The term problem-based learning (PBL) is open to misuse, 1 and educational researchers may well compound confusion by not clarifying its use in their specific curricular settings. 1–3 Furthermore, the literature gives few insights into how much PBL tutors share a working understanding of the term.
Traditionally, medical students acquire knowledge, as directed, within subject boundaries. For the purpose of this report, PBL refers to the educational method and philosophy of students learning what they identify as relevant to understanding PBL case scenarios. Such learning is problem-first learning, 1 in which practical application drives rather than follows knowledge acquisition (integrating across traditional subject boundaries):
Problem-based learning promotes efficient knowledge acquisition and use via self-directed collection, critical appraisal, and synthesis of evidence (and active self-evaluation) according to learning objectives generated by students from facilitated small groupwork on a clinical scenario. 4(p. 53)
[Students' activities in Liverpool PBL sessions]… Look for phenomena requiring explanation… Investigate prior knowledge and experience… Volunteer shared learning objectives… Explain the essence of the case scenario… Reflect and evaluate…” 5(p. 658)
In PBL, students should be self directed, hence tutors should be “trying not to teach” as such.
PBL is not synonymous with problem solving, so how do they relate to each other? The literature about clinical problem solving highlights hypotheti-codeductive or backward (from hypothesis to expected data) reasoning: perceiving and interpreting the situation, generating hypotheses, appraising data, formulating the problem, and making diagnostic and/or management decisions. This “works up” clinical cases for diagnosis and management. 2 The literature about critical thinking emphasizes creativity over systematic stages. Critical thinkers explore alternatives, are imaginative, flexible, and sceptical, and reflect habitually on their thinking and progress (metacognitive monitoring). 4 The problem-solving framework of PBL sessions does not necessarily improve problem solving directly (without targeting those skills specifically), but PBL might allow students' problem-solving processes to access knowledge better. 6
The relatively few studies that have highlighted PBL tutors' opinions about transforming to a problem-based curriculum 2,7 were mainly questionnaire based. PBL tutors generally favor PBL over traditional approaches 3 for personal satisfaction, and students' interest, reasoning, and clinical preparation. 8,9 Competing commitments and having to suppress their content expertise potentially frustrate tutors, 10 and they worry when and how to intervene during PBL sessions. 7
Much rarer in the literature, two semistructured interview studies focused on how foundation tutors (those facilitating the first-ever PBL groups in newly implemented problem-based curricula) conceptualize PBL and their motivation to be tutors. In 1991, Abdulrazzaq and Qayed reported that at the new United Arab Emirates problem-based medical school five of 18 foundation faculty were hostile to the PBL philosophy and five were unimpressed. 11 In 1988, Wilkerson and Maxwell reported on 27 of 31 PBL tutors in the first two years of Harvard Medical School's New Pathway curriculum. They found that these early tutors volunteered mostly because of their interest in education (89%), reform, and colleagueship. 12
In 1996, after about six years of planning, The University of Liverpool Faculty of Medicine admitted its first 208 entrants to a transformed problem-based medical curriculum. Consistent with General medical Council recommendations to reduce factual overload and promote curiosity, a five-year “coreplus-options” curriculum replaced the longstanding traditional curriculum. In the core curriculum of years one to four, PBL sessions drove learning of the knowledge base under four themes:
- Structure and function in health and disease (basic and clinical science)
- Individuals, groups, and society (behavioral science)
- Population perspective (population science: public health, epidemiology, statistics)
- Professional values and personal development (legal, ethical, professional issues)
Two-week integrated module, focused on PBL clinical scenarios (with suggested reading or computer-based materials) replaced the former preclinical—clinical and inter-subject boundaries and lecture-based format. First-year students had one (non-compulsory) 40-minute plenary session daily. Clinical and communication skills (simulation) training and PBL provided clinical context from the outset, blending progressively with clinical contact from year two, and culminating in a very vocationally orientated fifth year. Overall, curricular options comprised six special study modules and an elective placement.
The PBL tutors were pivotal in facilitating the breadth and depth of students' learning, yet most had graduated from and worked within traditional, lecture-based higher education. The PBL tutors' development involved one to two days' induction, then monthly one-hour update sessions, and started one and a half years before the curriculum. 13 PBL exposes students to the fallibility of their educators' subject knowledge and the medical knowledge base generally. The literature provides little insight into PBL tutors' collegial (dis)harmony about their role, yet this inevitably influences students' learning and collaboration. This preliminary study explored how a cohort of foundation PBL tutors characterized and made sense of PBL while trying not to teach.
The director of medical studies, the director of PBL, and a professor of medical education at The University of Liverpool Faculty of Medicine approved the study.
The study sample consisted of all the first-ever year-one, semester-one PBL tutors (October 1996—January 1997) in the problem-based undergraduate medical curriculum except the author, who was also a tutor (n = 34 tutors for 31 groups).
Literature reviews, 1,4,5 personal experience tutoring during the semester, and informal discussions at tutor-development sessions informed the design of a semistructured interview guide for telephone use. This was piloted and slightly amended with the first two interviews. (Beyond this cohort there were no other tutors for such piloting.) Openended questions explored how the tutors conceptualized PBL, problem solving, and the four curricular themes (results to be reported elsewhere 14). The interviews also involved closed and open-ended questions to ascertain tutors' personal details and experience.
An invitation letter outlined the study and assured confidentiality. Responders signed and returned a reply slip, in an enclosed envelope (stamped or internal-address), agreeing to an approximately 20-minute telephone interview (100% response rate). Interviews were booked from one to 21 days ahead by telephone, in writing, or in person, allowing the tutors to clarify queries. Non-responders received reminders after a month.
The author interviewed the tutors between February and June 1997 (semester two), interviewing those who were tutoring during semester two before tutoring restarted. All responders permitted audiotaping, receiving assurance (at a time of great uncertainty for staff) that the audiocassettes would be erased after their use in supplementing contemporaneous written notes. The author summarized for interviewees their descriptions of PBL before exploring its relationship with problem solving, clarified questions, and prompted full explanations. The interviews lasted 15–60 minutes (median 20 minutes).
The author transcribed the core answers with word processing, moving any text that pre-empted later questions to the relevant sections. Each interview received a unique record number.
The author read the transcripts repeatedly, color coding patterns and counting instances of common or important issues, and noting outlying elements. 15,16 Emerging themes were revisited and discussed over selected transcripts with an informed supervisory colleague (JS) from another faculty, maintaining responders' anonymity. Discussing some preliminary results at a tutor-development session stimulated further analysis. This iterative inductive analysis took a traditional empirical approach (aware of the lessons to be learned from grounded theory). 16,17
Quantitative data were analyzed with a statistical software package, producing simple frequency tabulations.
Characteristics of Tutors
Of the 34 tutors, 26 (76%) were men, 23 (68%) were medically qualified, graduating between 1964 and 1986; and 17 (50%) were aged 35–44 years (range 30–34 to 55–59). Collectively, they had accumulated 424 years of experience educating medical students (range 0–30, median 12.5). The tutors' backgrounds differed. The 25 university employees included two basic scientists (medical), nine basic scientists (non-medical), seven clinical doctors (mostly hospital consultants), two public health doctors, four medical educationalists (medical: including two practicing general practitioners), and a medical educationalist (non-medical). The nine non-university employees were a specialist nurse, four clinical doctors (hospital consultants), and four general practitioners.
For the 31 groups, 19 tutors (56%) facilitated all 21 sessions, ten tutored 19–20 sessions, and the others tutored five, nine, nine, 12, and 15 sessions, respectively. Five sessions had a reserve (apparently untrained) tutor or no tutor. Only ten tutors (29%) reported personal experience of PBL, but most described facilitating small-group work not involving PBL.
Essential characteristics of PBL (see Table 1).
Most of the tutors conceptualized PBL as being student-centered (68%) and involving small-group work (53%), active learning (41%), and the tutor (41%). They focused on its philosophy, e.g., activating prior knowledge and acquiring knowledge (efficiency, contextualization, integration). Their descriptions varied considerably in detail, occasionally betraying incongruous traditional approaches (but without any suggestion of subterfuge) (Table 2). Only two tutors mentioned the evaluative (reflective) component of PBL. Two tutors characterized PBL as not being predominantly about problem solving.
Main advantage and disadvantage of PBL (see Table 1).
For the main advantage, the tutors highlighted maintaining students' interest and enthusiasm (24%), promoting lifelong learning skills (21%), being active (18%), and being student-centered (15%). For the main disadvantage, they highlighted generating fear of knowledge gaps (32%) and being resource-intensive (24%).
Changes and expectations.
Responding to “How has your view [of the essential characteristics] of PBL changed (if at all) since you started as a PBL tutor?”, 19 tutors (56%) reported changing their views of PBL. Thirteen (38%) denied changing their views, and two remained evasive. Eighteen whose views had changed expanded, nonspecifically, about various concepts or practical implementation issues; 19 of 38 (50%) of these comments described reconceptualizing PBL. One tutor commented, for example, that Maastricht students “… go out to answer that problem whereas in our scenarios, there are often a dozen different aspects that the students could look up.” Another found it
… a lot more difficult… [because (i) when and how to intervene is problematic, and (ii) the expertise required to intervene properly (e.g., imagine being in an English or French tutorial; e.g., George Eliot's novels in an English tutorial)]; you don't have the jargon, the understanding, or the knowledge framework to be able to intervene properly… [extract]
Only one tutor was negative about PBL throughout the interview:
Absolutely not, no…. views not really changed at all. I'm still not convinced that PBL, despite the fact that [I will tutor again]… is the proper way of teaching.
Responding to “Has the PBL tutoring experience matched your expectations?”, three tutors avoided disclosing prior expectations. For 29 tutors (85%), their experiences matched or surpassed expectations, including one disliking tutoring, personally, yet remaining positive:
… I am more convinced that students can learn something this way. I'm much less convinced that I'm an appropriate person to be a PBL tutor. [extract]
Twenty tutors elaborated. Of their 46 positive comments, 20 (43%) involved students' progress, and 18 involved tutors' enjoyment and personal development. Their 51 negative comments were more disparate, mostly involving PBL's implementation (12) and time commitment (8). One tutor found it “very difficult to get down to concentrating on anything else like writing a paper after [PBL]… You're just in a totally different mode.” Even the tutor most negative and unconvinced about PBL “… quite enjoyed it….”
Two remaining tutors' experiences fell short of expectations, including the tutor previously finding it “… a lot more difficult….” That tutor blamed having insufficiently broad content expertise to facilitate stalled or inaccurate discourse (i.e., “… It's quite easy to keep asking the question ‘why?’” without knowing when to stop). The other tutor described a dysfunctional group, but said “… there was no crisis; we… all worked through; it could have been more fun [and in more depth, but was adequate].”
Reasons for volunteering (see Table 3).
The 34 tutors gave diverse reasons (three each) for agreeing to tutor, mostly citing: curiosity generally, educational interest, and personal development opportunities, 50%, 24%, and 24%, respectively.
Characteristics of a good PBL tutor (see Table 3).
Most commonly, the tutors characterized a good PBL tutor as knowing when and how to intervene, empathizing with students, and being enthusiastic, 41%, 29%, and 24%, respectively. The tutors did not focus primarily on their subject knowledge.
Comparisons with other tutors and Faculty of Medicine.
Responding to “How do you think that your view of PBL compares with other first semester tutors?”, 12 tutors (35%) said that their colleagues' views differed widely and six (18%) said that their views converged. Despite the training events, the other 16 tutors (47%) felt that they had not exchanged views enough to comment.
Twelve tutors (35%) disclosed suspicions about colleagues still “teaching” traditionally:
I suspect that less than 50% of the tutors had as clear an idea of what they were going into… [also based on students' feedback from other groups]… very much… hearsay [but] some tutors were very directive [and]… actually brought in lecture notes… gave them information. [extract]
[… Some have found it difficult to step back and not do a] straightforward question and answer tutorial slipping back into the old mechanism… substantially… most… will agree with [me] at least in part… —[… Some] say the words… whether… whether they've actually got that in their bowels….? [extract]
Responding to “How do you think that your view of PBL compares with the Faculty of Medicine?”, the 19 of 34 (56%) tutors who clearly disagreed with the Faculty of Medicine generally perceived that it initially overemphasized non-intervention:
… probably a bit more directive than the original message but I think that the message is shifting as a result of our experience… [at first], it was quite sort of loose and a lot of people thought it was a question of sitting there and not saying anything… [now] a clearer message coming through that that's not what it is. [extract]
Comparisons with students.
In their responses to the question, “How do you think that your view of PBL compares with your Semester 1 group of students?,” 15 (44%) referred to students adapting over one semester:
They became very “twitchy” when they were a bit unsure of how much to learn in a formal sense… I think they were getting used to it after the first five or six weeks and… all bar one were enjoying it… One… I think… may well still be twitched about it… ha ha.
Problem solving and its relationship to PBL (see Table 4).
All except one tutor discussed problem-solving skills. Most commonly, their generally impromptu descriptions conveyed problem analysis (74%), data collection (32%), and an organized process (29%):
It is the ability to approach a problem or situation to distil the essence of the problem from the… irrelevance around it, decide what information is needed to solve that problem, go and get that information, if it's not immediately at hand, and then see how… [to] formulate a… hypothesis or course of action and see how that… bears up with… experience and the information that's been obtained.
… looking at a problem and analyzing it, [seeing] what the problem actually is; [first] do you have the information that will help you solve that problem? [Second], what information is it that you require and where are you going to seek out that information from? [Then] knowing that you've actually solved the problem at the end of the process.
Only one tutor mentioned “thinking creatively.” Half of the tutors referred to outcomes, mostly to a single outcome, e.g., “… a clear endpoint—you know… actually coming out with an answer.” Seven tutors (21%) added theoretical or judgmental comments, e.g., that such skills are: task-specific “from my reading” and nonexistent generically; or about the “novice versus expert approach to things… it lies there somewhere.” Nine tutors (26%) invoked relevant practical situations, mostly clinical, e.g.: “[Problem solving] conjures up… 42 surgeries a week… about 300 consultations, ha ha” or adapting management to individual patients. Two broadened the definition of problem solving:
[It] covers anything from… “what am I going to eat tonight?” to solving the problems of the Liverpool economy.
It's like seeing a box of matches fall out on the floor; you have to line them up in a way that they can be put back in the boxes… the way you want them with all the heads up at one end…
The tutors answered “Do you think that problem-solving skills: can be taught… ?” and “… are transferable?”, and “How does ‘problem solving’ per se fit into you view of PBL?” rather superficially. Half of the tutors, 17(50%), agreed uncritically that such skills were teachable, but a further nine (26%) queried the word “taught” (preferring “improved,” “practiced,” “acquired,” or “developed”), including one tutor highlighting conscious deliberation:
No… wait a minute, taught? Ha! They can be acquired in a teaching situation… not in a lecture… If as a facilitator you reflect back to them [if working in a group]… “What have you just been doing?” “Why did that work?”… then [it helps them to] recognize it to use it again. It is like a good tennis stroke… when it goes right you say “Why did that go right?” [and then they identify it].
Four others referred to learning by “doing” or “experience.” The remaining four had reservations, two mentioning the content specificity of the skills.
Without further explanation, 19 tutors (56%) agreed that problem-solving skills are transferable. Only six (18%) mentioned content specificity or little evidence of transfer.
Over half of the tutors considered problem solving to be central to PBL, as if the shared word, problem, were self-explanatory. One tutor said, for example, “Yes. The students were presented with problems… and the two to me seem intrinsically bound.” The 26% of tutors conceptualizing PBL around knowledge acquisition more than problem solving gave more considered responses, e.g.:
… [they're going in] slightly different directions… [in PBL] trying to generate lots of options and a learning agenda… [in problem solving similar early on but then “hone down.” PBL is going wide, coming in a little bit but] staying broad; [problem as a basis, not necessarily as something to solve; can use problem-solving skills to answer learning objectives; PBL is a learning vehicle.] I don't think that PBL is about problem solving… although some of the skills that they're picking up… are going to help them… with differential diagnosis. [The endpoints are different.]
It doesn't. I don't think PBL is about teaching problem solving. [I have a major difference with Barrows' approach of using the approach to teach problem-solving skills.] I use PBL to teach the basic knowledge that people need… as a curiosity generator; not to teach problem-solving skills.
The former (plus a further, evasive response) suggested that problem solving could be acquired in passing. The latter was from one of the two tutors who characterized PBL as not being about teaching problem-solving skills.
… Oh, and by the way (see Table 5).
One tutor dismissed as “jargon” the author's introductory preamble about “… PBL process (both as method and philosophy)…,” then continued enthusiastically and helpfull, conversant with such philosophy. The tutors' closing words mixed contemplation, confession (“… If I'm honest, it's not as satisfying as giving a good lecture…”), commitment, concern, congratulation, and confusion.
This descriptive study gives unusual insight into how, in a problem-based undergraduate medical curriculum, the foundation PBL tutors made sense of their new role trying not to teach. Generally younger than Abdulrazzaq and Qayed's sample of male foundation tutors (of whom 17 of 18 came from other countries), 11 the tutors in this study also included some of the main curricular innovators, unlike those in Wilkerson and Maxwell's study (27 Harvard foundation tutors of whom one third were “senior” faculty of undisclosed age). 12 As elsewhere, Liverpool's foundation tutors were well placed to influence the educational expectations of the first intake of students, who, in turn, could influence the expectations of subsequent students as the curriculum extended beyond the first year. Such is the context for interpreting the findings, with caveats (and counter arguments) mostly about data collection and analysis.
First, during semistructured telephone interviews there is the potential to miss important nonverbal cues, generate more hurried answers than if face-to-face, and produce self-reported public accounts that differ from private accounts depending on different degrees of acquaintance between researcher (as a tutor—colleague) and interviewees. The method was, however, convenient for tutors under pressure both on and off campus, and it allowed for clarification of misunderstandings. It also allowed the 100% response, and shorter interviews appeared to signal competing commitments rather than responders' non-cooperation or disinterest. Second, potential criticisms of the study's robustness arise because no second coder verified content categories and coding, and the group's answers were summarized separately with minimal linking of answers from individual responders (or subgroup analysis, e.g., by sex, age, background).
Nevertheless, this approach did help to preserve individuals' anonymity at a politically sensitive time in the educational history of this Faculty of Medicine, especially as the responders in a 100% response were potentially more identifiable than if sampled. The study avoided a divisive “quantitative or qualitative” stance, 16 adapting complementary elements of both approaches consistent with overall purpose. The open-ended interview questions, informed by the literature, tutors, local context, and the author's personal experience as a foundation tutor, resembled those of Abdulrazzaq and Qayed 11 (e.g., “What do you believe about… what your role, in general, as a teacher should be?”), and Wilkerson and Maxwell 12 (e.g., “Why did you volunteer to teach in the New Pathway?”). Rigor lay in systematic, iterative, inductive analysis 19 that considered exceptions and refined emerging themes informed by critical supervisory discussion. The design was comparable to that advised 16,17 and reported by others. 11,12
Several sources reinforced the findings of this study. Subsequent tutor-development sessions raised comparable issues anecdotally, as did an external evaluator's contemporaneous study of the PBL process and philosophy (using informal interviews and observation) at the end of the first semester. 20 Despite tension between qualitative or quantitative stances in providing an analytic overview versus representative findings, the study generated important messages for this Faculty of Medicine, other curricula undergoing problem-based transformation, and staff development.
Three main messages emerge from the study. First, tutors mostly articulated PBL positively as a philosophy yet missed its reflective elements. In particular, they mostly characterized PBL explicitly and optimistically in terms of its student-centered philosophy and educational purpose (and knowledge acquisition) rather than detailing specific steps (the method) within PBL sessions. Furthermore, they highlighted most of the main philosophical tenets of PBL, but underemphasized the evaluative component.
In subsequent years, anecdotal evidence emerged from both staff and students that the foundation tutors were not alone in overlooking reflective practice. Consequently, staff development increasingly encouraged tutors to use the end of each PBL session to facilitate reflection on group process and progress.
Second, these tutors were particularly challenged by their own fallibility in knowing when and how to intervene without teaching. Their accounts hinted at their own, and others', fallibility (from traditional alter egos) in trying to translate the philosophy into practice, e.g., “… If I'm honest, it's not as satisfying as giving a good lecture…” (see Table 1). They implicitly valued both knowledge of broad content and knowledge of process when they characterized good tutors as knowing when and how to intervene. This was despite their not focusing on their own or colleagues' subject knowledge per se. Abdulrazzaq and Qayed asked, “What teaching skills will be important for you as a tutor in the system?” and the responses indicated communication skills, patience, and being good listeners. 11 (Wilkerson and Maxwell did not report the answers to a similar question. 12) Schmidt and Moust's empirical model of the effective tutor incorporated both subject knowledge and social congruence. 21 They perceived that the Faculty of Medicine had initially overemphasized holding back during PBL sessions, but their experiences showed a quite active role despite not teaching content. They perceived students' or tutors' fear of knowledge gaps as the main disadvantage, reflecting the epistemologic upheaval of transforming a traditional curriculum and complemented by sparing but heartfelt mentions of personal, student, colleague, or administrative dysfunction.
Third, the tutors' internal motivation and direct experience of PBL helped balance some of the confusion with the educational rationale, highlighting possibilities for future staff development. They were, unsurprisingly, slightly bemused at how problem solving fitted into this version of PBL, consistent with semantic and conceptual confusion in the literature. The few tutors aware of the evidence base were more circumspect. Appropriately, overall, few characterized PBL mainly in terms of problem solving, yet over half agreed, cursorily, that they were intimately related. They highlighted the analytic aspects of problem solving rather than creative thinking or evaluating achievement of objectives. Different traditions in a rather compartmentalized literature emphasize creativity or clear stages in problem solving. The medical education literature tends to make insufficient links between various descriptions of problem solving, critical thinking, clinical diagnosis, and reflective practice. 22 Mostly, the tutors in this study volunteered out of general curiosity (about something new), educational interest, and seeking personal development, much like Wilkerson and Maxwell's volunteers (interest in education, reform, and colleagueship 12). Nevertheless, these tutors could have been stronger role models for reflective practice. Some appeared to avoid their own self-directed learning and sharing experiences about PBL. Direct experience of PBL, however, did engage their interest and helped them to start making sense of the concept. Over half reported reconceptualizing PBL over that semester. (Bernstein et al. reported from an empirical study that Toronto tutors and medical students became more positive after personal experience. 23)
As suggested from other curricula, 3,8,9 most of the tutors were enthusiastic, were committed, and valued the experience despite discomfort at relearning how to be an educator. The most unconvinced and negative tutor still enjoyed the experience.
This group of PBL tutors did not appear to be playing subversive language games to undermine curricular reform, 24 but were merely struggling through the PBL conceptual fog 1 with old intellectual baggage. Local staff development needed to address their group facilitation skills, incorporating evaluation and reflection in PBL, and sharing good practice and its rationale.
This preliminary work raises further questions. First, how do PBL tutors use their content expertise to facilitate learning functional knowledge integrated across traditional boundaries, and how does this differ between medical and non-medical tutors? Second, to what extent do PBL tutors act as reflective (and vocational) role models for medical students? As noted in Mayo et al.'s commentary:
[The PBL tutors] set the stage for learning and present themselves as models of the learning process. In so doing, they exercise an unprecedented and unparalleled influence on students. PBL sessions reflect the tutor's imagination, creativity, personality, and temperament. These sessions succeed or fail in direct proportion to the tutor's preparedness and training for the task, organizational abilities, inter-personal skills, and sensitivity to students. 25(p. 126)
More empirical evidence about the tutor's influence might help to adapt PBL (and the tutor's role) appropriate to student progression.