How to use polarity thinking^TM to manage tensions between accountability and learner agency when using a multipurpose portfolio?

ABSTRACT

Portfolios are often implemented to target multiple purposes, e.g. assessment, accountability and/or self-regulated learning. However, in educational practice, it appears to be difficult to combine different purposes in one portfolio, as interdependencies between the purposes can cause tensions. This paper explored directions to manage tensions that are inextricably linked to multipurpose portfolio use. We used a systems thinking methodology, that was based on the polarity thinking^TM framework. This framework provides a step-by-step approach to chart a polarity map® that can help to balance the tensions present in specific settings. We followed the steps of the framework to chart a polarity map for multipurpose portfolio use. Based on literature and our prior research, we selected one overarching polarity: accountability and learner agency. This polarity seems responsible for multiple tensions related to multipurpose portfolio use. We formulated values (potential benefits) and fears (tensions that can arise) of the two poles of this polarity. Then, we organised a session with stakeholders who work with the portfolio of the Dutch General Practice speciality programme. Together we formulated action steps and early warnings that can help to balance accountability and learner agency during multipurpose portfolio use. In addition to previous recommendations concerning portfolio use, we advocate that it is important to create a shared frame of reference between all involved with the multipurpose portfolio. During this process, the acknowledgement and discussion of tensions related to multipurpose portfolio use are vital.

KEYWORDS:

• Portfolios
• accountability
• agency
• polarity
• tensions

Introduction

Over the last decades, portfolios are implemented in medical education with the aim to support the qualification of competent, lifelong learning professionals [1,2]. The use of portfolios is not an isolated practice within medical curricula, instead it strongly relates to other current educational practices. Two of these practices we would like to mention here specifically: competency-based education and programmatic assessment.

Competency-based education (CBE) is centred around the competencies that are considered vital to the field being educated [3]. CBE is outcome-based, as it prescribes the competencies that learners ought to possess in order to graduate. It is therefore considered more flexible and learner-centred than traditional education that also specifies how and when learners (and teachers) should achieve the desired results [3]. Programmatic assessment aims to combine two purposes of assessment: learning and decision-making. These aims are pursued by developing an integrated programme of assessment that is intertwined with the curriculum, rather than assessing knowledge and/or skills (once at the end) of individual courses [4]. In light of these educational practices, portfolios have been introduced within medical education around 1990 [5]. Portfolios are often integrated into the assessment programme and may fulfil a role for formative assessment, summative assessment and/or accountability [1,2]. The exact role of the portfolio within the assessment programme may vary between the portfolio being the place where evidence for assessments and/or assessment results are collected and the portfolio content itself being assessed. Furthermore, portfolios are used to support self-regulated learning (SRL), which refers to ‘the degree to which students are metacognitively, motivationally, and behavio[u]rally active participants in their own learning process’ [6]. Portfolio use is considered to support various SRL processes, in particular feedback seeking, reflection, self-assessment, goal setting and monitoring [1,2]. It is a common practice to implement portfolios that are intended to support multiple purposes [5].

However, in educational practice, it proves to be complicated to serve different purposes with one portfolio [1,7–9]. This is exemplified by perceptions of learners, who often experience their portfolio to be an instrument for faculty that can assess their performance with using the portfolio rather than a tool that is beneficial for their own learning process [10–12]. Apparently, the value of portfolio use for SRL can be compromised if assessment and/or accountability need to be served as well. Often discussed in this context is the combination of (summative) assessment and support of reflection for learning in one portfolio [1,7,13–15]. The interdependency between purposes can cause tensions during portfolio use. Although reflective skills are considered vital for physicians and therefore important during competency assessment, learners may be reluctant to document in-depth reflections in their portfolios, because they fear that exposing doubts or uncertainties may give faculty members reading the contents of the portfolio a negative impression of their competence level, leading to an unfavourable assessment [16,17]. Insights of researchers and educators differ whether and how to deal with these tensions [1,7,13,15].

The approaches that have been suggested to deal with tensions of multipurpose portfolio use (e.g. the advice to stop assessing portfolios in order for reflection to take place) do not acknowledge the complexity of the problem at hand. Complexity theory describes that complex problems are signified by a large number of interconnected elements, that share interdependencies, are dynamic and intransparent [18]. A pitfall is the expectation that a single, fit-for-purpose solution can solve a complex problem [19–21]. The mentioned solution regarding multipurpose portfolio use shows that such solutions do not sufficiently address the interdependent and dynamic nature of a complex problem: when portfolios are not assessed, learners can be left with the feeling ‘what am I doing it for?’ [9,22].

In contrast, systems thinking methodologies are considered suitable to deal with complex problems [21,23], as these ‘recognize that factors relating to a problem are connected and dynamic, are considerate of diverse perspectives, and are aware of boundaries surrounding issues’ [21]. An approach that fulfils the criteria of systems thinking methodologies is described by the polarity thinking ^TM framework. This framework focuses on polarities (also described as paradoxes or interdependent pairs) and offers an approach to find a balance for both poles of the polarity to co-exist [19,20].

This paper describes how the polarity thinking^TM framework was applied in order to explore possibilities to manage the tensions that are inextricably linked to multipurpose portfolio use.

Methods

Study context

This paper is part of a research project that studied the use of portfolios for the support of SRL within the Dutch General Practice (GP) speciality programme. During this project, a realist literature review, portfolio content analysis and qualitative focus group study were conducted [24–26]. The results of these studies informed the formulation of the polarity map that is described in this paper (Appendix A provides a summary of these studies). In particular, the results of the focus group study were an important source for our discussions and decisions during the process of polarity map building. During this study, we organised nine focus groups to gather the perspectives on portfolio use of trainees, supervisors and faculty from three Dutch GP training institutes.

Methodological approach

The polarity thinking^TM framework was selected as it fulfils the criteria of a systems thinking methodology, which was considered suitable due to the complex problems related to multipurpose portfolio use. Central to the polarity thinking^TM framework is the concept of polarities, which encompasses different poles of a situation that appear to be in competition, but in fact depend on and need each other to thrive [19,20,27]. Polarities are omnipresent and can be found within individuals (inhale vs. exhale), organisations (cost-effectiveness vs. service excellence) or nations (social security vs. self-reliance). Naturally, also medical education comes with polarities, as already shown by Govaerts et al., who outlined and reflected on two polarities of assessment: standardisation vs. authenticity and numbers (quantitative data) vs. words (qualitative data) [19].

Polarity thinking^TM is based on the observation that tensions between the poles of a polarity are usually accommodated by solutions that focus on one of the two poles, which are referred to as either-or solutions. However, these types of solutions are generally too simplistic, as they do not take into account the interdependency between the two poles of the polarity. For this reason, polarity thinking^TM suggests that instead of an either-or solution, one should opt for a both-and approach, i.e. a balance in which both poles can co-exist [19,20]. In order to reach such a balance, a polarity map is composed through a step-by-step approach. Figure 1 shows the polarity map with a description of the different components.

Figure 1. A polarity map

Study procedure

Govaerts et al. described an approach to formulate a polarity map that aligns with the principles and instructions of the polarity thinking^TM framework [19,27]. Four members of the research team (RG, BT, SH, NS) performed these steps to chart a polarity map for the use of multipurpose portfolios in medical training. As described by the approach of Govaerts et al., stakeholders were involved during the third step of the process [19]. The other members of the research team (JM, MS, AT) were involved on a consultation basis; they were regularly asked for their ideas and feedback.

Step 1: seeing

This step entails the appraisal of the polarities present in a specific setting. In accordance with the instructions of the polarity thinking framework^TM, we first formulated issues of multipurpose portfolio use. These issues had been identified during literature study and the other studies of this research project, thereby describing the challenges we had encountered (problem) and the intended purpose of portfolio use (possibility). We primarily used problems that we encountered during our focus group study, but made sure that these problems were also described in the extended literature on portfolio use.

Secondly, we formulated an explanation for these problems (moving FROM) and the situation that one intends to reach through portfolio use (moving TO). These actions were performed individually by the team members (RG, BT, SH, NS). Thereafter, the proposed polarities of the different team members were discussed, which resulted in the selection of one overarching polarity. We quickly reached consensus, as the selected polarity was identified as the most important hindrance of portfolio use during our focus group study and we also encountered this polarity in papers that were read during our literature review.

Step 2: mapping

During this step, the components at the centre of the polarity map (the Greater Purpose Statement (GPS), deeper fear, values and fears) were completed. First, the team members (RG, BT, SH, NS) formulated these components individually. Then, the different suggestions were assembled, discussed and adjusted during multiple iterations, until a final version was established.

Step 3: tapping/leveraging

This step encompasses the formulation of action steps and early warnings through discussion during a session with stakeholders. We wanted to weigh-in a diversity of viewpoints and therefore included representatives of relevant stakeholder groups of the Dutch GP speciality training. These representatives were purposively approached via email and asked to participate in the stakeholder session. The final participants were two residents, two faculty members, two clinical supervisors, and one team leader. They all provided consent to participate in the stakeholder session.

Due to the COVID-19 guidelines in place at the time, the stakeholder session took place through videoconferencing (Zoom) and was moderated by the first author. First, the polarity thinking ^TM framework was presented. Next, the stakeholders were introduced to the polarity map on multipurpose portfolios, which had been drawn up during the first two steps. When it was clear that all participants understood the concept of polarity thinking ^TM and the different components of the polarity map on multipurpose portfolio use, the participants were invited to propose and discuss potential action steps and early warnings. The suggestions of participants were noted and shared on screen by the moderator. After 50 min, the discussion ended, as the participants did not offer any new ideas. The stakeholder session was audio recorded and later transcribed verbatim.

After the stakeholder session, the first researcher critically reviewed the suggested action steps and early warnings, in order to make sure that they were clear (i.e. easily comprehendable), concrete (i.e. easily observable) and distinctive (i.e. only applicable to the pole they were assigned to). Some adjustments were proposed and this version was sent to the participants of the stakeholder session, together with an explanation of the changes made to the initial version. Participants were invited to reply to this new version.

Reflexivity

As described above, the system thinking methodologies are characterised by – amongst others – a recognition of multiple perspectives. We believe this aligns with a constructivist paradigm, that acknowledges the social construction of reality. Therefore, we aimed to gather varying perspectives by utilising the different experiences, opinions and knowledge of the research team and including stakeholders with various backgrounds. In concern to the research team differences were amongst others the result of our professional backgrounds – psychology (RG, AT), educational science (BT, SH, MS), health science (SH) and medicine (BT, JM, NS) – and our own involvement with portfolios, as developer (BT, SH, MS), user (AT), or researcher (all authors).

Findings

The polarity map on multipurpose portfolio use is presented in Figure 2. The (reflections on) choices made during the development of this polarity map are explained below.

Figure 2. The polarity map on multipurpose portfolio use in medical training

Step 1: seeing

The different polarities of multipurpose portfolio use that were proposed are presented in Table 1. Through discussion of this list, we recognised that the last four polarities could be seen as consequences of the first polarity, which focuses on accountability and learner agency. We defined accountability as ‘the responsibility of educators to the public’ and agency as ‘one’s capacity to act purposefully and autonomously’ [28]. When striving for accountability (polarity 1) chances are that the portfolio will be used to handle (polarity 3) portfolio content of learners in order to assess (polarity 4) in a structured (polarity 5) and data-driven (polarity 6) way. In contrast, striving for learner agency (polarity 1) will probably relate to coaching (polarity 3), reflection (polarity 4), personalisation (polarity 5) and intuitive knowing (polarity 6). The second polarity that focused on time constraints of portfolio use in a workplace setting was considered unrelated. However, given the interrelatedness, we expected that focusing on the first polarity could also inform us about the four polarities that can result from this overarching polarity.

Table 1. The list of possible polarities of multipurpose portfolio use in medical education that was proposed during the first step of the polarity map formulation.

Polarity	Possibility	Problem	Moving from	Moving To
Accountability & Learner Agency	Meaningful use of the portfolio for SRL.	Learners do not use the portfolio for personal learning (objectives), as they feel pressured to document information that is expected by others.	(Perceptions of) obligations to add portfolio forms.	Learner agency during the use/control of portfolio.
Need to prepare & Just-in-time	(Timely) embedding of portfolio processes during workplace learning.	Portfolio use is considered unfeasible in the busy chaotic workplace setting.	Perception of lack of time/taking too much time.	Natural/timely embedding in workplace learning processes.
Management/handling & Coaching	Coaching/mentoring that is aligned with the portfolio’s purpose.	Supervision meetings are used to check off responsibilities.	Perception of being kept in check.	Meaningful and supported relationship mentor-mentee.
Assessment & Reflection	Portfolio assessments are an inspiration for meaningful reflection.	Learners do not document their reflections due to fear of assessment.	Anxiety to report because of summative consequences.	Frank and open portfolio reports.
Structuring & Personalisation	The portfolio structure supports possibilities to personalise one’s portfolio.	Learners do not know how to adapt their portfolio to their personal needs due to a rigid/limited structure of the portfolio.	(Lack of) external structure limits the possibilities of portfolio use.	The portfolio supports learners to personalise their portfolio.
Data-driven & Intuitive knowing, i.e. ‘gut feeling’	Objective and subjective data are used in combination.	Information is not documented in the portfolio, as coaching and learning are considered intuitive processes that cannot be formalised.	A portfolio with limited information.	A rich portfolio that supports coaching, learning and decision making.

The description of this overarching polarity in Table 1 focuses on problems that can arise when accountability overshadows learner agency, as we had encountered this situation in our previous studies. In such a situation, portfolio requirements installed to ensure the portfolio’s value for accountability, restrict the (experienced) possibilities of learners to use the portfolio for their personal learning objectives. However, also the opposite situation might arise: portfolios that have no value for accountability, as personal choices of learners have resulted in a disparity of portfolio content. Given the importance of both poles of this polarity in medical training, we considered it important to explore possibilities to balance the polarity.

Step 2: mapping

We easily agreed on the values of both poles, as there was a high degree of similarity between the proposed values, e.g. (intrinsic) motivation was by most team members mentioned as a value of learner agency. Our proposals for the GPS were also similar, as everyone considered lifelong learning an important aspiration of balanced portfolio use. In contrast, the components on the bottom half of the polarity map required more extensive discussion, as the initial suggestions of the deeper fear and fears for both polarities were more varied. However, through discussion, we realised that these suggestions shared commonalities. For example, the following suggested fears of accountability,

• Deliberate choice for performance assessment of what is already mastered,

• Selective documentation,

• Adding socially desirable portfolio content in order to meet external standards,

• Avoidance of portfolio entries that are personal (and potentially meaningful),

were combined into the fear: learners only add portfolio content in order to meet external standards (tick-box exercise).

Our final version of the GPS, deeper fear, values and fears is shown in Figure 2.

Step 3: tapping/leveraging

During the stakeholder session, it became clear that participants recognised the selected polarity and the proposed values and fears from their own experiences with the multipurpose portfolio of the GP speciality training.

The action steps and early warnings that were proposed by participants of the stakeholder session are presented at the centre of Figure 3. While participants easily suggested possible action steps and early warnings, they often struggled to assign these to only one of the two poles of the polarity.

Figure 3. The action steps and early warnings that were proposed during the stakeholder session

The adjustments that were made after the stakeholder session and were shared with participants are explained in the boxes surrounding the table in Figure 3. While three participants replied that they agreed with the adjusted version of the polarity map, one faculty member responded specifically to the removal of the proposed early warning ‘there is little documentation in the portfolio’ by underlining that an empty portfolio could indicate all sorts of problems and was therefore interesting. Despite this response, we decided that no early warning would be included regarding the amount of documentation, the faculty member’s response did not refute the primary reason for removal by the research team (a lack of distinctiveness). The other participants did not reply to our invitation to comment on the adjusted version of the polarity map.

The final version of the action steps and early warnings is presented in Figure 2.

Discussion

Multipurpose portfolios are implemented in medical education with the aim to support the qualification of competent, lifelong learning professionals. However, in educational practice, it appears difficult to combine multiple purposes in one portfolio, as different tensions undermine the portfolio’s potential. We identified accountability and learner agency as an overarching polarity that can explain various tensions related to multipurpose portfolio use. Using the polarity thinking^TM framework, we charted a polarity map, in order to clarify the polarity and provide concrete advice on how to manage tensions that result from this polarity.

Tensions related to multipurpose portfolio use have been discussed previously, in concern to the combination of (summative) assessment and reflection [1,7,13–15]. Moreover, multiple fears described in our polarity map have been reported in studies that evaluated multipurpose portfolio use in medical education, thereby indicating imbalances between accountability and learner agency. Fears that indicate a focus on accountability at the cost of learner agency were reported most often: portfolio use as a tick-box exercise, motivational problems and/or stress and anxiety of learners [10,29–33]. However, also the opposite situation (learner agency overshadowing accountability) might occur, as studies discussed reliability and/or validity problems of portfolio assessment due to the disparity of portfolio content [34,35].

Given that multipurpose portfolios are an important component of medical curricula across the world, it is troublesome that intended purposes are often not met. As discussed, previously suggested solutions can add to imbalances between portfolio purposes, as they are usually focused on one of the purposes (e.g. rubrics that are intended to increase the interrater reliability of portfolio assessment) [36]. It is time to consider solutions that fit the criteria of systems thinking methodologies.

Implications for practice

Many of the action steps included in our polarity map are similar to previous recommendations regarding portfolio use. In line with research literature that emphasises the importance of clear guidelines, participants of the stakeholder session formulated the action step ‘provide sufficient information on portfolio use’ [5,12,37]. Furthermore, the literature stresses the importance of faculty development, which could be seen as a prerequisite for another action step ‘faculty actively engage with the portfolio and guide learners during the completion of learning cycles’ [5,13,37]. Moreover, it is not surprising that action steps have been proposed considering the content (i.e. ‘provide portfolio tools that focus on an awareness of learning’), use (i.e. ‘evaluate competences periodically’) and technical aspects of the portfolio (‘provide the opportunity to deliver content in a personalised way that corresponds with clinical practice’) [13,37].

In addition, there is one action step unique to our polarity map: ‘acknowledge and discuss the tensions between accountability and learner agency’. We advocate that it is important that all stakeholders (e.g. learners, faculty, and curriculum coordinators) exchange information on the purposes, expectations and assessment of the multipurpose portfolio in use, in order to create a shared frame of reference between all involved. During this exchange, the focus should not be on isolated aspects (or problems) of the portfolio, but instead a holistic approach is needed, i.e. systems thinking.

In these implications, we would like to highlight an important concept related to systems thinking: continuous quality improvement [23]. While (most of) abovementioned actions points will probably be considered during the introduction of a multipurpose portfolio, attention for the functioning often fades when the portfolio has become an established component of the training programme. At that point, imbalances between the different purposes, and therewith tensions, will often arise. Therefore, it is important that portfolio use is periodically evaluated. Evaluation should go beyond a strengths and weaknesses assessment, and the use of a comprehensive Plan-Do-Check-Act (PDCA) cycle is advised [38]. During this process, the action steps and early warnings of our polarity map can be helpful.

In concern to the above, we think it is important to emphasise that the polarity between accountability and learner agency is not unique to portfolio use, as it has also been described in concern to other aspects of current medical training programmes [38]. We have mentioned that medical training programmes are often defined by the use of CBE. While CBE is theoretically meant to be flexible and learner-centred, the outcome-oriented nature of CBE can also be seen as a fixed and prescriptive picture of how doctors should perform, think and act [38,39]. Another important component of many medical training programmes is programmatic assessment. While single assessments are intended to support learning, it has been reported that learners can also perceive these assessments to be high-stake (i.e. resulting in pass/fail decisions), which is stressful for learners and can reduce experienced learning opportunities [40]. Considering that the polarity between accountability and learner agency can be fundamentally present in medical training programmes, one might wonder whether it is possible to balance the polarity by means of portfolio use only. Instead, it is probably necessary to embed the abovementioned suggestions in all aspects of the training programme [41].

Strengths and limitations

Although tensions that can result from the use of multipurpose portfolios have been discussed previously [5–7,13], little guidance has been provided on how to adequately manage these tensions in educational practice. The polarity thinking^TM framework offered an approach that aligns with systems thinking and enabled the provision of concrete advice on how to deal with the complex problems related to multipurpose portfolio use.

However, the construction of a polarity map was also complex and possibilities to make alterations were limited due to the trademark agreement. Moreover, we considered one of the instructions regarding the formulation of the polarity map to be at odds with the principles of the polarity thinking^TM framework. The instructions specify that the values, fears, action steps and early warnings should be unique to one of the two poles. However, this seems to contradict the primary principles of the polarity thinking^TM framework that underlines the importance of interdependency between the two poles of a polarity. Consequently, it was difficult for participants of the stakeholder session to formulate action steps and early warnings that applied to one of the two poles only. Therefore, some suggestions were removed after the stakeholder session, as they were not considered distinctively related to one pole (e.g. accessibility of the portfolio). However, during later discussion of the polarity map, we realised that some of the action steps could be considered important to both poles (e.g. acknowledgement and discussion of tensions). Nevertheless, as we wanted to do justice to the suggestions resulting from the stakeholder session, we decided to make no further alterations to the polarity map than those shared with the participants directly after the session.

Finally, the inclusion of seven representatives related to the Dutch GP speciality programme summons us to mention that this is just one example of a polarity map on multipurpose portfolio use. The same exercise in other settings might provide different results.

Conclusions

Portfolio use in medical training has the potential to serve various purposes. However, in educational practice, it appears to be difficult to maintain a balance between the different purposes of a portfolio, as interrelatedness between the purposes can cause tensions. We considered accountability and learner agency as an important overarching polarity that can explain a multitude of tensions associated with portfolio use. Polarity thinking^TM enabled us to suggest action steps and early warnings that can help to balance accountability and learning agency. Most importantly, an awareness of the interdependency between this polarity is essential to support multiple purposes with one portfolio.

Acknowledgments

The authors would like to thank the participants of the stakeholder session for their valuable contributions.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.

Appendix A

– A summary of the studies that informed the formulation of the polarity map

Table

Realist review We performed a realist review in order to describe the evidence base that considers how and under what circumstances portfolio use supports SRL. This realist review was performed in two phases. In the first phase, stakeholder interviews and a scoping search were performed to formulate a program theory that explains how portfolio use may support SRL. In the second phase, an in-depth literature search was conducted, which resulted in the inclusion of 16 papers. We extracted context-mechanism-outcome configurations from these papers, that described in which ways and under what circumstances portfolio use resulted in changes in SRL. This process helped to establish two working mechanisms of portfolio use: documenting as a moment of contemplation (learners analyse experiences while writing portfolio reports) and documentation as a reminder of past events (learners are reminded of past experiences when reading previously documented portfolio reports). These two mechanisms can explain the positive relationship between portfolio use and self-assessment, reflection and feedback processes that was identified. However, there was limited evidence that portfolio use also supported the two other SRL constructs that health professions education literature considers related to portfolio use: formulation of learning objectives and plans, and monitoring. While this review showed that portfolio use has the potential to support SRL, it also showed that the working mechanisms of portfolio use – and with that SRL – are easily disrupted by interference of contextual factors (e.g. system of assessment) and other factors related to portfolio use (e.g. how the portfolio is used during mentoring). Many of these disruptions seem to relate to tensions between different portfolio purposes, which may undermine motivation of learners. Moreover, this review demonstrated that previous portfolio studies were falling short both theoretically – in defining and conceptualising SRL – and methodologically. Content analysis In order to develop the evidence base concerning portfolio use for the support of SRL, it was of relevance to gain insight into the presence of SRL constructs within portfolio documentation. Therefore, a content analysis of the portfolios of general practitioner (GP) trainees was performed. While SRL consists of many constructs, the focus of this study was on the constructs that health profession education considers to be supported by portfolio use: reflection and feedback (as these are expected to contribute to self-assessment), goal-setting and planning, and monitoring. A codebook was formulated by use of literature that describes criteria for good practice of these SRL constructs. The codebook was used to rate the presence of these criteria within ninety portfolios. This analysis revealed that reflective entries were scarce and mainly described events and experiences without explanations and context. Feedback was generally limited to comments on what went well and lacked specificity, context and directions for future learning. Learning goals and plans were short of specificity, but did contain challenging and consistent topics. Signs of monitoring were – limitedly – present in 75% of the portfolios. In sum, the criteria for good practices of the included SRL constructs were hardly present in the portfolios of trainees. We concluded that documentation of SRL constructs will not appear in portfolios without effort from all stakeholders involved, as documentation of SRL requires time and asks for a high level of introspection and writing skills. Focus group study The limited quality of portfolio documentation regarding SRL did not necessarily mean that GP trainees did not engage in SRL during portfolio use, as proper documentation is possibly not required for SRL to take place. In order to get a better perspective on this, a focus group study that explored the experiences with portfolio use of different stakeholders of the GP speciality training program was performed. Nine focus groups were organised to gather experiences of the three important user groups of the portfolio: trainees, supervisors and faculty. Template analysis was used to analyse the data. Three themes were identified: SRL with(out) the portfolio, stakeholder dynamics and ambiguities. Respondents were doubtful about the learning benefits of portfolio use, as most trainees only used their portfolio to check off what was considered to be required. Stakeholder dynamics clarified checking off behaviour in two ways. First, trainees often considered documenting learning activities as a superfluous, administrative activity, since the close relationship with their supervisor already adequately supported their learning process. Second, faculty often (unintentionally) took portfolio ownership away from trainees, as they instructed trainees to deliver portfolio content that was valuable for assessment. Without ownership trainees experienced little motivation and leeway to use the portfolio during SRL. Besides, ambiguities related to why and how the portfolio should be used amplified checking off behaviour, as trainees haphazardly completed portfolio reports in the hope that they would thereby fulfil potential requirements. Consequently, we concluded that portfolio use did not support SRL in the setting of the GP speciality program. Difficulties to serve multiple purposes (i.e. accountability and learner agency) with one portfolio were identified as the primary obstacle, since these difficulties resulted in competing interests and a lack of clarity concerning why and how the portfolio should be used.