Case-based Approaches to Professional Ethics: a systematic comparison of students' and ethicists' moral reasoning

This article provides a systematic analysis of the cognitive processes required for acquiring skill in practical ethical reasoning in a professional domain. We undertook this NSF-supported research project in part to study relationships between case-based instruction in professional ethics and cognitive analyses of ethical reasoning strategies. Using a web-based experimental design, we report striking differences in the students’ and ethicists’ use of knowledge and reasoning. Virtually all of the ethicists and some students’ protocols made signiŽ cant use of specialized professional knowledge and also used role-speciŽ c content in the ethical principles applied in their responses. In contrast, other student protocols made signiŽ cantly more use of common knowledge and applied more general ethical principles or appealed to consequences in the justiŽ cation of their responses. Our analyses show how certain strategies were superior to others in regard to identifying alternative moral issues, assessing the moral implications of actions, and providing alternative practical resolutions to con icts. The Ž ndings point to the importance of professional knowledge and role-speciŽ c professional obligations in resolving professional ethical con icts: the same knowledge and “middle-level” principles necessary to comprehend and apply professional codes of ethics. In considering professional ethical reasoning from the viewpoint of cognitive science, an area of primary interest concerns students’ abilities to recognise and respond appropriately to ethical problems characteristic of their professional practice. Increasingly, professional schools in higher education are including ethics courses in their programs and increasingly the pedagogical methods of these courses have relied on the use of case-based instructional strategies. The pedagogical value of focusing ethics instruction on individual cases, on case-based comparison, and on explicit procedures for conducting moral deliberation has been recognised by Jonsen and ISSN 0305-7240 print; ISSN 1465-3877 online/01/040377-22 Ó 2001 Journal of Moral Education Ltd DOI: 10.1080/03057240120094869 378 M. Keefer & K. D. Ashley Toulmin (1990), Jonsen (1991), GrifŽ n (1988) Arras (1991), Beauchamp and Childress (2000), Harris et al. (1999) and Whitbeck (1998). One way that educators have attempted to alert young professionals to the responsibilities that attach to professional knowledge has been to stress the importance of codes of ethics. Ethics codes are not only a fact of life in professional practice. Codes of ethics provide moral principles aimed at guiding practice; their provisions apply the principles of common morality to a speciŽ c professional Ž eld by specifying additional, more domain-speciŽ c principles. These role-speciŽ c principles extend common morality by articulating the special obligations pertaining to professional roles. Students need to have a working knowledge of the provisions of their code in order to align their ethical conduct to professional standards. Yet, while we know that knowledge of professional codes is important, many authors note that most professionals do not frequently consult ethics code provisions as a guide in making decisions (see Harris et al., 1999). This would suggest that much of the practical and ethical knowledge that is characteristic of a domain is acquired “on the job” (Resnick & Wirt 1996; Lave & Wenger, 1991). From a cognitive viewpoint, applying ethical codes is difŽ cult partly due to the open-ended quality of professional codes. For example, in the domain of engineering, numerous code provisions such as those involving “con icts of interest” or admonitions to “recognize the proprietary interests of others”, “hold paramount the safety, health, welfare, of the public in the performance of their professional duties” or “act in professional matters for each employer or client as faithful agents or trustees” involve deliberately open-textured terms which are not further deŽ ned in the codes (Harris et al., 1995, pp. 6–7). Hence, simple knowledge of code provisions is clearly not enough to help students learn to recognise ethical dilemmas they are likely to encounter in their professional practice. Knowledge of these role-speciŽ c obligations along with relevant specialised knowledge is necessary if professionals are to identify ethical problems that they are likely to encounter in their professional life. We believe that students will only understand and be guided by ethics code provisions if they gain experience working through case examples and explanations that identify the problem features which code principles mark as ethically relevant. SpeciŽ cally, we are interested in the way common morality is applied to professional contexts and understanding the role that specialised professional knowledge plays in the application of these domain-speciŽ c middle-level principles, e.g. those principles that might appear in professional ethics code provisions. Our Model of Case-based Moral Reasoning The model we have devised to analyse case-based moral reasoning is informed by our own previous work and uses techniques of discourse analysis, AI research in cognitive science, analogical reasoning and problem-solving (Ashley, 1990; Keefer & Olsen, 1995; Ashley & Keefer, 1996; Keefer et al., 1998). It is also closely aligned with new approaches to the teaching of practical ethics (e.g. GrifŽ n, 1988; Arras, Comparing Students’ and Ethicists’ Reasoning 379 1991; Jonsen, 1991; Whitbeck, 1998), especially those found in a text such as Engineering Ethics: Concepts and Cases (Harris et al., 1999). Our model and these recent philosophical approaches share a commitment to studying and teaching ethics using analyses of moral decision-making in practical contexts, usually in the form of realistic case examples. For example, professional ethics texts such as Engineering Ethics (Harris et al., 1995) provide many ethics problems and analyses invoking and interpreting general provisions of common morality and speciŽ c ethics code provisions. The pedagogical efŽ cacy of such an approach relies on the willingness of students to work through the numerous case examples and to compare cases in light of the ethicists’ detailed case analyses. The case examples in Engineering Ethics are presented to students in an interesting way. Often, the authors organise a series of cases around a principle drawn from common morality such as “it is wrong to steal or to commit theft” or from an engineering ethics code provision such as “Engineers shall not disclose conŽ dential information concerning the business affairs or technical processes of any present or former client or employer without his consent ([NPSE code] III.4)” (Harris et al., 1995, p. 126). Our model has been devised to help us classify in a systematic way the analyses of such ethical problems as performed by engineering students or by ethicists. The model supports classiŽ cation of four aspects of a case-based analysis: the components of the analysis, its complexity, the content of the moral justiŽ cation in the analysis and the knowledge invoked in the analysis. Components in Case-based Analysis According to our model, one may observe seven basic components in case-based moral reasoning, as summarised in Table I. A reasoner may (1) identify the moral issue at stake as well as (2) the relevant known and unknown facts in the problem, (3) offer a resolution and (4) a moral justiŽ cation, (5) consider alternative scenarios that argue for different conclusions, (6) identify and evaluate longer term moral consequences and (7) offer alternative resolutions. For example, according to our model, when we observe a subject’s analysis identifying moral issues, we encode it as Component 1. If the analysis also invokes some ethical principles as justiŽ cations, we encode it as Component 4 as well. Components 5 and 7 of our model account for some speciŽ c aspects of the process of moral analysis described in (Harris et al., 1995, Figs 4.1, p. 102; 4.3, p. 111; 5.1, p. 128; 5.2, p. 140). In our study, some subjects’ case analyses proceeded in a decidedly casuist way. They presented a series of examples along a spectrum from clear positive to clear negative (e.g. the “line-drawing” method in Harris et al., 1995, p. 131). Each successive case represents an addition of or a change in the magnitude of some morally relevant factor or condition which makes the application of common or specialised middle-level moral principles more or less necessary (Harris, et al., 1995, p. 131). In our model we code this tactic as reasoning Component 5, generating alternative scenarios. 380 M. Keefer & K. D. Ashley TABLE I. Component model of moral case-based reasoning Component 1 Identify the moral issue at stake Component 2 Identify the relevant knowledge and unknown facts in a problem Component 3 Offer a resolution Component 4 Offer a moral justiŽ cation Component 5 Consider alternative scenarios that argue for different conclusions Component 6 Identify and evaluate longer term moral consequences Component 7 Offer alternative resolutions using interactive planning Other subjects organised case analyses around con icting middle-level code principles, such as engineers should “hold paramount the safety, health, welfare, of the public in the performance of their professional duties” versus “act in professional matters for each employer or client as faithful agents or trustees” (Harris et al., 1995, p. 139). Subjects also annotated cases presented as calling for “hard choices” between con icting principles with a range of optional actions the protagonist could take (referred to as creative middle-way solutions). The actions are organised by the extent to which they serve each of the con icting principles (i.e. in a manner that satisŽ es all, or most, of the moral claims in the dilemma). In our model we classify this kind of reasoning as Component 7, alternative actions. In previous work, the Ž rst author established a strong empirical correlation between protocols using these alternative action resolutions (involving interactive plans and social episodes) and protocols scored as exhibiting Carol Gilligan’s Care Orientation to moral reasoning (Keefer & Olson, 1995). Complexity of Case-based Analysis Based on our empirical investigations, described below, students often provide a basic case analysis comprising either a simple plan for resolving the problem (Component 3) or a small set of alternative resolutions (Component 7) as well as a moral justiŽ cation, for instance, by invoking some ethical principles (Component 4). Other responses are more complex, and our model includes a scheme for classiŽ cation of complexity of analyses as shown in Table II. The basic response of Component 3 and 4, or 7 and 4, are classiŽ ed as Component Level A. If a subject includes not only a resolution but also a combination of two components the response is classiŽ ed as Level B. These may include a justiŽ cation (Component 4), identiŽ cation of a moral issue (Component 1), identiŽ cation of known and unknown relevant facts (Component 2), identiŽ cation of moral consequences (Component 6), or alternative scenarios where conclusions differ (Component 5). Level C involves a combination of three such elements, and four elements merit classiŽ cation as Level D. Comparing Students’ and Ethicists’ Reasoning 381 TABLE II. DeŽ nition of component complexity levels