Discordance Between Resident and Faculty Perceptions of Resident Autonomy

Can Self-Determination Theory Help Interpret Differences and Guide Strategies for Bridging the Divide?

Abstract

Purpose 

To identify and interpret differences between resident and faculty perceptions of resident autonomy and of faculty support of resident autonomy.

Method 

Parallel questionnaires were sent to pediatric residents and faculty at the University of Rochester Medical Center in 2011. Items addressed self-determination theory (SDT) constructs (autonomy, competence, relatedness) and asked residents and faculty to rate and/or comment on their own and the other group’s behaviors. Distributions of responses to 17 parallel Likert scale items were compared by Wilcoxon rank-sum tests. Written comments underwent qualitative content analysis.

Results 

Respondents included 62/78 residents (79%) and 71/100 faculty (71%). The groups differed significantly on 15 of 17 parallel items but agreed that faculty sometimes provided too much direction. Written comments suggested that SDT constructs were closely interrelated in residency training. Residents expressed frustration that their care plans were changed without explanation. Faculty reported reluctance to give “passive” residents autonomy in patient care unless stakes were low. Many reported granting more independence to residents who displayed motivation and competence. Some described working to overcome residents’ passivity by clarifying and reinforcing expectations.

Conclusions 

Faculty and residents had discordant perceptions of resident autonomy and of faculty support for resident autonomy. When faculty restrict the independence of “passive” residents whose competence they question, residents may receive fewer opportunities for active learning. Strategies that support autonomy, such as scaffolding, may help residents gain confidence and competence, enhance residents’ relatedness to team members and supervisors, and help programs adapt to accreditation requirements to foster residents’ growth in independence.

Self-determination theory (SDT) has been studied extensively and applied to enhance motivation in business, education, and health care settings.^1–4 Academic medicine, however, has been slow to integrate this theoretical framework into educational practice.^5–7 SDT may prove useful in illuminating the dynamics of teacher–learner interactions in the residency training environment, where evolving evaluation systems and intensifying regulatory mandates^8–10 as well as generational differences^11,12 cause stress for residents and faculty alike. Insights provided by SDT may also help to guide residency programs’ responses to the dissonance created by these pressures.

SDT is relevant to all levels of education because motivation is the primary energy that drives learning.^2,3 SDT is centered on three constructs that define an individual’s innate psychological needs:

• autonomy—the drive to be the origin of one’s behavior and to exercise free will in choosing one’s goals;
• competence—the need to feel efficacious in the actions one pursues and performs; and
• relatedness—the desire to feel connected with others, to belong to and to be valued by one’s community.

All three constructs are relevant to independent medical practice.^1,2,5 Exercising autonomy can help physicians think for themselves and pursue self-directed learning; feeling competent is key to confident decision making and leadership; and feeling a sense of relatedness can enhance physicians’ rapport with patients and the health care team.

Individuals who receive autonomy support (e.g., sensitivity to their perspectives, acknowledgment of their feelings, provision of choices, minimization of controls) from important authority figures are more motivated to pursue their goals, are more satisfied with their work and lives, and ultimately become higher achievers than individuals who are forced or persuaded to pursue the goals of others.^2,5 Adult learners, in particular, are more motivated when they are given choices and when their learning agenda is directly relevant to their professional needs and interests.¹³ Yet residency training programs are replete with external mandates that challenge the autonomy of faculty and residents. The Carnegie Foundation for the Advancement of Learning’s 2010 call for medical education reform suggests that such regulatory constraints can interfere with the individualization of learning and with physician identity formation.^10,14

Several insightful studies from the University of Utrecht have examined SDT in the context of undergraduate medical education.^6,15–17 These have found application of SDT-related methods in curricula that adopted problem-based learning (autonomy and competence), employed small-group learning (autonomy and relatedness), and exposed students to patients early in medical school (relatedness). Analogous applications of SDT to graduate medical education are needed.

The pediatric residency program at the University of Rochester Medical Center (URMC) is engaged in a longitudinal curriculum development and evaluation project that is applying SDT strategies to improve the learning climate.¹⁸ Over the past decade, our faculty have repeatedly stated at faculty meetings that residents lack initiative and are overly dependent on them in making patient care decisions. Yet in 2010, residents formally expressed their frustration to the faculty via an internal “Pediatric Resident–Faculty Autonomy Contract,” writing that they were not being given enough opportunities to make decisions, were having patient management plans dictated to them, and were not receiving enough feedback to learn from their errors. In response to these dissonant views, our curriculum has been modified to include discussions of SDT with faculty and residents, as well as resident workshops on understanding individual autonomy needs and working more effectively with faculty to ensure that those needs are met.

To clarify and interpret these dissonant perceptions of resident autonomy and faculty support of resident autonomy, we used the three constructs of SDT to create parallel resident and faculty surveys. These surveys probed areas of disagreement and asked about elements of resident education related to autonomy, competence, and relatedness. We compared resident and faculty responses using quantitative analysis of Likert scale data and qualitative content analysis of SDT-related factors in written comments.

Method

Setting and sample

Our residency program includes approximately 15 residents per postgraduate year (PGY) in the three-year categorical pediatric program and 8 residents per PGY in the four-year combined medicine–pediatric program. Categorical residents have a choice of five longitudinal tracks for their curriculum. Graduate medical education at URMC emphasizes the biopsychosocial model,¹⁹ which addresses patients holistically and encourages trainees to reflect on how to support patients in all the ways that illness may affect their lives.

This study targeted all pediatric and medicine–pediatric residents in our program in academic year 2011–2012 and all faculty who interacted significantly with these residents. Resident completion of the survey was a required program activity, but residents were given the option of refusing use of their responses for research purposes. No incentives were offered. Faculty participation was voluntary. This study was approved by the URMC institutional review board.

Survey design: Parallel questionnaires

To enhance content validity of survey responses,^20,21 the parallel questionnaires for residents and faculty were designed using an iterative consensus process among the authors, who included senior and junior faculty and fellows. Items were created according to the following primary criteria: (a) limited number of items (to maximize response rate), (b) balanced numbers of items on behaviors of residents and of faculty, and (c) inclusion of items addressing each of the three SDT domains (as explained in Table 1, footnote c).

Table 1: Description of Parallel Items and Open-Ended Items Included in Resident and Faculty Questionnaires and Their Relationship to Self-Determination Theory (SDT) Constructs, University of Rochester Medical Center Pediatric Residency Program, 2011 Surveys

To enhance validity of the response process, item wording avoided specialized SDT terminology, and the survey employed an online response format familiar to all respondents. To avoid social desirability bias, items were phrased to avoid prompts for a specific response, and respondents were guaranteed confidentiality. To enhance reliability (validity of internal structure), the questionnaires were pilot tested repeatedly by former pediatric chief residents and current fellows, refined, and re-reviewed to ensure alignment with SDT constructs.

Table 1 describes the 11 items that used parallel wording to ask residents and faculty to rate the frequency of their own behaviors or the behaviors of the other group. Because 6 of these items elicited separate faculty responses regarding the behaviors of interns (PGY-1) and senior residents (PGY-2–PGY-4), a total of 17 parallel items could be compared between residents and faculty. Ratings were based on a five-point Likert scale ranging from “very seldom” = 1 to “very frequently” = 5. Residents were required to respond to all Likert scale items; faculty could skip items because not all applied to everyone. Optional written comments were invited on 2 items on the resident survey and 3 items on the faculty survey; these were not parallel. The resident and faculty surveys are available as Supplemental Digital Appendixes 1 and 2 at https://links.lww.com/ACADMED/A237.

Survey distribution

Residents and faculty were sent informational letters that explained the survey and its goal: to improve residency training in our program. The letters stated that responses would be confidential (residents) or anonymous (faculty). During the study period (June–October 2011), potential participants were sent an e-mail notification with a link to the online survey at SurveyMonkey.com (SurveyMonkey, Palo Alto, California). Nonrespondents received a maximum of three e-mailed reminders.

Quantitative analysis

Distributions of intern, senior resident, and faculty responses to the 17 parallel items were compared using Wilcoxon rank-sum tests in two ways: (1) intern and senior resident responses separated, and (2) all resident responses combined. Statistical differences at the level of P < .05 were considered significant.

Qualitative analysis

Because the items requesting written comments were not parallel across the two questionnaires, we analyzed resident comments as one set and faculty comments as another. We conducted a directed qualitative content analysis²² to evaluate SDT themes that might elucidate quantitative group differences.

To minimize generational bias, we divided ourselves into two groups of mixed ages: One group began with resident comments, the other with the faculty comments. Each author independently examined resident and faculty comments in several iterations to identify themes and participated in meetings of the full group of authors to reach consensus on common themes. Next, we independently reexamined the comments to identify SDT-related factors and illustrative comments. We reconvened to compare and agree on new themes and consider how the theoretical concepts of SDT helped to elucidate resident and faculty differences.

Results

Sample

Of the 78 residents and 100 faculty who received the survey, 62 residents (79%) and 71 faculty (71%) responded. The 62 residents included 18 interns (29%) and 44 senior residents (71%; 17 PGY-2, 21 PGY-3, and 6 PGY-4 [of 8 medicine–pediatrics residents]). None of the residents requested exclusion from the research analysis.

Quantitative results

Table 2 summarizes mean ratings of the frequency of resident and faculty behaviors by interns, senior residents, and faculty. Differences in resident and faculty ratings were statistically significant for 15 of the 17 parallel items. Both groups agreed that faculty provided too much direction with moderate frequency.

Table 2: Comparison of Resident and Faculty Ratings of Parallel Items on Resident and Faculty Behaviors, University of Rochester Medical Center Pediatric Residency Program, 2011 Surveysa

Intern and senior resident self-ratings were typically higher than faculty ratings of resident behaviors, whereas faculty self-ratings were higher than resident ratings of most faculty behaviors. This consistent pattern of differences is depicted in Figure 1, where intern and senior resident data are combined. (Analysis of separate ratings by/of interns and senior residents provided very similar results and is not shown.)

Figure 1: Comparison of resident and faculty perceptions of resident behaviors (panel A) and faculty behaviors (panel B) as rated on 2011 parallel surveys, University of Rochester Medical Center Pediatric Residency Program. Data are mean ratings by group on a Likert scale, where very seldom = 1 and very frequently = 5. P values compare all resident responses (interns and senior residents combined, n = 62) and faculty responses (n = 71) regarding all residents. Resident behaviors (panel A) were rated in three settings: during admissions, during rounds, and when cross-covering. In panel A, “HPI” refers to resident presentations (answering pertinent questions regarding the patient’s history of present illness [HPI], medications, etc.); “plan” refers to presenting a thorough assessment and plan (e.g., patient’s symptoms, differential diagnosis). For descriptions of behaviors and parallel survey items, see Table 1.

Qualitative findings

Seven major themes emerged from our initial, open-ended qualitative analysis of resident and faculty written comments: (1) faculty direction of residents, (2) patient care planning and resident learning, (3) resident confidence and preparedness, (4) faculty expectations of residents, (5) cultural or generational differences, (6) time and scheduling limitations, and (7) trust and support (see Table 3).

Table 3: Themes and Representative Written Comments From Residents and Faculty on Resident Autonomy and Faculty Support of Resident Autonomy, University of Rochester Medical Center Pediatric Residency Program, 2011 Surveysa

This initial analysis of the written responses confirmed our expectation that resident autonomy was a contentious issue in our residency program. We next proceeded to a directed content analysis of the themes and comments in relation to the SDT constructs of competence, autonomy, and relatedness. Our findings are presented below. (Parenthetical numbers and letters indicate pertinent themes and comments, respectively, from Table 3.)

Faculty reasons for limiting residents’ exercise of autonomy.

Faculty agreed that residents were no longer given (or had failed to “earn”) opportunities to exercise the autonomy that they themselves had enjoyed as housestaff (e.g., 5a, 5b). About one-third of faculty comments described worrisome inconsistencies among residents. For example, faculty complained:

I think that the quality of residents is quite variable and between the shifts, the cross-coverage, etc., it is hard to really feel safe giving housestaff “authority.”

Many [residents] are not invested enough and seem to have little desire to be independent and accountable.

[Some] don’t even seem to know what their job as an intern or resident is.

Collectively, faculty reported granting more independence to residents who were confident and took initiative (3d), who were committed to knowing the patient well and being accountable (1c, 2d), and who understood the rationale for their own patient care plans (2c). In SDT terms, faculty wished to see motivation and competence (both demonstrated and self-perceived) in residents in order to feel “safe” allowing residents to make autonomous decisions.

Negative effects of limiting resident autonomy opportunities.

Faculty and residents agreed that, typically, residents were given relative independence in care planning only when the stakes were low. Residents expressed frustration about faculty restrictions on their opportunities to make decisions (2a) and about unexplained changes in their patient care plans. One resident explicitly associated this frustration with a loss of learning opportunities:

[On some critical care services,] residents are not even involved at all in the consults. They are forced to babysit the old patients and consequently miss out on a valuable opportunity to assess sick versus not sick.

Environmental factors related to resident autonomy restrictions.

Some faculty attributed their restriction of resident autonomy to environmental barriers. Both faculty and residents noted that a high patient census, fragmented resident time, frequent changes in attending faculty and resident assignments, and the pressure of completing clerical work were impediments to resident autonomy and self-directed learning (6b, 6c). Competing demands on time were identified in more than three-quarters of the resident comments on this topic (6a).

Faculty strategies to support resident autonomy.

Some faculty described dealing with resident passivity by setting clear expectations for performance (4c, 4d). One faculty member commented that this method was effective, but also exhausting:

After two weeks of this, there is only so much teeth pulling I feel like doing to have them make a decision/plan.

However, other faculty admitted that they may be too quick to take control of the planning process:

I have definitely had a resident stop me, appropriately, and say, “Do you want to first hear what I want to do?” This was a bit of a wake-up call to me to allow her to provide a plan on her own. [Cf resident comment 2a]

I encourage housestaff to think independently—I tell them that I want to hear their thoughts, and I try hard to stay out of the discussion until they have communicated their thoughts.

Relatedness in the learning community.

The few comments that focused explicitly on trust and support were a mix of positive and negative opinions (7a–7d). Faculty and residents often expressed dismay and frustration about differences in “expectations,” “priorities,” or “cultures” between their respective groups. However, some faculty comments implied supportive attitudes toward residents and concern about meeting residents’ learning needs (3d), and a few residents expressed appreciation for faculty support, implying that their autonomy needs were being met (7b). In our educational environment, relatedness appears to be strained for many, but not all, residents and faculty.

Discussion

This study demonstrates significant discordance between faculty and resident perceptions of resident autonomy and of faculty support for resident autonomy within one residency program. Faculty and resident responses to parallel survey items showed strikingly consistent differences. Respondents’ written comments may help explain the underlying factors driving this dissonance and suggest potential strategies for resolution. Our analysis of these comments suggests that support of autonomy, competence, and relatedness—the three constructs of SDT—are closely interrelated in residency training. Strategies that help residents exercise autonomy appropriately are likely to encourage them to develop competence and enhance their relatedness to their team members and supervisors. Hence, our study affirms the relevance and potential importance of SDT in resident education, as has been demonstrated previously in other settings.^1–4,15–17

The consistent differences between faculty and resident perceptions of autonomy (Table 2) suggest that a common underlying factor or set of related factors may be driving the observed differences. Here, we will explore three possible underlying drivers: (1) generational differences, (2) inaccurate self-assessment on the part of both faculty and residents, and (3) challenges to self-determination. In the context of challenges to self-determination, we will consider strategies that may help to bridge the divide between residents and faculty.

Generational differences

Our results are generally consistent with findings of studies of generational differences in medical education.^{11,12,23–26} Descriptions of Generation Y (individuals born after 1982¹²) predict some of the attitudes expressed by our residents: For example, our residents indicated that they would like faculty to provide more specific work expectations, more support, better explanations when residents’ treatment plans are changed, and frequent feedback. However, we did not design this study for generational analysis: Although most of the responding residents belong to Generation Y, the faculty are a composite of generations.²⁵

Physician self-assessment

Our results are also consistent with reports that physicians are often inaccurate in their self-assessments.^27–30 The resident and faculty groups consistently rated themselves higher than the other group did, and some written comments suggested a lack of insight into how their own behaviors could help drive the behaviors of the other group. For example, overdirection by faculty may make residents more passive, whereas resident passivity may stimulate faculty to exert more control. We anticipate that efforts to improve the self-understanding and mutual empathy of both groups, through ongoing discussions of SDT, may enhance their insights into themselves and one another.

Challenges to self-determination

SDT provides a useful lens for interpreting resident and faculty responses to our survey. In written comments, several faculty said they wanted to see evidence of motivation and competence in a resident before trusting him or her to participate autonomously in patient care. Residents, in turn, expressed frustration about faculty withholding trust and limiting opportunities to exercise autonomy and demonstrate competence. Lack of mutual trust is a serious threat to relatedness in a learning community and in patient care teams.

The written comments also suggested strategies emerging from SDT for bridging the divide between residents and faculty. Some residents reported that they had learned to earn decision-making opportunities by giving faculty evidence of their motivation, competence, and confidence. Certain faculty commented that some “passive” residents had more potential for self-direction than was initially evident; these faculty described assiduous efforts to engage residents in active planning by clarifying and reinforcing their expectations. This autonomy-supportive teaching strategy, often called “scaffolding,” has been widely recommended to foster self-directed learning in learners at all levels.^31,32

We encourage faculty to scaffold the learning of residents they view as passive—including novices who are appropriately reluctant to take on independent roles—by giving them decision-making responsibilities in increasingly complex situations after they have proven themselves in less challenging settings. Dijksterhuis et al³³ have analyzed how faculty assess the degree of independence they feel comfortable giving to learners, and these authors argue that faculty need to find ways to move learners beyond their current level of competence without jeopardizing their confidence, relatedness, and capacity for autonomous action. We suggest that faculty who develop empathy with residents and learn to identify their concerns and needs are more likely to achieve this delicate balance.

The dissonance evident in our program could be symptomatic of the stresses other residency programs may experience as faculty face the challenge of evaluating resident achievement of milestones and entrustable professional activities.^34–36 Success of the Accreditation Council for Graduate Medical Education’s Next Accreditation System^37–39 requires that faculty learn to effectively support and at the same time evaluate residents’ progress toward independence.^33,40–42 The Carnegie Foundation’s 2010 call for reform recommends that we “promote relationships with faculty who simultaneously support learners and hold them to high standards” in order to facilitate identity formation in physicians-in-training.¹⁴

Faculty and residents labor together in a highly regulated environment and could potentially create partnerships to help one another cope. According to SDT, arbitrary external controls that conflict with personal value systems would be expected to challenge the autonomy of both groups.^2,3 Developing new approaches to help both faculty and residents live more autonomous lives, while sharing coping strategies and bonding around common frustrations, may make our learning communities healthier.¹⁴

Strengths and limitations

This study has limitations. First, generalizability may be restricted. The study was conducted within one residency program where frequent discussion of autonomy issues may have elicited differences between residents and faculty that would not be evident elsewhere. Moreover, most respondents represented the discipline of pediatrics. However, given that generational divides have been reported in medical school and in residencies in other disciplines,^23–26 we believe that the study of resident–faculty discord in other settings may also be clarified by the insights offered by SDT.

Another important limitation of this study is that our interpretation of survey responses was not formally evaluated for construct validity (as defined by Cook and Beckman²⁰). The brevity of our tool, although necessary to ensure a good response rate, may have increased the risk of construct underrepresentation.⁴³ However, the items we included were developed with careful attention to content and response-process validity, and the tool was pilot tested iteratively to enhance its reliability. All items related to one or more SDT constructs, thus minimizing the threat of construct-irrelevant variance.⁴³ In addition, measures to minimize response bias from social desirability were implemented. The consistent differences in the responses from resident and faculty groups suggest considerable internal consistency, although factor analysis was not feasible. Evaluation of validity by comparison of our findings with those from a “gold standard” assessment was impossible, and evidence from consequences was not relevant.

The consistency between our quantitative and qualitative data helps to cross-validate our findings, but neither source can be considered objective because all responses depended on self-report. The purpose of this study was to compare perceptions of the respondent groups about themselves and each other in relation to resident autonomy, and in studies of attitudes and emotions, self-report data—despite their limitations—are more informative than objective data.⁴⁴ Objective measurement of self-determination factors in our residents is ongoing.

Conclusions

This study identified consistent differences between faculty and resident perceptions of resident autonomy and of faculty support for resident autonomy. Analysis of written comments showed that autonomy, competence, and relatedness interacted in the way faculty directed the clinical activities of residents, thus confirming the relevance of SDT to resident education in our residency program. Faculty expressed reluctance to support residents’ needs for autonomy in patient care if residents failed to demonstrate motivation and competence.

A major concern raised by this study is that when faculty restrict the independence of “passive” residents whose competence they question, these residents may receive fewer opportunities for active learning. Some faculty, however, described success in motivating and activating passive residents by enforcing clear expectations for active participation in patient care decision making. Residents who are reluctant to act autonomously may benefit from more scaffolding in their education, so that they can gradually build the confidence they need to assume a more independent role in patient care.

We propose the following hypothesis for future study: Residents with weak autonomy and limited competence will benefit if faculty give them gradations of independence to scaffold their learning and development. The Next Accreditation System³⁹ will offer concrete tools to facilitate this process. Future research could also address interactions between resident and faculty autonomy in the educational environment—exploring, for example, whether faculty who feel their own autonomy is restricted are more likely to limit independent learning opportunities for residents. Such a study might inform new strategies for enhancing resident learning in an environment bristling with external controls.

Acknowledgments: The authors wish to thank Aaron Blumkin, MS, for assistance with data analysis, and MacKenzi Hillard, MD, for her contributions to the authors’ Self-Determined Learning and Improvement Project. Geoffrey Williams provided early guidance in project planning.