The socialisation of mistreatment in the healthcare workplace: Moving beyond narrative content to analyse educator data as discourse.

Workplace learning is central to healthcare students' and trainees' education, providing them with valuable opportunities and experiences and socialising them to become healthcare professionals.1 Yet, clinical workplaces are highly complex spaces, fraught with hierarchy and (very often) workplace pressures.2, 3 Within this environment, students and trainees exhibit and encounter wide-ranging professionalism behaviours and lapses.4-6 Alongside professionalism, discrimination and bullying (indeed, all forms of mistreatment) are pervasive, having serious adverse consequences for learners.7-10 So, the learning process can sometimes come at a cost, particularly when it involves mistreatment by those in authority. This mistreatment not only affects learners' well-being but also negatively impacts patient care and safety.6 We applaud Vanstone et al,11 therefore, for contributing to ongoing conversations around student and trainee workplace mistreatment, adding to voluminous prior research and keeping the spotlight on this continuing challenge. We have been researching this topic internationally since 2007 (using the terminology ‘professionalism dilemmas’ and ‘workplace dignity’).6 Sadly, workplace mistreatment remains an ongoing issue far from resolution. Nevertheless, Vanstone et al's11 articulation of this topic propels us forward, specifically with their examination of well-worn events from the perspective of the usually silent stakeholder—the educator (rather than typically privileging learner perspectives). Their analysis, guided by constructivist grounded theory, led the authors to conclude that, excepting extreme behaviours, participants' judgements about whether events comprised mistreatment varied because of differences in individuals' sense-making. As such, developing objective definitions of what comprises mistreatment is problematic. Workplace mistreatment remains an ongoing issue far from resolution. Whereas Vanstone et al11 highlight actors' active (versus passive) constructions of their social world through experiences, we think it valuable to add the important constructivist concept of socialisation to their data analysis. Socialisation is the active process through which objective events in the world are incorporated into existing understandings to become subjectively meaningful.12 Essentially, this individual sense-making shapes who we are (our identities) from birth (primary socialisation).1 Socialisation into a profession (secondary socialisation) is particularly prominent for students and trainees during workplace learning. Here, we come to understand our professional identity, the workplace culture, and internalise (to a greater or lesser degree) cultural values and norms. Socialisation reduces uncertainty, facilitating mastery of a new environment.13 Research suggests four main elements to socialisation: (i) learning job components (task mastery), (ii) understanding one's organisational role (role clarification), (iii) adjusting to the organisation's culture (acculturation) and (iv) developing co-worker relationships (social integration).13 The longer someone inhabits a profession (e.g. physician, nurse) or specific organisational environment (e.g. surgery, general practice), the more inculcated they become within that culture. And, if that profession or organisational culture could be described as abusive to a reasonable outsider, the more acclimatised they become to abuse. Given this background, we believe that differences in the data presented by Vanstone et al11 can (at least in part) be explained by the length of time participants had spent in specific workplace settings. The longer someone inhabits a profession (e.g. physician, nurse) or specific organisational environment (e.g. surgery, general practice), the more inculcated they become within that culture. It is impossible to contextualise Vanstone et al's11 data according to potential narrator levels of socialisation because they do not report the demographic features of their sample (to preserve confidentiality). It is also difficult to make sense of data privileging socialisation whenever investigators focus on the surface level of what participants said. This is commonplace amongst qualitative researchers in health professions education, and we also begin our analysis at this important foundational level.9 However, by digging deeper, and focusing attention on the process of talk (i.e. how participants speak, what they do with words), it is sometimes possible to reveal a different, richer story. This is because narratives are products of someone's social construction (the content of talk), as well as reflecting the construction process itself (how people talk). So, rather than analysing these narratives as if they are true representations of a real event (the so-called discourse as data perspective),14 there is value to analysing them as social constructions in a particular context (a data as discourse perspective).14 If one considers such narratives as social constructions in particular contexts (including degree of narrator socialisation), it can enable one to think deeply about the important impacts of socialisation on the people constructing the narratives and through the narratives themselves when they are narrated to others. By digging deeper, and focusing attention on the process of talk (i.e. how participants speak, what they do with words), it is sometimes possible to reveal a different, richer story. To provide an example of working with data as discourse, we discuss the educator's narrative in Vanstone et al's11 article. Here, the educator narrates a justification for why it may be appropriate to raise your voice in a surgical operating room: ‘If somebody is in the middle of an [operating room] and is about to do something that he must stop doing immediately and you feel you need to raise your voice to make [the undesired behaviour] stop. Just like a child who is going to touch the stove, you may have to swipe their hand away because they're going to get hurt. To me, that's not mistreatment’. On reading this excerpt, we immediately wanted more context about the excerpt to better understand it. What went on beforehand? How did the conversation get to this point? Who is the narrator? Who is the story protagonist? But even without context, we can see how the narrator discursively constructs the information as truth. This is revealed though the use of the pronoun you in the excerpt (as bolded by us). You can be used referentially (i.e. person in front of me) or impersonally (i.e. everyone and anyone).15 When read carefully, we can see how the narrator uses you both referentially and impersonally as they justify an act of shouting (softened to voice raising) and slapping away a learner's hand (softened to swiping). Furthermore, both impersonal uses are followed by relatively softer, hedged verbs (feel, may) rather than harder and more directive verbs (will, should, must). Together, this suggests the narrator is using soft persuasion to convince the listener that their argument is rational. They then employ multiple analogies—likening the student to a vulnerable child needing protection from danger and likening surgical equipment to a hot stove. Ultimately, these analogies serve to justify the surgeon shouting at a learner and slapping their hand. The narrator ends their talk by owning their claim: This is not mistreatment. We cannot know from the paper how our interpretation reflects the data, but we offer it to ask, more generally, what it could mean for further research in this area. Admittedly, this is just one, acontextual narrative, from which we extrapolate to make points about the importance of socialisation of mistreatment in the healthcare workplace and of moving beyond the content of talk to analyse educator data as discourse. Our central argument is that to really understand what is happening in important data such as these, we encourage researchers to think deeply about what social constructionism means to optimise the richness of the data they collect. All too often qualitative researchers in health professions education jump to surface-level meanings, focusing on the content of talk. But language is more than this: We do things with language: ‘Language is not simply a neutral medium for generating subject knowledge, but a form of social practice that acts to constitute as much as to reflect social realities’ (p. 119).16 It is crucial, therefore, that researchers consider the constitutive power of their textual data, thinking about and analysing it for its content and linguistic messages.4, 15 As we have written previously, constructionism values language, dialogue and context.17 All too often qualitative researchers in health professions education jump to surface-level meanings, focusing on the content of talk. But language is more than this: We do things with language. If we ignore language, dialogue and context when exploring the persistent problem of learner mistreatment within the healthcare workplace, we risk being forever stuck in a cycle of reporting student mistreatment without understanding fully why it is legitimised by those socialised into this culture. More importantly, we risk not being able to break this cycle to advance towards cultural change in healthcare education. If we ignore language, dialogue and context when exploring the persistent problem of learner mistreatment within the healthcare workplace, we risk being forever stuck in a cycle of reporting student mistreatment without understanding fully why it is legitimised by those socialised into this culture. Lynn V. Monrouxe: Conceptualization; writing—original draft; writing—review and editing. Charlotte E. Rees: Conceptualization; writing—review and editing. Open access publishing facilitated by The University of Sydney, as part of the Wiley - The University of Sydney agreement via the Council of Australian University Librarians.