for the attainment of number facts in early childhood. Learned number facts form complete networks in long term memory between mental arithmetic s and the phonological loop. However, the association was no longer significant in an adolescent population (Reuhkala, 2001).


The teaching and learning processes take place through the medium of language. Mathematics is considered to be the language of science, and as Baroody and Standifer (1993) indicate, “For children, Mathematics is essentially a second or foreign language.” The translation of ordinary language in mathematics into the symbolic language creates a conflict of exactitude which leads to overload of the working memory. The usage of common vocabularies in mathematics causes another language problem, because their meanings in the mathematics context differ from the normal English usages. Gardner (1972) examined the accessibility of words to students at a number of levels in secondary school by testing commonly used words in science and the results showed:
• “Pupils lacked precision in their use of words as they moved from context to context.
• Students were easily confused by words (which) ‘sound alike’ or ‘look alike’.
• In a significant number of cases, students chose meaning exactly opposite to the accepted meaning.
• There was an improvement in performance with age.”


Durkin and Shire (1991) demonstrate that one of the feature in mathematics is that the meanings to convey them are often endowed with other meanings, which may be more familiar to children in every day language and the vocabulary of mathematics includes many words which have multiple meanings and there is evidence that students often fail to interpret the words as teachers intend them. Cassells and Johnstone (1983) have emphasised the great problems, which are caused when normal words are used with precise meanings. Macnab and Cummine (1986) specify words such root, solution, product, matrix, differentiate, integrate, function, coordinate, prime, factor, multiply, power, index, whose use in the mathematics context cause difficulties of the semantic confusion involved. Durkin and Shire (1991) also listed a table of individual common ambiguous words (multi-meaning), used in mathematics and they discuss the meaning of each word is likely to have for child before he or she encounters it in a mathematical context at school . In this case, the mathematical meaning of the words and holding information occupy the student’s working memory space. Therefore, there is no empty room or space for manipulating this information.


IDENTIFYING COGNITIVE TRAITS AND LEARNING STYLES

To incorporate cognitive traits and/or learning styles in educational systems, information about cognitive traits and learning styles need to be first collected. One approach is to let students perform comprehensive tests or questionnaires to find out the cognitive traits or learning styles. Such an approach; however, has potential to suffer from the biases and indecisiveness of the learners. A more meaningful approach is to track the students’ behaviour and infer the required information from their behaviour. Cognitive Trait Model (Kinshuk & Lin, 2004; Lin & Kinshuk, 2005) uses this approach to profile learners according to their cognitive traits. For the identification of learning styles, approaches for detecting the dimensions of Felder-Silverman learning style model are introduced.


Identification of Cognitive Traits

Cognitive Trait Model (CTM) is a student model that profiles learners according to their cognitive traits. Four cognitive traits, working memory capacity, inductive reasoning ability, processing speed, and associative learning skills are included in CTM so far. The CTM offers the role of ‘learning companion’, which can be consulted by and interacted with different learning environments about a particular learner. The CTM can still be valid after a long period of time due to the more or less persistent nature of cognitive traits of human beings (Deary, 2004). When a student encounters a new learning environment, the learning environment can directly use the CTM of the particular student, and does not need to “re-learn the student”.The identification of the cognitive traits is based on the behaviour of learners in the system. Various patterns, called Manifests of Traits (MOT), are defined for each cognitive trait. Each MOT is a piece of an interaction pattern that manifests a learner’s characteristics. A neural network (Lin & Kinshuk, 2004) is responsible for calculating the cognitive traits of the learners based on the information of the MOTs.


Identification of Learning Styles

There are a number of adaptive systems available in the literature incorporating learning styles. For example, CS383 (Carver, Howard, & Lane, 1999) was the first adaptive hypermedia system that was based on Felder-Silverman learning style model (FSLSM). The course conducted in the system included comprehensive collection of media objects. The system offered students the option to order these objects in accordance with how well they fit to the learning style of the student. Also MAS-PLANG (Peña, Marzo, & de la Rosa, 2002), a multi-agent system which has been developed to enrich the intelligent tutoring system USD (Fabregat, Marzo, & Peña, 2000) with adaptivity with respect to learning styles is based on FSLSM. Another example is INSPIRE (Papanikolaou & Grigoriadou, 2003) that is based on Honey and Mumford’s learning style theory. In all these systems and in most other systems which incorporate learning styles, the learning style is identified based on a questionnaire that needs to be filled out by learners before using the system. These questionnaires are based on the assumption that learners are aware of how they learn. Jonassen and Grabowski (1993, p. 234) pointed out that “because learning styles are based on self-reported
measures, rather than ability tests, validity is one of their most significant problems”. García, Amandi, Schiaffino, and Campo (2006) studied the use of Bayesian networks (Jensen, 1996) to detect students’ learning styles based on their behaviour in the educational system SAVER. Based on the Felder-Silverman
learning style model, they determined patterns of behavior, which are representative for the respective dimensions, as well as the different states these variables/patterns can take. Because SAVER does not incorporate the visual/verbal dimension, this dimension is left out from investigations. 


While the above approaches are developed for specific systems, Graf and Kinshuk (2006) proposed an approach to detect learning styles in learning management systems (LMS) in general. Equal to the approach by Garcia et al. Felder-Silverman learning style model is used as the basis but in this case all four dimensions are incorporated. The patterns of behaviour are derived from commonly used features in LMS such as forums and exercises. Regarding the calculation of the learning styles, the approach used in the Index of Learning Styles (Felder & Soloman, 1997), a questionnaire for identifying the learning style according to Felder-Silverman learning style model, is applied.