The role of the teacher in remediating at-risk medical students

Abstract

Background: Previous work identified complex ingredients of a remediation programme for at-risk medical students: mandatory, stable, facilitated small groups promote both cognitive and affective developments, with improved self-regulation, metacognition and reflection resulting in significant performance gains.

Aim: We explore the teachers’ role in this intervention, aiming to expand and deepen understanding of remediation methods in medical education.

Methods: Extensive qualitative data from student surveys and in-depth teacher interviews, along with quantitative student performance data, produced a rich description of remediation processes.

Results: Remediation should support emotional needs and foster cognitive and metacognitive skills for self-regulation and critical thinking. Teachers of remediation need to motivate, critique, challenge and advise their learners, applying teaching and contextual expertise in a constructivist, student-centred environment that fosters curiosity and joy for learning. Teachers of remediation can mediate these processes through embodiment of five core roles: facilitator, nurturing mentor, disciplinarian, diagnostician and modeller of desired skills, attitudes and behaviours.

Conclusion: Remediation of struggling medical students can be achieved through a cognitive apprenticeship within a small community of inquiry that motivates and challenges the students. This community needs teachers capable of performing a unique combination of roles that demands high levels of teaching presence and practical wisdom.

Introduction

With the trend towards admission of students from diverse, non-traditional backgrounds to medical education (Howe et al. 2008; Jolly et al. 2008), comes increasing numbers of students who do not know what to expect or what will be expected of them (Fitzmaurice 2010), and thus a need for increased academic support to help them succeed (Winston et al. 2010a). Common difficulties include organizing and integrating large amounts of information, test-taking and test anxiety (Paul et al. 2009), time management, ineffective study strategies, insufficient background and content knowledge, weakness in literacy and numeracy skills, and a general lack of self-regulatory, metacognitive and critical thinking skills (Winston et al. 2010b). However, these weaker students tend to avoid seeking assistance (Devoe et al. 2007), and frequently fail a course in medical school (Winston et al. 2010b).

Students who fail are typically allowed to repeat, and yet repetition without remediation rarely leads to long-term success (Mattick & Knight 2007; Winston et al. 2010a). A handful of successful remediation programmes has been reported in the medical education literature (McGrath & McQuail 2004; Alexander et al. 2005; Kies & Freund 2005; Saxena et al. 2009; Cleland et al. 2010), but there appears to be an absence of guidelines about how best to accomplish remediation, along with frequent calls for further research into details of effective, replicable strategies for remediation of struggling medical students (Kies & Freund 2005; Paul et al. 2009; Saxena et al. 2009; Cleland et al. 2010).

Table

Practice points

Remediation of at-risk medical students works well in small, student-centred, communities of inquiry that address both cognition and emotion, by fostering self-regulation, critical thinking and a joy for learning. Teachers of remediation can mediate these processes through embodiment of five core roles: facilitator, nurturing mentor, disciplinarian, diagnostician and modeller of desired skills, attitudes and behaviours. Performing this complex blend of roles requires contextual and teaching expertise, teaching presence and practical wisdom.

Winston et al. (2010a) describe the details of a mandatory remediation programme for students who fail their first semester at a large Caribbean medical school. Students repeating that first semester on probation, in addition to retaking basic science courses, must meet twice a week, for 12 weeks, in a small group of five or six students, with a faculty member from the school's Centre for Teaching and Learning. The key themes in the carefully designed syllabus of this remediation course, which draws on a blend of theories and research-proven techniques, include: self-regulation, metacognition and reflection; explicit, think-aloud discourse; active learning; giving and receiving feedback; application of learning processes to basic science course content; care with language; accuracy of thinking and attention to affect and motivation. Participants have performed significantly better than students who repeated the first semester at the school prior to the introduction of this mandatory programme: pass rates have increased from 58% to 91% for examinations taken at the end of the repeated first semester (n = 511; p < 0.0001, 15 cohorts over 5 years) and from 28% to 51% for national examinations (United States Medical Licensing Examination (USMLE)) taken five semesters later (n = 216; p < 0.0001; remaining 295 participants have yet to take USMLE).

Subsequent clarification research (Cook et al. 2008) showed that the participating at-risk medical students believe that successful remediation should be mandatory, works best in small stable groups supported by honest teachers with rigorous expectations and good facilitation skills, needs to challenge students’ conceptions of learning, encourage their development as flexible, reflective learners, and must take into account a blend of cognitive and affective factors. Exploration of these students’ perspectives made it clear that remediation is a complex, context-dependent process, in which elements of the learning environment – students, teachers, syllabus, school and more – all play major, mutually constitutive roles (Winston et al. 2010b).

In order for such a remediation programme to be used effectively in other contexts, it is essential to gain a deeper understanding of the role of the teachers and precisely how they affect the students’ outcomes. Previous work has noted that teachers of remediation need a good understanding of the learning and assessment environment (Mattick & Knight 2007), and should involve a teacher–student relationship that builds confidence while maintaining high expectations and challenging students’ thinking (Winston et al. 2010b). However, we have found no literature at all on teachers’ views of their role in effective remediation of students in difficulty in medical education, nor, indeed, in other areas of higher education. The goal here is to begin to address this literature gap, to identify the qualities of good teaching that most apply to remediation in medical education, and to relate these qualities to the broader literature on teaching expertise.

The importance of the teacher

Dewey (1938) says that the learning environment includes the conditions, materials, people and total social set-up with which an individual interacts, and teachers are certainly an element of any pedagogic setting. There seems little doubt that the quality of teaching influences student learning and achievement (Thompson et al. 2008; McNeill 2009; Patel et al. 2009), that teachers are key factors in both small group learning (Jaarsma et al. 2008; Postholm 2010) and successful student support programmes (Muraskin 1997), and that academic advising can impact student performance (Tekian et al. 2000). Students perceive teachers as having strong influence on their experience (Harris 2010; Kreber et al. 2010), and both faculty and students rate intangible teacher variables, such as interest, enthusiasm and approachability, as having greatest importance for learning (Lammers & Smith 2008). And yet, Foster (2010) found no association between specific tutors and either student attrition or improvement of weaker students. Indeed, there is considerable uncertainty about which teacher characteristics positively affect student achievement (Konstantopoulus & Chung 2011), and about the precise role of teachers in supporting at-risk students (Barkas 2011).

Considerable work in education research has shown that teachers’ beliefs and conceptions affect their teaching practice (Bolhuis & Voeten 2004; Song et al. 2007; Feucht & Bendixen 2010; Mahlios et al. 2010), which in turn influences students’ help-seeking strategies (Kozanitis et al. 2007) and approaches to learning (Prosser et al. 2003; Kember et al. 2008), and, ultimately, student outcomes. Thus, in order to understand the attributes and practices of teachers required for excellence in assisting at-risk medical students, we need to explore teachers’ beliefs (Pajares 1992), using sources that include practitioners’ own descriptions of their work (Kelchtermans 2009; Fitzmaurice 2010). Hopefully, by drawing upon the wisdom of practice (Shulman 2007), we can develop a deeper understanding of the key elements that inform the pedagogical content knowledge important to remediation in medical education (Shulman 1986, 1987).

Aims

This study aims to build towards a practical theory of remediation in medical education, and, in particular, to examine the role of the teacher in this process. Thus, the main research question is:

What is the role of the teacher in remediation of struggling medical students?

This broad question can be broken down into sub-questions that allow for exploration of different perspectives that may steer us towards a deeper understanding of remediation. In previous work, student focus groups suggested that the teacher is one of several key factors in the complex process of remediation (Winston et al. 2010b). To probe deeper into the student view, we aim to answer the following question:

How do students believe the teachers influence student success?

The teachers’ voice is essential to understanding their role, and thus we ask:

How do remedial teachers in medical education see their role? What do they perceive to be the most effective elements in their remediation of at-risk medical students?

Finally, we look at the student performance data, asking:

Are there any differences in outcomes of students allocated to different teachers? If so, how can these be explained?

The results of these explorations will then be examined for links to theoretical perspectives in an effort to continue to build essential bridges between theory and practice (Sparapani et al. 1996; Kaufman 2003). It is, of course, important to acknowledge our own perspectives (Dewey 1938): theory building inevitably depends on the conceptual framework used by researchers. Since education and learning are inherently complex processes (Jörg et al. 2007; Winston et al. 2010b) in which all elements of the pedagogic setting are mutually constitutive (Varela et al. 1992), we take on a complexivist framework (Jörg et al. 2007), which we hope will enable us to go beyond the limitations of any single conceptual framework and draw upon a series of conceptual lenses (Bordage 2009) to develop a deeper understanding of the complexity of remediation.

Methods

A complexity framework requires developing understanding through the interactions between the related components of a system (Thomas 2006), so the study of complex interactions benefits from a mixed methods approach (Kennedy & Lingard 2006; Schifferdecker & Reed 2009). Here we attempted the triangulation of some quantitative and considerable qualitative data to seek a valid balance of stakeholders’ voices (Denzin & Lincoln 2005).

First, a brief survey was sent electronically to 310 students who had successfully completed our remediation programme when they repeated their first semester at our medical school (Winston et al. 2010a). These were all current students at the school (in semesters 2 through 10), many of whom had moved on to clinical rotations by the time of this study. The survey asked students to name their teacher during the programme, to what extent that teacher had influenced the outcome of their participation (Likert-type scale) and to describe what that teacher did that was helpful and how that influenced their learning. A limitation of this approach is that, lacking contact details of those who failed out of the school, the views of unsuccessful students are not represented. However, given the appreciative nature of this part of the enquiry, we feel these successful students’ views contribute to our understanding of what makes remediation work.

At the time of this research, six faculty members had been teachers on the programme, and all students included in the data here were members of a group taught by one of these six teachers. One declined to participate, and one is the first author of this article (KW). The remaining four were sent questionnaires, which asked for demographic data, and included a series of open-ended questions inviting views on the remediation programme and their role within it. The teachers were then invited to participate in 1-h flexible semi-structured interviews (Feucht & Bendixen 2010), for which the questions were formulated to explore and clarify the questionnaire responses: three interviews were conducted, while one teacher chose to engage in written dialogue to answer the interview questions. KW also completed responses to both the questionnaire and interview questions.

The inclusion of data from KW requires some explanation. The attempt to understand the teachers’ views is a phenomenological approach, an exploration of lived experience aiming to discover, describe and classify the teachers’ conceptions of their reality (Varela et al. 1991; Allan & Clarke 2007). In all qualitative research, results are suffused with researcher perspectives (Denzin & Lincoln 2005), the researcher is an intrinsic part of the method, upon whose understanding of the field the outcomes of interviews depend (Kvale 1996). Further, there is a need for insider understanding, for self-study research that is essential for the development of practical wisdom (Postholm 2008; Lunenberg & Korthagen 2009). Thus, given that KW is both designer of and teacher in this programme, any attempt to claim objectivity would surely ring hollow: neither data collection nor analysis can be neutral (Mauthner & Doucet 2003), both are influenced by the researchers, and here we admit and embrace the inevitable inclusion of researcher perspectives.

We believe that this does not exclude analytical rigour. The phenomenological method involves looking for constants among varieties of experienced phenomena to construct descriptive categories of collective understanding (Kvale 1996; Greasley & Ashworth 2007). The qualitative data from both students and teachers were initially subjected to bottom-up analysis, with comments coded and grouped into pools of related meaning (Harris 2010). These pools were then further abstracted and refined to derive themes that could be analysed top-down for comparison of new findings with previous research and the extant literature. Finally, the findings were cross-checked and verified by the participants (Kitto et al. 2008).

The third method in this triangulation approach was a quantitative analysis of the short- and long-term outcomes of the students, grouped according to which teacher they worked with during their participation in the programme, the goal being to determine whether any differences in perspectives could be correlated with differences in objective performance.

The School's IRB has approved this research project. All participants signed an informed consent form. Identities have been kept confidential and removed from all the presented data.

Results

In qualitative research, where procedural rigour and reflexivity are essential (Mauthner & Doucet 2003; Kitto et al. 2008), and boundaries between data collection and analysis are blurred (Pope et al. 2000), there is inevitably some overlap between methods and results. So, in an attempt to render the interpretative process a little more visible, our results are presented in the order they were found.

Students’ views

It is important to note that allocation of students to groups in the remediation course (and thus to specific teachers) is essentially random: for scheduling purposes, we make use of their laboratory group designations, which are random. The 188 students who responded to the survey (61% response rate) were spread in approximate proportion to the number of students taught by each teacher. When asked whether their teacher influenced their outcomes, the responses were: strongly positive influence, 76%; somewhat positive influence, 20%; no influence at all, 3%; somewhat negative influence, 1%; strongly negative influence, 0%. Clearly, then, the students believe that their teacher makes a difference, and for the better.

For all six teachers, students’ made similar comments on how their teachers helped their learning. The terms encouragement and motivation featured prominently, as did approachability and honesty. These students had failed, and appreciated the support their teachers provided:

She was very supportive and encouraged us to succeed despite minor setbacks.

He always managed to help me bring my self-esteem and confidence level up.

They stressed the importance of listening and advising, of simply being interested in them:

He consistently asked us how we were doing, was interested in our progress, and showed a sincere desire to help us pass.

Facilitation skills were acknowledged by many respondents:

‘She always made sure everyone participated’, and ‘gave the group the opportunity to discuss the answers and possible explanations, guiding us so the session does not go off topic’, while remaining ‘sensitive to our individual needs, identifying weaknesses and suggesting changes to overcome them’.

There were many comments about specific elements of the course, and the teachers’ role in enforcing them, while also allowing some student flexibility:

He required us to work outside of our comfort zone and try techniques we weren’t accustomed to, and provided the foundation for me to figure out what worked best for me.

Frequently mentioned was the importance of challenging students’ thinking and holding them accountable for details:

He expected us to perform at a high level and led us very far in our thinking in a very short time. Making us pay attention to all details and pick out words we didn’t understand was very beneficial because now I don’t let other things slide.

He makes you explain things step by step, and asks questions that keep you focused on the problem. It allows you to see flaws you wouldn't have otherwise.

Teachers’ views

The teachers’ questionnaire and interview responses conveyed a rich description of a complex role consisting of a broad range of behaviours and attitudes. All five participants shared similar views, and all followed the same course syllabus and handbook (available by request from KW) with their students. The following description provides a flavour of these teachers’ shared conceptions of the essential aspects of their work.

There was a shared sense of purpose, a common belief in the value of the course for repeating students.

We have to believe in what we’re doing, enjoy what we’re doing, take responsibility for putting everything that we have agreed upon in front of them.

There was a sense of collegiality, of mutual trust among us that I think allowed us to meet the students fully focused on what we thought we could do for them, and I don't doubt that was communicated to the students.

Key to the success of this programme is an emphasis on student-centredness, on active learning, critical thinking, dialogue, reflection and relevance:

Teaching is about allowing people to discover truths through inquiry, experimentation and reflection.

We provide a space for questioning assumptions and making them explicit, for sharing ideas and developing communication skills. The use of illustrations, analogies, examples and stories makes the process more active and enjoyable.

Use of (medical science) curricular content as a vehicle for all exercises is fundamental to creating acceptance from students.

It seems that for this course, medical science content knowledge is helpful but not essential, whereas teaching and context expertise are both considered indispensable:

Knowledge of how people learn, of pedagogical theory, is needed to explain why the ideas are important, and to probe deeper into methods that help them.

Contextual knowledge is very important – the curriculum format, not necessarily details, (though content knowledge is helpful, and also helps earn students’ respect), rules and regulations, the set of behaviours needed for success in this school, and medical school in general.

To successfully convey the core programme to the students, the teachers described five multi-faceted, overlapping roles that encompass and expand the behaviours described by the students. The first of these is as active facilitator, akin to ‘Socrates’ midwife’ (Plato 1997), key for eliciting student talk and providing a dynamic, non-intimidating learning environment that encourages both cognitive and emotional support:

Facilitation skills are essential – enables the group to work together, ensures everyone feels their needs are met, and prevents non-beneficial discussions on unrelated topics.

It's a hard-working process, you’re constantly working at them to maintain order and direction. You have to establish, and maintain, basic ground rules from the start.

Facilitating – do not re-teach the content. Ask probing questions and expect students to give and explore meaningful answers.

The next role might be described as nurturing mentor, which involves advising, counselling, and demonstrating empathy, kindness and belief in the students. The students are dealing with failure and exposed vulnerability in the high-functioning environment of medical school, and this role is crucial to student trust, acceptance and motivation:

I am a friendly, concerned, accessible, and firm mentor. I have belief in them and let them see that I have this, and care about their performance.

I foster an atmosphere that encourages the students to feel at ease in the sessions. The fact that this program exists shows the students that we care about them, so they see it as a positive thing and they accept it.

I should not tell them to work harder unless I know they are not diligent. I should remember that the students who are performing worst are typically working hardest.

The students need relationships that they can trust here. I am not above discussing some of my own problems, past or present, with students, and believe it helps our relationship to do so.

However, this needs to be tempered by the disciplinarian, the strict teacher who has high expectations and demands diligence:

Successful learning is largely a matter of self-discipline. Don’t allow them to indulge in denial or half-reasoned answers; don’t accept unjustified excuses.

Ultimately, my role is to hold students accountable for their thinking and reasoning, to enable the students to become independent learners, to attain self-mastery.

If they report things as going well, I require them to elaborate on what this means, specifically with regards to the activities that we propose.

I believe that many students have been misled by being taught study strategies they can employ without much mental effort. There is a serious communication problem in describing just what that mental effort consists of, and successful teaching of these skills consists of overcoming that communication problem.

Although this course is quite prescriptive, and most students have problems in common, they are still individuals, and this requires the teacher to be a keen diagnostician, who observes and listens attentively, and provides clear, honest feedback:

I must learn to know the students, be astute, knowledgeable and observant enough to be able to identify student weaknesses and provide ‘remedies’ as much as possible.

Know when to ask questions, when to ask why. It's partly knowing the content, partly feeling for when they’re floundering, partly recognizing body language, and watching the group, seeing what's happening.

I have to listen carefully to what the students say, to point out alternative ways of thinking. And it's not just what's happening in the classroom, or what's happening in study time, it's what's happening at home, the boyfriend, that sort of thing.

Provide feedback, clear and specific feedback. Comment on specific positives in their discussions, point out very specific errors in logical reasoning. You also have to catch them doing things right.

Occasionally, I must engage in difficult conversations, in which I advise on pursuit of an altogether different career path to which they may be better suited.

Finally, these teachers talked about being a model for the behaviours they are trying to teach, the need to embody enthusiasm for learning, curiosity, critical thinking, reflective acceptance of feedback. You have to enjoy the work to do it well:

I model ways of thinking and learning. I feel that my biggest contribution is in the area of making connections. There is an essential pleasure in seeing someone suddenly light up with understanding.

I enjoy the relationships that form in the group, and the stability of the group, over the extended time that we get to meet with them.

It's socially responsible work. Helping students to become physicians impacts their lives and the lives of countless people I will never know. I think that doing the job well is a very worthy aim.

Show them how to find joy and interest in the material – demonstrate a thrill at the concepts, awe at medical science. This is our gift to them, that sense of ‘isn’t this amazing!’

Quantitative performance data

The outcomes of students allocated to different teachers are presented in Table 1. Teachers are labelled t1 through t6. The fifth semester marks the beginning of the clinical part of the school's curriculum, so passing fourth semester, completion of the basic science curriculum, is taken as a long-term measure of the success of the remedial programme taken when students repeat first semester.

Table 1  Performance of students allocated to different teachers

Teacher	Teaching experience	Qualifications	Number of students	MCAT mean	Mean fails^a	Pass sem 1 (%)	Pass sem 4^b (%)
t1	40+ years	MEd	112	21.5	1.85	90	65
t2	3 years	MD	39	22.1	1.71	80	49
t3	39 years	Phd	25	20.8	1.76	92	60
t4	3 years	DVM	90	21	1.84	88	54
t5	25+ years	MEd	124	20.6	1.84	93	69
t6	4 years	PhD	39	22.3	1.67	92	54

Notes: MCAT, Medical College Admissions Test, standardized US medical school entrance exam; sem, semester. ^aStudents may have failed 1, 2, 3 or 4 courses in first semester originally. Long-term success correlates with number of fails, p < 0.014.

^bPrior to this programme, 38% passed semester 4. With the programme, overall, 61% passed sem 4. For full details, see Winston et al. (2010a, b).

Short-term success of students was similar with all teachers, appeared to be unrelated to teacher qualifications (t2, t3 and t4 have recognized basic science content expertise), and students of all teachers performed significantly better than students without the aid of the programme (Winston et al. 2010a). The one striking difference seems to be teacher experience: Pearson's correlation between experience and long-term outcomes is 0.78. Table 2 shows this clearly, and significantly. Indeed, given the measures of prior knowledge (mean number of fails during first semester and MCAT, Medical College Admissions Test scores), one would not have expected outcomes to favour the experienced teachers.

Table 2  Experienced versus inexperienced teachers

Teachers	Number of students	MCAT mean	Mean fails	Pass sem 4 (%)
Experienced (t1, t3 and t5)	261	21	1.84	66
Inexperienced (t2, t4 and t6)	168	21.7	1.76	53

Note: χ², p < 0.007 that difference in passing sem 4 is due to chance.

This finding prompted us to look back at the qualitative data for any differences we may have missed. Some isolated comments hint at possible differences. For example, students described experienced teachers as friendly, pushed us outside of our comfort zone and required us to try techniques; in contrast, inexperienced teachers became a friend, could have been meaner, made us feel comfortable and allowed us to try different techniques. The experienced teachers more strongly stressed the importance of discipline and of verbal and non-verbal language; they also expressed more uncertainty about what works best, and sought feedback from the students. The less experienced failed to mention seeking feedback, one acknowledged difficulty in providing feedback, and one discussed a growing realization of the need to push students harder.

However, the similarities between the teachers’ views far outweigh any differences, and the sample size is much too small to generalize from. Still, it is worth noting that contradictions between qualitative and quantitative findings are not uncommon, and, while disconcerting, do suggest areas for further work (Schifferdecker & Reed 2009).

Discussion

The fact that both students and teachers agree on what makes this remediation programme work exemplifies one of its greatest strengths, a shared message successfully conveyed. Drawing together the results, we can summarize the role of the teacher in remediation of at-risk medical students, as experienced by both students and teachers. These teachers need to support their students’ cognitive and affective development, applying teaching expertise and contextual knowledge to a constructivist, student-centred approach that promotes self-efficacy and independence. They need to engage, motivate, observe, critique, challenge and advise their learners; they need to be curious, rigorous, flexible thinkers who are interested in their students’ growth and enjoy teaching. These teachers perform five broad, interrelated roles that embody these behaviours: facilitator, nurturing mentor, disciplinarian, diagnostician and modeller of the skills these students need to turn failure into a successful medical school career. Unsurprisingly, the ability to play all these roles seems to improve with experience.

A broad range of extant literature documents the value of these attributes in education. The cognitive and affective are inseparable (Vygotsky 1986; Lakoff & Johnson 1999): engagement, motivation and the will to learn are dependent on emotion and, when fostered, result in deep learning (Vansteenkiste et al. 2009; Taylor 2010; McCune & Entwistle 2011). There is good agreement that the ability to motivate, engage, explain, challenge and connect with students are core elements of good teaching (Trigwell et al. 1999; Bartram & Bailey 2009). Effective teachers have patience, enthusiasm, high expectations and genuine interest in their students (Hull et al. 1991; Jaarsma et al. 2008; Rosenfeld & Rosenfeld 2008; Thompson et al. 2008). Enjoyment of teaching, conveyed to the students, matters (Confucius 1979; Frenzel et al. 2009). In this nurturing mentor role, the teacher acts as a friendly companion who provides the psychological safety, healthy intimacy and advice that nurtures this safe environment conducive to reflection and learning (Reeve 2009; Sibii 2010).

Context gives meaning to words and actions (Bateson 1979), and awareness of contextual effects on learning, especially the interconnected relationships between various parts of the curriculum, pedagogy, learners and content is essential to effective teaching (Shulman 1986; Stooksberry et al. 2009). In constructivism, where other students are part of this context, the focus on collaboration, dialogue, metacognition and articulation of meaning enhances creative thinking and self-regulation, and thus improves student performance (Dewey 1910; Mercer 2000; Kaufman 2003; Postholm 2008; McCune & Entwistle 2011), and this is especially important for weaker students (Brigman & Webb 2007; Kistner et al. 2010). This combining of autonomy support and structure by student-centred teachers is known to increase conceptual change, self-efficacy and independence (Kaufman 2003; Prosser et al. 2003; Reeve 2009; Mitchell 2010): more distributed power results in more productive argument and discourse (Bloom & Volk 2007) and enhanced pedagogical efficacy (Hmelo-Silver & Barrows 2008).

It is the role of the facilitator to enable these constructivist dialogues, to ensure all group members are heard, to join in the discussion by challenging ideas and asking probing, Socratic questions (Martin & Arendale 1992; Plato 1997; Postholm 2010; Taylor 2010) that foster the curiosity central to reasoning (Dewey 1910; Dyche & Epstein 2011). The facilitator is an active participant who helps create a community of inquiry by asking for clarification, identifying connections between ideas, summarizing and posing different views; the facilitator should be a disruptive influence who prevents groupthink and premature convergence of opinions by promoting wider exploration (Stoyanova & Kennedy 2010; Sayama et al. 2011), especially important with less successful students who tend to readily agree without consideration of alternative explanations.

Many similar skills are also needed by the expert diagnostician. Teachers have to study their students, to carefully observe nuances of language, gesture and interaction to detect signs of misunderstanding (Dewey 1910, 1938; Mercer 2000; Boudreau et al. 2009). A hallmark of expert educators is the ability to problematize, take significant facts as evidence, cherish error as opportunity, and to act on their diagnoses to provide feedback (Dewey 1910; Tsui 2009; Davidoff 2011). Accurate, timely, specific feedback facilitates the development of self-efficacy and self-regulation, and particularly helps lower achieving students (Harlen 2003; Fenollar et al. 2007; Hattie & Timperley 2007; Carless et al. 2011). It takes considerable discipline to act on feedback and change habits, so the compassion of the nurturing mentor has to be applied judiciously (Jonas 2010), tempered by the strict disciplinarian who critiques candidly (Lim & Seet 2007), stretches learners with required, scaffolded activities (Vygotsky 1986; Mercer 2000; Wilson & Bai 2010) and demands explicit, non-vague use of language for training rigorous thinking (Dewey 1910).

The skills required for these teacher roles overlap those the learners need to be successful students, and, ultimately, good physicians: doctors need to be trusted by their patients, critically consider evidence, context and patients’ lifeworlds and be willing to jointly construct explanations and interventions (Groopman 2008; Dahlberg et al. 2009; Fuks et al. 2009). The relationship between a teacher and remediating students parallels the doctor–patient relationship in many ways; authentically modelling skills, attitudes and behaviours is an integral part of this work (Kreber et al. 2010; Dyche & Epstein 2011).

To combine all these roles, an educator needs the capacity to handle complex emotional and cognitive tasks (Steiner & Peifer 1990), to be highly adaptive, able to draw on both physical and ever-changing social surroundings (Dewey 1938; Corno 2008). Expertise in any complex domain takes extended practice (Alexander et al. 2009), and experience is known to correlate positively with student performance (Darling-Hammond & Youngs 2002). Inexperienced teachers may overvalue affective over cognitive student outcomes (Pajares 1992), even while convinced of underlying principles (Bolhuis & Voeten 2004). Understanding and practical wisdom are not the same (Aristotle 2008): becoming an expert teacher is a lifelong process (Hultberg et al. 2008) requiring not only skills and competencies, but also development of personal practical knowledge derived from experience of many particular situations (Field & Latta 2001; Aristotle 2008; Ben-Peretz 2011). In this study, the three experienced teachers have taught a wide range of disciplines in various institutional contexts, so, although the data may be an artefact of individual differences within a small sample, the literature suggests experience should count.

This brief literature survey suggests that the teaching of remediation requires similar skills to teaching in general. However, given the skills deficits and vulnerabilities of at-risk students, especially in the context of a high pressure, fast-paced medical school environment that demands extremely high functioning, we would suggest that remediation of struggling medical students needs teachers who can bring a unique, flexible combination of roles to their daily practice. These students have previously struggled in largely anonymous, lecture-based courses, and our work suggests that the teaching of remediation cannot be considered as somehow subordinate to teaching (Barkas 2011), nor left to the most junior faculty or student tutors, as is frequently the case, but rather requires considerable teaching expertise.

Thus far, we have given a rich description of the roles of the teachers, rather than provided a theoretical model; Wittgenstein (1967) would say we should stop there, that ‘explanation’ adds nothing to description. Nevertheless, drawing ‘fox-like’ on several theories has been shown to be applicable to the real world (Tetlock 2005; Chinn & Samarapungavan 2009), so here we reflect on ideas that may help towards our construction of a theory of remediation.

Previous work has shown that successful remediation is a complex, context-dependent process, ideally delivered as a mandatory course to small groups. It should be based upon a blend of constructivist, student-centred techniques that acknowledge the importance of both cognitive and affective factors in the development of self-regulation, metacognition and critical thinking (Winston et al. 2010a, b). Now we can add the five teacher roles that mediate this process: facilitator, nurturing mentor, disciplinarian, diagnostician and role model. The cognitive part of this is reflected in the ‘cognitive apprenticeship model’, in which teachers model, coach and scaffold learning using questioning (articulation), reflection and exploration within a positive learning climate, gradually withdrawing support as students progress (Collins et al. 1989; Stalmeijer et al. 2009).

In remediation, we suggest that this apprenticeship plays out within a ‘community of inquiry’ (Stoyanova & Kennedy 2010), a methodology that considers education as a complex dynamic system in which components are structurally coupled and mutually constitutive (Maturana & Varela 1987; Jörg et al. 2007). A community of inquiry provides a safe environment for a group of students to learn critical thinking through dialogue, where members build on each other's ideas to reconstruct their understandings, guided by a facilitator who provides the optimal amount of ‘noise’, feedback and challenges that allow enough space for student exploratory talk while preventing groupthink and provoking discussion (Proulx 2010; Stoyanova & Kennedy 2010).

The teacher is thus an integral component of this community (Dewey 1938; Proulx 2010), performing complex roles that require ‘presence’ (Rodgers & Raider-Roth 2006). Teaching presence involves the whole self, fully alert and connected to the individual students and the group, their cognition and emotion, the subject matter and all interactions within the classroom context, responded to with compassionate, flexible, intelligent action (Dewey 1910; Rodgers & Raider-Roth 2006; Stooksberry et al. 2009). Presence develops with reflection and experience (Rodgers & Raider-Roth 2006), much like practical wisdom. Perhaps, it is this combination of presence and practical wisdom that provides the artistry, the feeling, that ‘just know’ factor, that allows teachers to go beyond competency to attain excellence (Dewey 1910; Eisner 2002; Lunenberg & Korthagen 2009; Fitzmaurice 2010).

Limitations and further work

Categories are abstractions imposed by researchers for tidiness (Bateson 1972; Ricca 2009), conceptual understandings trail life (Thomas 2006), and all words come laden with prior meanings (Bakhtin 1986), which will inevitably affect our understanding of our subjects, and our readers’ understanding of us. Objectivity is never easy, especially so for practitioner researchers, and we acknowledge that our own perspectives suffuse our work. Of course, given the added value of practitioner–researcher viewpoint, this could be considered both strength and limitation.

Most people are poor at self-assessment (Kruger & Dunning 1999; Sui & Reiter 2009), and what they tell researchers may be new to themselves, ambiguous and not fully representative of underlying conceptions (Mauthner & Doucet 2003; Roth 2007). There is also uncertainty about the relationship between teachers’ beliefs about their actions and their actual practice (Norton et al. 2005; Ben-Peretz 2011; Peeraer et al. 2011), hinted at by the similarity of teachers according to our qualitative findings when compared to the differences in the quantitative results. Thus, we are unable to directly link these qualitatively described attributes of good remediation to students’ success and failure rates. Additionally, complex systems might be better studied at the level of their emergence, where interactions are visible (Jörg et al. 2007; Kauffman 2008; Radford 2008). This suggests that a fruitful area for further work would be direct observation of classroom teaching, which would allow comparison of perceptions with more objective reality (Wittgenstein 2006; Ben-Peretz 2011), may help us identify the most important attributes for student success, and could also result in more specific guidance for supporting novice teachers’ development of practical wisdom (Lunenberg & Korthagen 2009). Indeed, both the teaching and the remediation course itself may benefit from direct observation of the enactment of the different lessons included in the course handbook (Winston et al. 2010a), since observation of and learning from small changes may be the best way to improve an intervention (Morris & Hiebert 2011).

Conclusion

Although human functioning in complex systems is unpredictable (Kauffman 1995), and absolute educational laws may never be established (Eva 2008), we are now able to offer a tentative theory of remediation applicable to one particular context of at-risk medical students. Remediation should support emotional needs and foster cognitive and metacognitive skills for self-regulation and critical thinking. This is best done through a cognitive apprenticeship within a small community of enquiry that motivates and challenges the students. This community should foster curiosity and joy for learning, using collaborative exploratory dialogue that revolves around their experiences of failure, success and deep engagement with the subject matter of their courses. Teachers of remediation can mediate these processes through embodiment of five core roles; facilitator, nurturing mentor, disciplinarian, diagnostician and modeller of desired skills, attitudes and behaviours. This combination demands high levels of teaching presence and practical wisdom to flexibly manage a complex and uncertain remediation process.

While broader applicability remains to be tested before we can generalize, this practical theory draws together elements of teaching excellence supported by a wide range of literature to create a unique blend tailored to the high cognitive and affective demands of remediation in medical education, and promises value in remediation in other higher educational contexts.

Acknowledgements

KW would like to thank Mrs Catherine Schipul: our daily conversations about the details of teaching have been invaluable to this study.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.