Sheldon, K M; Krieger, L S --- "Understanding the Negative Effects of Legal Education on Law Students: A Longitudinal Test of Self-determination Theory" [2007] LegEdDig 11; (2007) 15(1) Legal Education Digest 29

Understanding the Negative Effects of Legal Education on Law Students: A Longitudinal Test of Self-Determination Theory

K M Sheldon & L S Krieger

[2007] LegEdDig 11; (2007) 15(1) Legal Education Digest 29

FSC Col Public Law Research Paper No. 206, 2006, pp 1–49

The popular notion that law school is an exceptionally stressful experience for many students has been substantiated by longitudinal studies. Indeed, the emotional distress of law students appears to significantly exceed that of medical students and at times to approach that of psychiatric populations. These findings have substantial human and social significance, given that the level of adjustment of graduating law students is likely to carry over into professional practice and may set the stage for the unparalleled frequency of psychological distress and other problems seen broadly among lawyers today.

In recent longitudinal studies, Sheldon and Krieger confirmed earlier findings of emotional distress among law students, and deepened the inquiry to address the motivation and values of the students. They applied self-determination theory (SDT), a humanistically-oriented but also rigorously empirical theory of human motivation, which has been under development for three decades. Consistent with earlier longitudinal studies, Sheldon and Krieger demonstrated that law students experience precipitous declines in their mental health during their first year. More importantly, they showed that these declines were correlated with negative changes in motivation and valuing as defined by SDT. Specifically, students evidenced radically reduced intrinsic motivation, i.e., a loss of engagement in behaviour because of its interest and enjoyment potential. In addition, students showed maladaptive changes in valuing during their first year of law school: they shifted towards (extrinsic) image and appearance values and away from (intrinsic) community and helping values, and also exhibited a general decline in the overall level of valuing (both intrinsic and extrinsic). These first-year effects were essentially replicated across two different samples at two different law schools, and were shown to persist into the third year within one of the samples, from which second- and third-year data were also collected.

However, they did not take full advantage of the conceptual resources of SDT because they did not take into account two other important components of SDT — psychological need satisfaction, and the nature of the social context. In addition, the earlier studies did not consider SDT’s dynamic process model based on these constructs.

The integrated model begins with the social context. According to SDT, the development of positive motivation is importantly forwarded or impeded by the characteristics of the social environment. Specifically, when authorities provide ‘autonomy support’ and acknowledge their subordinates’ initiative and self-directedness, those subordinates discover, retain and enhance their intrinsic motivations and at least internalise non-enjoyable but important extrinsic motivations. In contrast, when authorities are controlling or deny the self-agency of subordinates, intrinsic motivations are undermined and internalisation is forestalled.

Autonomy support has three prototypical features: (1) choice provision, in which the authority provides subordinates with as much choice as possible within the constraints of the task and situation; (2) meaningful rationale provision, in which the authority explains the situation in cases where no choice can be provided; and (3) perspective taking, in which the authority shows that he/she is aware of, and cares about, the point of view of the subordinate.

The next step in the integrated model links autonomy support to psychological need satisfaction. Psychological need concepts have become increasingly important in SDT, because they help explain the positive vs. negative effects of social context while also grounding the theory within an overarching view of evolved human nature. According to SDT, all human beings require regular experiences of autonomy, competence, and relatedness in order to thrive and maximise their positive motivation. In other words, people need to feel that they are good at what they do, or at least can become good at it (competence); that they are doing what they choose and want to be doing — e.g., what they enjoy or at least believe in (autonomy); and that they are relating meaningfully to others in the process — e.g., connecting with the selves of other people (relatedness).

The final link in the integrated model leads from psychological need satisfaction to three positive outcomes: well-being/psychological health, grade performance, and self-determined career motivation. The prediction of the first outcome, psychological well-being and health, derives from a prominent strand of SDT research that has focused on mood, life-satisfaction, and psychological vitality.

The second positive outcome in the integrated model is law school grade-point average (GPA). This reflects another strand within SDT, in which the determinants of optimal performance, e.g., outcomes such as mastery, creativity, conceptual learning, and level of achievement, are investigated. Because SDT postulates that psychological need satisfaction is important for positive outcomes of all types, the integrated SDT model contains a path from changes in need satisfaction to final law school GPA (although there are doubtless many other factors besides psychological need-satisfaction which influence GPA).

The third positive outcome is self-determined job motivation. According to SDT, when social contexts support autonomy, and therefore promote psychological need satisfaction, individuals gain the inner resources to develop and follow intrinsic motivations, and are also able to identify with and internalise appropriate extrinsic motivations within those contexts.

As previously stated, in this research we placed self-determined motivation at the end of the model, as a positive outcome of need satisfaction rather than as a cause of satisfaction and well-being outcomes itself. We did this because commentaries indicate that by the third year of law school most students have begun disengaging from law school itself, instead focusing on their future work.

We also hoped to extend the SDT model in a very different way, by considering variations across different law schools. Again, SDT views the social context as an important initial determinant of a downstream chain of causes and effects. Obviously, however, such effects might reflect individual dynamics as much as social dynamics, and thus it is important to manipulate or at least measure a social-level grouping factor that might correspond to objective differences in contextual autonomy support. In these studies we addressed this problem by collecting data from students at two different law schools, with institutional differences that might well result in differing levels of perceived institutional autonomy support.

The two law schools we selected (LS1 and LS2) were located in different regions of the country. Although both schools admit highly qualified candidates, with largely equivalent undergraduate grades and LSAT scores, the schools appear to have somewhat different educational and pedagogical philosophies.

In sum, our overall goal was to test the entire SDT model of thriving, including the basic model and also a further initial predictor reflecting group-level variations in institutional environments.

There were some initial demographic differences between the two samples, which were expected given the different settings and program offerings at the two schools.

At the first assessment, participants were asked to self-report their undergraduate GPA using the typical 4-point scale (e.g., 2.9, 3.6). In the final assessment, participants were asked ‘what is your cumulative GPA so far in law school?’ They checked one out of eight options, including ‘F’, ‘D’, ‘C’, ‘C+’, ‘B’, ‘B+’, ‘A-’, and ‘A’. We converted these reports to the typical 4-point scale.

Subjective well-being was assessed at the beginning and end of law school via the positive affect scale of the positive affect/negative affect scale, the Satisfaction with Life scale, and the 6-item depression scale of the Brief Symptom Inventory. The positive affect scale contains items such as ‘inspired,’ ‘active,’ and ‘proud’; the life-satisfaction scale contains items such as ‘I am satisfied with my life’ and ‘the conditions of my life are excellent;’ and the depression scale contains items such as ‘feeling lonely’ and ‘feeling no interest in things.’ All items were administered with 1 (not at all) to 5 (very much) scales, with reference to participants’ experience of ‘the last two months.’ We chose the two-month time frame because we wanted to assess participants’ general state of mind leading up to the assessment.

Because principal components analyses revealed that these three measures formed a single factor at each time of administration (after depression was recoded), we computed Year 1 and Year 3 subjective well-being composites by subtracting depression from the sum of positive affect and life-satisfaction.

To assess psychological need satisfaction we used the nine items employed by Sheldon et al. to assess autonomy, competence, and relatedness need satisfaction in their study of ‘most satisfying events’ (3 items per need).

As preliminary analyses, we first present descriptive statistics and zero-order correlations for the whole sample, and also present simple mean differences between the two law schools. We also consider the effects of sample attrition. Then, we begin our hypothesis testing by evaluating mean differences in autonomy support between the two law schools, controlling for several demographic variables. This is important to verify that the two schools indeed differ on this crucial social-contextual variable. In addition, we examine institution effects on changes in need-satisfaction and on SWB. Next, we conduct three three-step regressions, one for each dependent measure, in which law school membership and several demographic variables are entered at the first step, autonomy-support is entered at the second step, and the three change-in-need satisfaction variables are entered at the third step. Finally, we test the entire model simultaneously using structural equation modelling (SEM) techniques, including a multigroup analysis to evaluate the model’s invariance across the two schools.

This three-year prospective study provides new support for SDT’s dynamic-process model of human thriving, and also provides considerable new insight into the problems in contemporary legal education. To date, few longitudinal studies of this duration have been reported, either within past SDT research or within past studies of the effects of legal education.

The first step of the validated model addressed group-level variations in institutional culture. Specifically, faculty at LS2 were felt to be more focused on student concerns, and to provide more choices and more meaningful rationales for mandatory rules and requirements. In comparison, faculty at LS1 were felt by students to be more controlling and insensitive to their perspectives.

At the second step of the model, which concerns the combined sample regardless of institutional membership, perceived autonomy support predicted greater psychological need satisfaction over the three years. Conversely, students who rated faculty within their program as more controlling also experienced declining psychological need satisfaction. This is consistent with SDT’s emphasis on the fact that negative social contexts can deprive people of the psychological nutriments that they need to thrive.

Another innovation for the SDT model derives from our longitudinal examination of two different institutional contexts as predictors of perceived autonomy support. Although SDT focuses conceptually upon social dynamics, social-level variables have typically been assessed via individual differences in perceptions of a single social context. With such a methodology, autonomy support, a social-level factor, may sometimes be confounded with personality differences that are unrelated to the social context. In our data, perceived autonomy support differed as expected across two institutional contexts, and these group-level differences in turn had downstream effects upon those immersed within those contexts.

These results suggest that, in order to maximise the learning and emotional adjustment of its graduates, law schools need to focus on enhancing their students’ feelings of autonomy. Because such feelings can have trickle-down effects, predicting changes in students’ basic need satisfaction and consequent psychological well-being, effects which may also carry forward into the legal career. Given the excessive incidence of depression among practicing lawyers and the likely negative spillover of that phenomenon to society and the justice system, these findings may have substantial practical importance.

Autonomy support also predicted (via need satisfaction) what many would consider the most important outcome of all — final law school GPA. Indeed, despite the equivalence of academic predictors, the school populations differed markedly in their multistate bar exam results, in ways that would be predicted from our findings of institutional differences in reported autonomy support and need satisfaction.

The current findings, that autonomy support predicted intrinsic motivation for work, suggest that the motivation-dampening effects of law school may indeed have negative effects that extend well beyond graduation. These effects are also consistent with recent commentators’ claims that there are major problems in the legal profession. Of course, much longer-term longitudinal studies are required to confirm these potential connections.

Given our findings that greater autonomy support predicted improved need satisfaction, subjective well-being, and motivation for career, and may have led to enhanced learning and bar performance as well, schools would be well-served to evaluate how they can best consider the priorities of their students and provide choices consistent with those priorities.

Changes toward employing faculty with more teaching and lawyering (including public service) experience, offering a balance of practical skills training, or providing more training and rewards for teaching excellence, might also ultimately enhance students’ sense of autonomy and engagement.

This study found certain negative effects that were common to the two diverse schools sampled, and also found important differences in those effects. To further explore the differing effects, future research should expand the number of samples and seek to pinpoint the institutional factors most linked to student perceptions of autonomy support. In particular, studies might focus on schools with systematically varying commitments to scholarship and teaching, varying faculty qualifications and orientations, varying grading systems, and varying program support and flexibility. Finally, future research could also follow law students into their careers to determine the longer-term consequences of their educational experiences.