Learning By Doing

European Journal of Engineering Education Vol. 30, No. 1, March 2005, 105–119

‘Learning by doing’: a teaching method for active learning in scientific graduate education LUDOVIC BOT†*, POL-BERNARD GOSSIAUX‡, CARL-PHILIPPE RAUCH§ and SAFOUANA TABIOU§ †Department of social sciences, ENSIETA, 2 rue François Verny, F-29806 Brest, cedex 09, France ‡SUBATECH (Department of Nuclear Physics), Ecole des Mines de Nantes, France §Centre for the Enhancement of Engineering Education, Ecole des Mines de Nantes, France (Received 21 December 2002; accepted 26 August 2004) This article describes an active learning method for the teaching of physical sciences and mathematics to engineers. After defining the challenges involved in the training of engineers, we shall describe the answers provided by our method, ‘learning by doing’ (named ‘Apprentissage Par l’Action’ in French), by introducing four key points: real-life simulation, the management of non-success, the result requirement and the different roles of the teachers. An assessment of this experience is carried out which emphasizes the factors paramount in the success of this pedagogical innovation. Similarities between our experience and other well-known methods such as problem-based learning, problem solving and, more generally, the concept of learning by doing coined by John Dewey in his philosophy of education, are mentioned. Keywords: Active learning; ‘Learning by doing’; Scientific graduate education; Mathematics; Physics; Multidisciplinary approach; Professionalization of engineers

1.

Introduction

It has been some years now since the Ecole des Mines de Nantes School of Engineering (EMN) acquired original pedagogic experience for training engineers with APA, or ‘Apprentisage Par l’Action’.1 This method has been used in modern language teaching, in in-company training, and in the teaching of physical sciences and mathematics. This article deals with the last two areas, given that the authors have specialized in them, and also because, as regards the pedagogics, they are the areas in which APA is seemingly most cutting-edge. EMN is an engineering school that selects its students through an entrance exam after a science-focused preparation of 1 year following the Baccalauréat (Higher Leaving Certificate). Students come into the school between 19 and 20 years of age and follow a 4-year programme until they pass their engineering degree (Master’s level). The students come from the French education system, which is selective over short periods (final year in high school and preparatory school) and mainly focuses on theoretical teaching. *Corresponding author. Email: [email protected]

European Journal of Engineering Education ISSN 0304-3797 print/ISSN 1469-5898 online © 2005 Taylor & Francis Ltd http://www.tandf.co.uk/journals DOI: 10.1080/03043790512331313868

106

L. Bot et al.

The EMN’s programme seeks to train executives who are functional within industry, and it is based on general scientific and technical training in industrial systems engineering. After the course, few students go in for research or design. Even if its pedagogic and industrial legitimacy still hinges on its research departments, the EMN has also allotted considerable resources to pedagogics. The relative newness of the school (the first class graduated in 1996) and the absence of internal tradition have made it possible to conceive of a programme that breaks with the frequently over-impersonal vision of traditional programmes. The numbers for the induction year stand at around 130 engineering students. Independently of the initial testing in languages, physics teachers were asked to look into new ways of teaching their subject based on a method called ZAP!,2 which was developed in the USA. The ZAP method had originally been imported from CALTECH in 1996 to be offered to year 1 EMN students. Given the teething problems, ZAP was adapted in 1998 to suit French engineer training and was called APA. Applying this type of pedagogics to the teaching of formal sciences was a challenge that the EMN and its teachers rose to as from 1999, with the setting-up of a mathematics module. Both these ‘APA-physics’ and ‘APA-maths’ experiments involved 20 teachers and researchers and made up 20% of the EMN year 1 teaching programme.

2. What are the pedagogic challenges facing the teaching of sciences? Firstly, it is useful to explain some of the pedagogic challenges that sciences encounter in today’s society. The challenges comprise three levels: challenges for society in general, challenges surrounding the professionalization of engineers and the more strictly scientific challenges. 2.1

Challenges for society in general

This first challenge is upstream of teaching and consists of reacting positively to the disaffection of young people with scientific syllabuses. This has been proven over significant periods of time and concerns all western countries (Ourisson 2002). The current generation of students has been greatly influenced by ‘virtual’ technologies from a tender age. Despite the fact that this adaptability and receptiveness to innovation are extremely positive features in these generations, the notions of intellectual effort and patience have become less credible. However, both these values are prerequisites for the scientific strategies that link theoretical reasoning to actual experimentation. A ‘dose’ of reality might make it possible to get them back on track. This objective does not just come from a scientific and/or technical challenge. It is also a challenge to society so that its members learn to be responsible. For certain youngsters, this never-ending presence of non-structured data, virtual reconstructions and the blurring of the different levels of representation makes it difficult for them to build up their inner natures or their relationships with reality, and this includes their social dimensions. The challenge of intellectual autonomy has consequently become more urgent today than for previous generations. In fact, this entails training through the sciences, not only for the sciences. This was the aim of the US experiments that inspired APA, whose initial findings were social ones. APA is therefore not a counter proposal to social trends. A repressive or normative vision of teaching sciences would destroy certain values of emancipation, creativity and freedom that scientific activity contains, including its technical dimensions related to innovation and the creation of wealth. The challenge is to make the learning of sciences a source of personal development and to revive the pleasure that young people have in using them. Freedom is the cornerstone, provided it is matched by responsibility in the face of results.

‘Learning by doing’

2.2

107

The challenges for the professionalization of engineers

Another challenge is directly linked to the French context in which engineers are trained. Preparatory classes and school entrance exams have endowed the scientific disciplines with a selective role. Students must focus their learning on exercise standards that must be reproduced within a given time in order to pass extremely formal examinations (the French Baccalauréat and graduate school entrance exams). As regards mathematics, educational studies (Castela 2002) have demonstrated that the best students from these years have acquired solid ‘scientific understanding’but often lack ‘scientific-mindedness’. For the majority of them, knowledge is acquired through repetition, not by any investigative aspiration for autonomy. In fact, science does not function by repetition or ‘transmission’. The development of experimental techniques and new knowledge always go hand in hand. It is seemingly impossible to separate the results of speculation, culture or knowledge from what is due to pragmatism, endeavour, or skills. Knowledge and know-how are inseparable in any innovative scientific and technical process. The aim is more to reposition this complementarity within the teaching of science. If science is taught from this perspective, it becomes a particularly efficient tool for developing skills in innovation, autonomy, self-learning and creativity. These days such qualities are recognized by a large number of people in charge of human resources (HR) and economic policies as key factors in the output of added value in western economies (Crozier 1989, Jarrosson 1991, Genelot 1992, Milton and Stinson 1995, Hunot 2000, Goldratt and Cox 2002). It should be noted that these same people have never defined any scientific programme reflecting the needs that they can identify in industry. It is improbable that such a deduction will be contemplated. For long-term profitability, executive training must at least partly avoid the changing trends in management. While including the already mentioned social challenges in their ideas, HR people also stress the fact that executives will have to develop in a very diversified and rather unpredictable environment.

2.3

The challenges for scientific development

Modern science is suffering from a compartmentalization of disciplines. Students must learn to resolve ‘real’ problems that do not necessarily comply to a subject-by-subject classification. Such problems are complex ones in that they contain inseparable knowledge and know-how. This need to incorporate skills in order to meet a ‘concrete’ challenge that can directly concern them is approached; putting the students in front of a real-life simulation, a starting point for all our ‘action’ training sessions. These simulations are not carried through by churning out pre-acquired knowledge or skills. They require ‘ad hoc’ adaptation, reasoning and learning which must be validated by real experiments performed within the students’ means. Within this approach, called reasoned scientific investigation, the student is expected to formulate his own hypothesis and then to design experimental devices or deductive reasoning in order to test them. At the same time, with the experience of such well-framed and time-tested disciplines such as physical science and mathematics as a basis, the teaching seemingly cannot fail to formulate clear expectations and a result requirement. When applied to science, active learning methods require longer and necessarily more directions than those generally provided by the constructivist visions of knowledge. Associated with the terms of a situation, daily experience, the personal experiences of the students or teamwork emulation are relevant input for teaching management and corporate life (Hunot 2000). The objects of knowledge that physical science and mathematics seek to attain are less directly accessible.

108

L. Bot et al.

The scientific teaching that we are contemplating here thus meets the expectations situated ‘downstream’ of the training, i.e. in the firms. It views seriously their needs, identified in terms of ‘frames of mind’ and ‘adaptability’. However, it also endows itself with the means of minimum resistance so that the minds of future executives can be shaped to go beyond the vicissitudes of modern economies. Our pedagogics also considers the ‘upstream’ expectations of the training, namely, putting young people in front of the need to forge themselves a lasting professional future, but who are hardly receptive to the arguments of the powers that be, or tempted by an intellectual effort without the accompanying personal development. Finally, with a view to decompartmentalizing the disciplines in order to query openly the objects, the ambition that is followed is authentically a scientific one.

3. The elements of the APA method to meet the challenge The primary objective is to put the student in a setting where he really uses his freedom and his responsibility, therefore to put him face-to-face with the object of knowledge before he faces the knowledge itself. The four pedagogical aspects shown to be significant in the conception of APA sessions are: real life simulation; the management of non-success; the result requirement; and the monitoring of students’ progress. These aspects imply different roles for the teachers. The implementation of APA is performed in front of the students following four learning phases: the technical, inductive, deductive and applicative phases. Despite the fact that different terms are used for them, these elements are present in most active methods. As the text progresses, we shall point out similarities identified with problem-based learning (PBL) and the problem-solving. To our knowledge, PBL is the most formalized and prevalent of active methods. A PBL experience, close to APA in the fact that it concerns mathematics in another engineer training, is described in Lecomte (2002). 3.1 Real-life simulation APA functions by confronting the students with a problem, or a situation in the larger sense, which they must react to. The situation is prepared and written by teachers in relation to their teaching goals. Given the many works on active learning methods, we call this initial concept a ‘situation-problem’, which means this is not simply a ‘problem’, as the term could have been used in the academic sense, but neither does it imply a simple gratuitous situation. The situation should be problematic and urge the students to react. The reaction required from the students assumes new knowledge and know-how on their behalf (this is the objective of the teaching), and it is generally useful that the situation-problem is both inciting and has an aspect of play. For the teaching of physical sciences, the construction in 1 year of a complete measuring apparatus prompts several situation-problems: electrical supply, transformation of currents and voltage, as well as gauging (Rauch 2001). In order to teach mathematics, the introduction to the concept of numerical series is carried out by a game of construction. A question is asked that is simple, even for the layman. After various construction tests, it is easy to observe that certain intuitive replies are erroneous. The students are then asked to perform measurements on their constructions and to put forward laws by recursive reasoning. When the modelling phase is achieved they are confronted by the harmonic series. They do not know that this object is infinite and they can only understand it by using their previous knowledge on sequences.3 When he is first up against the situation-problem, the student is entitled to waste time. This awareness-forming phase is difficult to manage as it is situated between two contradictory

‘Learning by doing’

109

focal points. On the one hand, students do not have the same faculties in order to grasp a situation and connect its problematic features with their previous knowledge. On the other hand, the simulation phase must not be followed by a phase of distraction because this would prevent the student from using his initial intuition. The simulation must be extended through both the formulating of hypotheses and the finalizing of experimental strategies to validate or invalidate these hypotheses. Despite its initially naïve nature, the experimentation must be active and concrete because this situation-problem must be surmounted and the initial pitfalls that it really represents identified, all of which requires learning. The situation-problem goes beyond a simple disciplinary classification. It is not a ‘mathematical problem’ and it is not a historical ‘physical experiment’ that is redone to illustrate a lesson. In fact, it is an outside ‘reality’, a concrete challenge, a game of logic and a surmountable challenge for each and everyone, or a situation that all students, having reached a given stage of their programmes, can understand and come to grips with. Of course, the situationproblem has been designed by teachers in relation to their teaching goals and with the help of their in-depth subject skills. Yet this must be introduced to the students as much as possible as a uniquely intrinsic need. Certain students, of course, do not relate to the session. However, for those who are willing enough to give it the necessary attention, the situation-problem must be as self-supporting as possible. It is important that the teacher organizing these APA sessions keeps away from any notion that a ‘correction’ or ‘solution’ is to be provided for the situation-problem. The natural reaction of a teacher is to become the centre of conversation again once he realizes the students have become aware of the situation-problem. This is both easily accepted and requested by the students, but such a procedure actually means transmitting an illusion of freedom to them during their learning. It is important that at the end of the simulation, each student experiences the difficulties pertaining to true freedom along with its contingencies and its occasional sources of dissatisfaction. Unaided, he must relate this freedom to his responsibility for assuming the consequences of his hypotheses. This does not mean that in the end the situation-problem has no ‘solution’. The existence of such a solution is in fact often a prerequisite for the session planner so that the teaching goals can be both defined and attained.

3.2

Management of non-success

Consequently, the students’ entitlement to error must be included the moment the teaching is designed and measured. Such concern is not just a simple matter of time. Given that the situation-problems require new learning, it is actually not recommended that the student does not commit at least one error. New learning often derives from moments when difficulties are encountered. There are countless cases of discoveries or scientific inventions being made due to errors or changes of direction in relation to some prescribed programme. It is therefore necessary that situations-problems are not solved by over-simplistic strategies or initial intuitions that are subsequently not disproved. Failures and their ensuing reappraisal also facilitate mental assimilation. This non-success must be ‘managed’ as well as possible so as not to jeopardize the actual sessions or the individual tasks required within deadlines that are both reasonable and in keeping with the rest of the programme. During the sessions, the supervisor must make sure that the students do not waste too much time on matters of principle or technique, which will be solved only when they move on to the experiments. Moreover, additional aspects concerning the situation-problem must not be allowed to distract the students and lead them away from the specified result requirement. The paramount role of non-success in APA means the aims

110

L. Bot et al.

of each session are scaled down. The learning must remain focused on basics, not through erudite teaching, for example, with its plethora of anecdotes. Students’ entitlement to error means a substantial amount of individual work and APA explicitly requires this. Error or hesitation cannot be planned. The starting point for freedom is in managing one’s time on one’s own. This said, it is important to have some idea about the typical amount of time required by students in order for them to solve a situation-problem. In the case of an engineering school, the teaching is directed towards individual projects and active learning methods so individual working time must not be overloaded in relation to the in-house training set out in the students’schedules. As a result, the measurement of the teaching includes a typical amount allocated to individual work while satisfying two contradictory trends: not to overly formalize work that is to be carried out on one’s own, in order to respect the differences between individuals, but also to keep within reason when the task is defined so as not to lessen their motivation.

3.3 Result requirement and student assessment The texts drafted for APA require students actually to produce an apparatus or measurement in physical science or an accurate response to resolve a mathematical problem. These result requirements correspond to objectives set by the situation-problem designer. They are formulated a minima. They are essential for the continuation of the programme and only they can be assessed. Besides these a minima objectives, each APA session refers to the acquisition of related knowledge without seeking any particular or single path. As much as possible must be done so that the student freely takes this path according to his means, his perception of the situation-problem and to his previous knowledge. Initially, the teacher cannot formalize through objectives, given that this part belongs to the creativity and freedom of the learner; nor can the teacher subsequently correct it. Such a ‘correction’ would have two consequences that do not meet APA objectives. Firstly, it would give the student a poor idea of his progress and may cause him to lose confidence in himself, and confidence is essential if he is to be creative. Secondly, this would make the student dependent on one strategy chosen for one type of problem and so weaken his autonomy when he will have to develop new strategies in the future. As a result, it is difficult to assess related learning that the students experience in order to obtain the necessary result. This related learning, such as the ability to reason, to check one’s work, the acceptance of one’s own errors and the wish to rectify them, the management of one’s work time, tailoring one’s behaviour to a situation, teamwork, creativity and so forth, are obviously all important and the teacher will try to develop these attributes in each student. Needless to say, they are extremely useful in the professional context that the student will later experience and they are much discussed topics for educational scientists dealing with executive training (Hunot 2000, Lemoult and Tellier 2003). However, this is a matter of final objectives (or rather ‘meta-objectives’), not points of departure. They require indirect strategies as well as the mediation of tangible contents. For the sake of objectivity, the result requirement only concerns contents, whereas the ‘meta-objectives’ are approached via implicit dimensions that every simulation contains. The assumption is that the student must get through the implicit dimensions of a given situation in his own way, in order to attain the explicitly specified result requirement. It is equally important to dissociate learning phases from assessment phases in order to avoid more or less dissimulated assessment sessions occurring on a permanent basis. Experience proves that too much time given over to assessment distorts the inspirations that underlie APA, i.e. creativity and the pleasure in doing sciences. The assessment must keep its human side

‘Learning by doing’

111

and not try to be over rigorous. It will merely mirror the development of the teaching with its contingencies and the social relations involving the students, the teachers and the general institutional, academic and economic surroundings. To avoid permanent assessment, APA specifically devotes certain sessions to assessment, and these are organized in conjunction with teaching procedures (teamwork, experimentation, respecting students’ acknowledged freedom and not a proctored test). In the large majority of cases, the students who have not got used to working spontaneously during the training sessions visibly encounter difficulties in the assessment sessions. 3.4

Monitoring the students’ progress

As the sessions unfold, so too the students obtain a return on their efforts which enables them to make progress. However, this return is not actually an assessment. It is done in a log in which each student or group records his or their progress. They are required to write down their responses and results, but also their difficulties and their aborted ideas. After each session, the supervisor makes notes in the log and uses several means of communication. The first two means are the simple ‘commentary’ and ‘advice’ to students (‘please realize that your strategy is more general than the case in hand’, ‘you could have also done this calculation or made this assembly’, ‘I suggest I ask you this question at a later date’, etc.) Both these means of communication are not necessarily followed up by the students. When students fail to comply with an element of the result requirement, a ‘demand’ is made. If the students do not reply to these ‘demands’ through extra work, this can affect the assessment (their work being considered by the end-of-year examining teams). When the supervisor considers that a student’s failure in a matter is too significant to surmount, he uses ‘correction’ mode to get the student back on course. The toing and froing with the monitoring log takes place in phase with the teaching. The supervisors are compelled to return the annotated logs within 24 h. If these conditions are both submitted to the students and complied with early in the year, the log becomes a working tool for extra dialogue further to the supervised sessions in which the students willingly express themselves. It is equally indispensable for the supervisor so that he can form an objective opinion on the real progress of each student. 3.5

Different roles for the teachers

Our experience has led us to distinguish several roles in APA-type teaching. They correspond to the different tasks that must be carried out for a module to be completed. They require different skills and it would probably be possible to imagine professional establishments endeavouring to make the greatest use of the skills and preferences of each teacher. At least concerning the first two, the roles in APA are similar to those identified by authors on active methods, especially on the PBL (Hunot 2000, de Graaff and Kolmos 2003). The first role concerns the conception of situation-problems. This probably contains the largest added value for the experienced teachers or researchers as they are themselves used to introducing new ideas given their activities in research. Their scientific culture enables them to recognize the existing links between disciplines and to shun the often school-like manner in which problems are addressed to students, for example, recalling that such and such calculus features in a mathematical exercise book actually derive from the modelling of a physical phenomenon. The levels of perception of or abstraction from the different disciplines and their intrinsic problems are frequently more intricate than described by the teaching programmes. This is the complexity confronting the engineer and which APA attempts to

112

L. Bot et al.

re-establish. The situation-problem designer can benefit from the complex relations among disciplines in order to transform a now commonplace problem into a productive simulation. Using his experience as a scientist, the situation-problem designer also appreciates the constraints involved in experimentation and the difficulties in validating hypotheses and reasoning through discriminatory experiences. He is capable of designing pedagogical situations to gauge reasonably the result requirement pertaining to a given situation-problem. Among these sometimes rich potentialities, he must identify or choose the required minimum and any underachievement of this must be solely due to the students’ own weaknesses. It is a delicate matter to inform students of this result requirement because it must also be as obvious as possible without introducing it as a foregone solution. The situation-problem designer must then have an idea of the paths and possible overruns that can be offered to the student so that not only can the designer choose among the different subjects, but he can also use this anticipation to ‘stage’ the situation-problem. In the PBL, this role corresponds to the ‘designer’. The second role concerns the monitoring of the sessions. Although APA favours individual work, it also uses time-sensitive sessions monitored by a teacher, and groups together 20–25 students. Such supervision enables the students to work more efficiently, either alone or in a group, and the teacher’s authority facilitates this. The teacher attempts to reply to the students’ questions while avoiding two possible pitfalls. Firstly, not to reply with a standard solution disconnected from the students’standpoint. The starting point should be the students’questions or certain angles that can be considered at a given stage of understanding without influencing what follows. It is therefore important to pay sufficient attention to the students in order to obtain a precise idea of their level of understanding and their progress. Another pitfall is to leave students to their own devices for too long so they do not progress and lose interest in the problem. Supervised sessions are an opportunity to get certain students back on course at little cost. These students often get bogged down in insignificant details of the situation-problem; this is caused by intellectual barriers and a lack of global vision. In science, one has to be wary of an overly constructivist vision of knowledge. Therefore, teaching should not just put a student up against a situation-problem on his own. It is also the role of the teacher monitoring these sessions to eradicate inhibitions, to encourage and guarantee a role for each student in the teamwork. Contrary to the traditional teacher who teaches, corrects or transmits, the supervisor acts as a guide more like an experienced co-learner than a transmitter of previously developed knowledge. The literature on active learning methods calls the supervisor ‘the facilitator’ or the mentor, and it is exactly applicable to our experience. The third and final role of the teacher is in the learning assessment. As regards the real ‘action’ sessions, the assessment focuses on result requirement specified by the designer of each session. Above all, one has to avoid the situation where students feel that the presence of the ‘facilitator’ is like the ‘eye of an assessor’ permanently appraising them. This primary level of assessment does not prevent other more traditional assessments if the teaching contains more traditional sessions with lessons and class work that can complete the action sessions in full pedagogical sequences and bears a relation to them. The other roles linked to the conception and performance of these traditional sessions are more academic and will not be described at this point. From our experience so far, the three roles appraised herein are assumed by the same teachers. They all participate in every APA stage. We do not have a fixed idea on the possibility of dissociating them, for example, to maximize teacher involvement by sharing out the tasks. An advantage of the current situation is that we have noticed that the contact with the students is all the better because the teaching team is both involved and coherent. It would also be strange to entrust the planning of the situation-problems to professionals who had no teaching experience with the relevant public. However, it is a real problem to separate totally the learning

‘Learning by doing’

113

phases and assessment phases in the event that they are performed by the same person. The assessor and ‘facilitator’ should possibly not be the same person and communicate through objectives via the result requirement linked to each situation problem. 3.6

The four phases of APA

APA makes it possible to clarify four phases in student learning. The time-keeping allocated to each phase and the structuring are probably the most universal elements in our feedback. These phases are actually four key moments in the construction of new knowledge and go far beyond our pedagogical experience in most cases. They are similar to the cycles of the problem-solving (Faucher and Martin 2000). • The first phase is technical. It involves drawing up a resource evaluation that is available prior to resolving a situation-problem. For example, it comprises measuring instruments, tools, material and measuring techniques that will probably be used. It also includes previous general knowledge and know-how as well as prior knowledge in the relevant discipline or in related disciplines. The technical phase means becoming aware that one is not totally helpless when meeting a new situation, both materially and intellectually speaking. It preludes the simulation. • The second phase is inductive. Faced with a situation-problem, hypotheses have to be made, new reasoning has to be induced and one requires intuition for possible solutions. The inductive phase is very important for creativity. Often a large place is given over to the pleasure of doing science. This comes straight after the simulation. • The third phase is deductive. For purposes of experimentation and validation, consequences have to be surmised from intuitions and hypotheses formulated during the inductive phase. This phase corresponds to the validation of experimental procedures in physics by effective measurements or to the acquisition of formal reasoning ensured in mathematics. The deductive phase is a very important for objectivity and responsibility. It sometimes appears laborious but it is a prerequisite for scientific rigour. It paves the way for the result requirement. • The fourth and final phase is applicative. It means becoming aware of the potential provided by new and fully understood knowledge and possibly to accept changes in the resulting readings. This phase is equally important for getting pleasure out of science and is the moment of synthesis following the efforts volunteered during the inductive and deductive phases. In physics, this corresponds to assimilating fully controlled know-how in the building of a more complete device. In mathematics, it means solving a situation-problem very much as in a professional situation. In all cases, the applicative phase provides the means of using acquired knowledge and enables the progression towards new technical and inductive phases in the future. Another simulation will make it possible to carry out a new four-phase cycle.

4.

Evaluation of current experience and prospects

4.1 APA as a part of the EMN programme At present, APA concerns EMN year 1, which has comprised around 110 students in the last 7 years for physics and 4 years for maths. The physics module is tailored to contain 50 h of in-house teaching and 100 h of individual homework. The maths module is tailored to contain 70 in-house hours of teaching and 80 h of individual homework. Apart from year 1, a physics APA module is available to students entering the school in year 2. These marginal sources

114

L. Bot et al.

of recruitment average 25 students per annum. This specific module comprises 40 in-house hours and 40 h of individual homework. Equally in year 2, an APA module was designed in 1999 to teach complex analysis, statistics and probabilities. It was tested out in 1999–2000 and 2000– 01 on a pilot group containing 24 students with 75 h of in-house teaching. Owing to the lack of resources available to monitor the whole year, the module was converted into project teaching, which now rounds off the mathematics programme in year 2. 4.2 The organization of APA modules at the EMN There are six full-time members of staff teaching the physics module, not forgetting the technical assistance for the planning of the individual sessions and the maintenance of individual carrying cases placed at the disposal of the students for their experiments. The areas of activity involving the members of this team are divided into optics, nuclear physics, electronics, microelectronics, mechanical engineering and thermodynamics. The team includes teacher-researchers also involved in fundamental research activities or in contractual research activities. The teaching team is co-ordinated by an engineer with experience in design in the aircraft industry and who is involved in the management of the ‘APA-physics’ project on a part-time basis. He guarantees the pedagogical coherence of the modules provided. There are 10 teachers involved in the maths module and they are either permanent members of staff or part-timers. The areas of activity are divided into maths, theoretical physics, statistics and probabilities applied to logistics and operational safety, process engineering and numerical resolution of transport equations (fluid mechanics, heat transfer). An engineer from the mining profession is also included, who is in charge of industrial development involving the Ministry of Industry. The teaching team is co-ordinated by a mathematician who manages the ‘APA-maths’ project. 4.3 Differences between ‘APA-physics’ and ‘APA-maths’ A significant difference between ‘APA-physics’ and ‘APA-maths’ modules is the fact that the physics module contains only APA-type sessions, whereas the maths modules includes APA-type sessions, traditional lectures and class work. The objective of the physics module is to teach experimental physics. It is not organized in relation to progress referring explicitly to contents or to an academic programme. It seeks to incorporate knowledge acquired previously or in parallel with other physics teaching into each student’s end of year achievement of a complete experiment based on a subject of physics (heat regulation, cosmic rays, etc.). Throughout the module, the students must validate the intermediary stages (electronic assemblies, electrical supply, procedures, etc.) necessary to develop a full installation. The construction of this appliance is sequenced through several situation-problems corresponding to the different types of know-how that have to be assimilated. The resolution of each situation-problem is spread over several in-house sessions that are spaced out over several days so as to allow the individual to work. The level of the analysis teaching in the mathematics module is 2 years after Higher Leaving Certificate. Such an academic programme is organized into several sequences that provide a complementarity between learning by action sessions, lecture and supervised class work. A sequence corresponds to a particular subject of the lesson or to a pooling of notions (e.g. numerical series). A sequence begins with guided explorative work (TDD in French, the last D standing for ‘discovery’), which generally spans two in-house sessions with several days between each to allow for individual work. Each TDD (the technical and inductive phases)

‘Learning by doing’

115

corresponds to a precise situation-problem and may often require overlapping experimental and/or modelling work. After this TDD, the students feel the need for maths and reasoning that surpass their previous knowledge. These notions provide the basis of the lectures (CM in French), followed by traditional class work (TD), which enable the students to acquire the know-how essential in mathematics, i.e. calculus and demonstrations rounding off the intuition that is often caught out in the TDD situation-problems (the deductive phase). Time permitting, and depending on the teaching team’s identification of a useful application of the notions defining the sequence, this is brought to a close by practical work (TP in French). The situationproblems used in the TPs are of a technological kind and are often much more complex than those used in the TDDs. They are an opportunity to assimilate mathematical skills in related areas and part of the engineers’ training programme (applicative phase). They provide an open view of the life of an engineer and avoid bringing maths down to being a mere ‘tool box’, as this would go against the autonomy that APA seeks to encourage. These differences come from circumstantial choice, not from any opposition between maths and physical science in the planning of this APA type of teaching. It would be possible to design the academic teaching of a particular discipline in physics in sequences, which in relation to discipline would assimilate APA-type sessions, lectures and class work into a coherent whole. The TDD–CM–TD–TP progression stated in the case of maths along with its logic for incorporating fundamental knowledge and technical know-how are applicable for all scientific learning. Likewise, it would be possible to design a mathematical module without relying on any subject programme, and uniquely aimed at training students in modelling and resolving problems. This would thus test the extent of their previous mathematical knowledge and their ability to implement it on their own. Beyond these differences coming from the context in which our experience was managed, different versions of APA are possible and can be complementary. On the one hand, a version containing only APA-type sessions (such as ‘APA-physics’ described in this article) seems to be adapted for the assessment or synthesis stages of training. It may force learners to build for themselves summaries of previous knowledge, to link together concepts learned from different courses, in order to solve situation-problems that do not respect frontiers of disciplines. Such an assessment stage is structuring before the next stage in the training devoted to a totally new learning. During it, students can ask themselves questions for which they need to stand back a bit from the courses. We think that such assessment stages, using widely APA-type sessions, could be systematically organized at strategic moments of training (determination of future course of studies, . . .) in order to validate students’ learning in terms of know-how. On the other hand, an intermediary version including APA-type sessions and traditional lectures (such as ‘APA-maths’ described in this article) seems to be interesting, to link concepts of one discipline with concepts of others. This need is obvious when knowledge is not an autonomous objective for training (i.e. maths in engineer training), but a way toward learning know-how essential to disciplines that are central in training aims (i.e. solving equations used in physics) or toward learning transversal skills (i.e. reasoning, . . .). If we can formulate general advice about the choice of pedagogic, we do not think that a complete training should rely on a unique pedagogic device, even if time and resources would allow the exclusive use of APA-type sessions. An important value of APA for teachers who took part in it was to (re)stimulate all their lectures, including traditional ones. To the claim of active methods to make the learner an actor of his learning, we would associate the necessity of allowing the teacher to set an example, being the author of his pedagogic. We recommend that our colleagues maintain a minimum diversity of pedagogics, taking inspiration from existing and well-formalized methods, without excessive denigration of traditional lectures or too much improvization or incessant changes. This diversity is helpful to break the routine inherent in the exclusive practice of a unique pedagogic, whatever it is; routine that can put to

116

L. Bot et al.

sleep teachers as well as students. The diversity stimulates evolution and the hybrid solutions, allowing teachers to take into account all the limiting factors of training (time, material and human resources, etc.), but keeping the essentials of their pedagogical aims.

4.4 The costs of the method and the difficulties in assessment To design a situation-problem takes up around 2 or 3 weeks of a teacher-researcher’s time and often needs a period of time for partial redrafting after it is first taught. The technical resources are not very costly. Firstly, they are often available in the research laboratories that all higher education establishments use. Secondly, APA has no use for overcomplicated or integrated machines or technology that will function like so many ‘black boxes’ and that will not allow the student to exercise either his freedom or his critical mind. In the event that these tools are necessary, above all if this means computer software and measuring devices, efforts will be made to use only the ones that seem to be ‘universal’. The main difficulty to surmount is the gathering of a team of multidisciplinary teacherresearchers. Practising active learning methods means putting the notion of the programme into a framework that favours the ‘hard core’ notion of knowledge. In relation to the teaching goals that are fixed, a ‘hard core’ makes it possible to sequence the disciplines in a sufficiently rigorous way, while allowing the mobilization of related skills that are useful for designing the situation-problems. These mobilized related skills that are on the edge of the main teaching goals offer as much space for the free expression of the students. The partial relativization of the academic programmes re-enhances the objects of knowledge that are not the exclusive property of a particular discipline. Given our mentality, this is difficult to adapt as it itself has been forged by a very structured and compartmentalized academic vision. Finally, the quantification of the real added value of a particular teaching method in terms of its long-term efficiency for student learning appears extremely difficult. On a generic level, as far as active learning methods are concerned, it has been shown (Norman and Schmidt 1992) that relatively more ‘passive’ traditional teaching allowed a sizeable acquisition of short-term knowledge (the assessment is carried out at the end of the teaching), but this was small in the long term (the same assessment is again made on the same individual a year later), whereas active learning methods restricted learning in terms of the amount of the ‘programme’ that was followed (the same assessment shows knowledge not to be as broad in the short term as after traditional teaching), but better used in the long term. The question of experimental validation remains a tricky one. On the one hand, the making up of samples of individuals representative and stable within a time frame and at a scale that is sufficient to validate the long-term efficiency of a method is beyond the scope of an establishment like the EMN. On the other hand, the criteria of knowledge validation used during the above cited study remain formal and close to the academic situation. In keeping with engineer training objectives, the criterion of ‘professional operationality’ must be considered, but we are totally lacking in validated scales of references concerning scientific training. This legitimate preoccupation with validation must not divert us from what would be a paradox if we were to announce our subscription to a particular method and to convince the student to avoid it when making reference samples. The first factor governing the success of such methods is in the subscription of the teachers, in the resumption of a debate among them and in front of the students, and in emulation through something new. The intrinsic dimension of all pedagogic activity is seemingly a de facto limitation to the full formalization of any ‘new method’ and its assessment. Our remarks point towards propositions relating to scientific teachings that are more lively and are inspired by what goes on in the laboratories, the studies and state-of-the-art activities. The pedagogics must not be treated as a complete

‘Learning by doing’

117

science permitting formal assessments unconnected with the tangible contents that have been taught. Active learning methods should rather be experienced at first hand along with their contingencies and their enthusiasms. Whether this means learning or teaching, the temptation to assess everything is something to avoid. The freedom and efficiency of the action supposedly accepts a residual doubt on the question of ‘essentials’.

5.

Conclusion

We have described the essential features of our experiment in APA, a method developed at the EMN used to teach in maths and physics. This type of teaching has several degrees of adaptation between active learning and traditional methods and has the advantage of retaining the quintessence of traditional teaching qualities while rectifying its most recognized flaws. As a response to the pedagogic challenges that our society addresses in engineer training, the key points of APA are real-life simulation, management of non-success, result requirement and the monitoring of students’ progress. The implementation of the teaching is performed in four important phases so as to acquire each new piece of knowledge, i.e. the technical phase, the inductive phase, the deductive phase and the applicative phase. These phases are similar to cycles of problem-solving. Most students certify that APA provides them with greater insight into the role and place of scientific disciplines in relation to each other, as regards both their programme and their future engineering professions. Although the assessment of the methods remains a rather delicate matter, the undeniable advantage of our approach is in its ability to bypass the educational manifesto so as to transform our ideas into concrete experience. Contact with the student is all the better as the teaching teams are both involved and coherent. The multidisciplinary approach of these teams is an important factor in the success and is seemingly the major precondition for organizing such teaching. Their coherence is not organized around linear disciplinary programmes, but rather the ‘hard cores’ of knowledge and skills expected in a given training programme calling extensively on the complementarities of the disciplines involved. As an important outlook for future developments of our method, we cannot ignore links between its aims and the pragmatic philosophy of education formulated by John Dewey (1916, 1933, 1938) in the early 20th century. As far as we understand references consulted when writing this paper, it seems that ‘learning by doing’ is a good translation into English from the French ‘apprentissage par l’action’, even within Dewey’s meaning and respecting relations between this author and constructivism (de Graaff and Kolmos 2003). Of course, we cannot assume large developments of the constructivism and experiential learning generated in psychology, cognitive and educational sciences from Dewey’s intuitions because they are very much more elaborated and validated than our method. Nevertheless, the scientific method and the development of technologies were used as a paradigm by John Dewey and his continuators. Among them, the model of ‘discovery learning’,4 developed by Jerome Bruner in the 1960s, is particularly close to the experience described in this article. Descriptions of ‘discovery learning’are similar to key concepts of APA: ‘. . . One of Dewey’s most important contributions to instructional technology is his conception of instruction in terms of scientific method, which he termed the “reflective method”. In this method, the learner first recognized a problem, then formulated a hypothesis that offered possible solutions or outcomes. Through reflection and experimentation, the hypothesis was then tested so that the learner could draw a conclusion. . . . Discovery learning, reception learning, and assigned learning (scaffolding) are three instructional models based on this method. . . . Discovery learning, proposed by Jerome Bruner in 1966, is one on the most influential cognitive models as

118

L. Bot et al.

it has many applications to science and related fields. Discovery learning encourages students to discover principles for themselves through experimentation and exploration. It is the teacher’s job to provide guidance when needed, but not before the student is allowed to explore a problem on her own, using previous knowledge and experiences, personal motivation, and experimentation’.5 We recognize the result requirement, the facilitator, as one of the teacher’s roles and the real-life simulation.

Notes 1. The translation from French into English for the expression ‘apprentissage par l’action’ would be ‘learning by doing’; but this translation could be ambiguous after the formalization in the early 20th century of ‘learning by doing’ by John Dewey as a key concept of his philosophy. The expression ‘learning by doing’ is still used by Dewey’s continuators. For example, psychologists such as Jerome Bruner or Jean Piaget developed Dewey’s intuitions from their observations of a child’s behaviour in front of learning challenges in complete theories of learning. Our experience is far from being so elaborated. Therefore, we chose not to use the translation in the text. As an outlook of this article, we shall argue that ‘learning by doing’ seems to be a good translation of ‘apprentissage par l’action’ because of similarities of our method with Dewey’s thinking about learning sciences and technologies. 2. The Zap! method was developed by CALTECH and is part of the wider hands-on methods common in the USA at all levels of teaching of experimental sciences. In France, the movement inspired ‘La Main à la Pâte’ method (literally meaning ‘getting down to it’ method) for primary teaching. Please see Germinet (1997) for the relationship between US and French methods and the implication of certain researchers in the development of this active learning methods. See Pine et al. (1996) for the ZAP! method. See Rauch (2001) on the APA method in sciences and the publications cited in this article. See Education Development Centre (2003) for examples of hands-on teaching. 3. For details and other examples of real-life simulations used in ‘APA-mathematics’, see http://www.ale2002.dtv.dk/presentations/Ludovic%20Bot.pdf. 4. http://home.earthlink.net/∼dougary/ITEC_800/final_project/dewey.htm 5. Cited from http://home.earthlink.net/∼dougary/ITEC_800/final_project/dewey.htm and http:// home.earthlink.net/∼dougary/ITEC_800/final_project/constructivism.htm; italics are added to show similarities with APA.

References Castela, C., Les Objets du Travail Personnel en Mathématiques des Etudiants de l’Enseignement Supérieur: Comparaison de Deux Institutions: Université et Classes Préparatoires aux Grandes Ecoles, 2002 (Cahier de DIDIREM: Paris). Crozier, M., L’Entreprise à l’Ecoute, 1989 (InterEditions: Paris). Dewey, J., Democracy and Education. An Introduction to the Philosophy of Education, 1916 (Free Press: New York). Dewey, J., How we Think, A Restatement of the Relation of Reflective Thinking to the Educational Process, 1933 (D.C. Health: Boston). Dewey, J., Experience and Education, 1938 (Collier Books: New York). de Graaff, E. and Kolmos, A., Characteristics of problem based learning. International Journal of Engineering Education, 2003, 19, 657–662. Education Development Centre (collection), An elementary hands-on inquiry sciences curriculum (INSIGHTS) 2003. Faucher, G. and Martin, F., Méthode de Résolution de Problèmes, 2000 (Presses Internationales Polytechniques: Montréal) (educational software). Genelot, D., Manager dans la Complexité, 1992 (INSEP Editions: Paris). Germinet, R., L’apprentissage de l’Incertain, 1997 (Odile Jacob: Paris). Goldratt, E.M. and Cox, J., The Goal, A Process of Ongoing Improvement, 1984 (North River Press, Croton-onHudson: New York). Hunot, F., Former les Nouveaux Managers, Problem-based Learning, 2000 (Editions Liaisons: Paris). Jarrosson, B., Invitation à une Philosophie du Management, 1991 (Calmann-Lévy: Paris). Lecomte, M., L’enseignement des mathématiques pour ingénieurs par une méthode d’apprentissage par problèmes. European Journal of Engineering Education, 2002, 27, 257–266.

‘Learning by doing’

119

Lemoult, B. and Tellier, F., Compétence et comportement en situation d’apprentissage. In Actes du 2nd Colloque, Question de Pédagogie dans l’Enseignement Supérieur: Réflexions, Projets et Pratiques, 2003 (Brest: ENSIETA et ENST-Bretagne). See also publications cited in this article. Milton, R.G. and Stinson, J.E., Education leaders for the new competitive environment. In Economic and Business Administration: The Case of Problem-based Learning, 1995 (Kluwer Academic: Dordrecht). Norman, G.R. and Schmidt, H.G., The psychological basis of problem-based learning: a review of the evidence. Educational Researcher, 1992, 67, 557–565. Ourisson, G., Désaffection des étudiants pour les études scientifiques, rapport au ministre de l’éducation nationale, gouvernement français, mars 2002, http://www.education.gouv.fr/rapport/ourisson/default.htm Pine, J., King, J. and Morrison, P., ZAP!, Experiments in Electrical Currents and Fields, 1996 (Jones and Bartlett Publishers, Sudbury, MA). Rauch, C.P., Approche Inductive d’un Problème de Régulation Concret, 2001 (actes CESTIS-EEA: ClermontFerrand).

About the authors Ludovic Bot was born in 1971. He got his PhD in Nuclear Physics in 1998 at the University of Nantes (France). He was Associate Professor at the Ecole des Mines de Nantes from 1998 to 2003, where he taught physics and mathematics. In May 2003, he left EMN for another engineering school. His main research activity is now dedicated to educational sciences. Pol-Bernard Gossiaux was born in 1967. He got his PhD in Nuclear Physics in 1993 at the University of Liege (Be). After three postdoctoral stays at the LPN Nantes (France), MPI Heidelberg (Germany) and at MSU (USA), he joined the Ecole des Mines de Nantes in 1997 as an Associate Professor, where he has since taught physics and mathematics. His main research activities are dedicated to the collisions of ultrarelativistic heavy ions and the formation of quark gluon plasma. He has been responsible for the teaching of mathematics from 1999 to 2003 and is now responsible for the teaching of theoretical physics. Carl-Philippe Rauch graduated in Aeronautical engineering in 1982. After working for a few years in industry, he left to teach physics at national colleges of engineering and graduated in optics in 1989. In 1996, he introduced learning by doing to the Ecole des Mines de Nantes, at both undergraduate level (engineering students) and for accompanying primary schools. Nowadays, his work aims to connect educational methods at primary, middle and high schools and university and college levels. Safouana Tabiou was born in 1961. She got her PhD in Mathematics in 1986 at the University of Jussieu (France). She taught mathematics and developed her interest in active learning methods at the University of Lome, Togo. In 1996, she joined the Ecole des Mines de Nantes to apply her experience in an engineering curriculum context. Since 1998, she has taken charge of the expansion of the learning by doing method from physics to mathematics at the Ecole des Mines de Nantes, where she is involved in the creation of a centre for pedagogical development.