Moral Dilemma Discussion Revisited – The Konstanz Method

The classical Blatt-method of dilemma discussion to foster moral-democratic competencies is discussed and compared with the new Konstanz Method of Dilemma Discussion (KMDD). The KMDD is better teachable, easier to be applied and has higher effect sizes. In a democracy, the rule of a king or a dictator has been replaced by the rule of moral principles of human conduct and interaction. Therefore, it is highly important that citizens do not only hold high moral and democratic ideals (which, according to the polls, most citizens do) but that they also possess a highly developed ability to apply these ideals in everyday-life, and solve inevitable conflicts through discussion and moral discourse. Especially when we are confronted with a conflict situation, we need the competence to make judgments in accordance with our moral principles and act upon them as well as to enter a moral discourse with our opponents rather than use violence and power to solve the conflict (Kohlberg, 1964; 1984). Indeed, moral and democratic competencies seem to be very important for various fields of behavior: low moral judgment and discourse competencies may be a cause of criminal behavior, may keep people back from acting upon their moral ideals of helping and compassion, and may get into the way of learning (Kohlberg, 1984; Sprinthall et al., 1994). In experimental studies it was found that people who have learned how to solve conflicts between opposing values and principles can learn better, allow others to help them learn and apply what they have learned in every-day contexts (Heidbrink, 1985; Lind, 2003. These moral abilities or competencies are not inborn nor can they be instilled through simple lectures but must (and can) be educated (and self-educated) throughout the life-cycle, that is, from early age through childhood, youth and adulthood (cf. Lind, 2002). Moral and democratic competencies are acquired in various places like family, school, university, workplace and in the public domain through neighborhood activities, political participation and engagement in universal community building. In the early years, children's learning depends heavily, yet not solely on the assistance of caring and competent educators like parents and teachers. Children would not learn if they were not, from the beginning, active learners themselves. Later in life children learn more and more through self-monitored activities, which must be encouraged and fostered rather than regarded as disturbing. The best known way to foster moral and democratic competencies is to provide proper learning opportunities in which s/he feels safe to freely express his/her moral ideals and arguments and in which s/he also respects others and their right of opinion. Such a learning opportunity is provided by the teaching method of the moral dilemma discussion, first suggested by Blatt (1969) and his mentor Kohlberg (Blatt & Kohlberg, 1975), and later improved by Lind (2003). The method of moral dilemma discussion is well-founded in the philosophy of education and in psychological and educational research, and its effects have been thoroughly analyzed in well-designed studies (Higgins, 1980; Leming, 1985; Lockwood, 1978; Lind, 2002). From the beginning, the dilemma discussion has shown to have a substantial effect size ( r = 0.40); hardly ever were negative effects reported. It has shown to be highly effective in various age-groups, from ten-year-olds to adults. Teachers of all subjects can be trained to use this method effectively in their classrooms on various age-levels, starting from grade five upward. Critical reviews of three decades of intervention studies using the Blatt method (e.g., Berkowitz, 1981; Berkowitz et al., 1980; Lind, 2002; Walker, 1983) tell us that dilemma discussion and other methods of moral and democratic teaching work most effectively if two basic conditions are met: * First, mutual respect and free moral discourse in the classroom. The teachers must see himor herself as a facilitator not as a master of students' learning. Especially in regard to the moral and democratic domain of learning, the teacher must not use his or her authority to impose his or her aims and pace of learning onto the students but tune into their moral ideals and their pace of development. This tuning-in is supported by a good understanding of the basic processes of teaching and learning. Yet, ultimately it must be achieved by engaging in a free discourse with the students in the sense of Habermas' (1983) discourse principles. In a dilemma discussion session, everybody is respected equally and has one vote and one voice regardless of power and status. * Second, high level of persisting attention. Learning takes place most effectively in situations in which the level of attention of all participating students is kept up as high as possible throughout the session. This is achieved best in sessions in which cycles of support and challenge alternate in a certain rhythm. Through challenges students get emotional and attentive, eager to solve a problem or to ease bad feelings. Yet, challenges must never last for too long or get too strong so that emotions don't raise too high and prevent learning. Over-excitement can be avoided by switching from phases of challenge to phases of support, in which the individual is reassured and his or her emotions can calm down to allow again intellectual activities and reflection to recuperate. Yet, the Blatt-method left room for improvements. They are much more teacher-directed than one would expect and those who run dilemma discussions frequently report problems with keeping up students' attention. Although the Blatt-method of dilemma discussion has generally a very strong impact on students' moral development, this method has some severe drawbacks: * The core rationale of the method, the so-called "+1-convention" has attracted much criticism (cf. Berkowitz, 1981; Berkowitz et al., 1980). With this the teacher is required to confront students always with arguments one stage above their own. This requirement is not only difficult to meet but also seems to be more in line with classical socialization theory than with the cognitive-development theory by Kohlberg and Piaget, which recommends that the teachers only stimulates development but does not try to model it. In a series of very thoughtful laboratory experiments, Walker (1983) found that counter-arguments are just as effective for enhancing students' moral judgment scores as "+1"arguments, implying that +1-arguments do not work because they "model" the reasoning of children but by challenging it like counter-arguments do. * Presenting several dilemmas in only one 45-minutes session seems to be effective, yet this practice leaves little room for the students to understand the nature of the dilemma and to develop a high level of reasoning. Therefore it is not surprising, that the effects sizes crumble some time after the end of the dilemma discussion program. * To discuss one or even several moral dilemmas within 45-minutes leaves too little room for students' active reflection and discussion activity. If the students (rather than the teacher) are to produce some number of good and challenging moral arguments, they need more time (I recommend 80 to 90 minutes) to fully understand the moral core of the dilemma and to think of all arguments they are capable of. * Too many sessions within a short time span don't make such programs more effective. On the contrary, fewer session spread over a longer period seem to have the same immediate effect and besides a more lasting effect. As the meta-analysis by Lind (2002) and also other practical experiences suggest, a good time span for a dilemma discussion is one session every two or three weeks. * The role of the teachers should be that of a facilitator rather than a master or moralizer. If the teacher gives reasons "one stage above" the students' stage of reasoning, the hidden message is that the teachers knows it better and the students must follow the lead of the teacher not only in matters of fact but also in moral matters. This hidden curriculum is certainly not in agreement with moral-democratic education as we understand it. * Programs to foster moral and democratic competencies must be evaluated as adequately as possible. This means that instruments for measurement are needed which tap the competence aspect of moral behavior rather than attitudinal aspects. Kohlberg's moral judgment interview does this to some degree. Yet it does not clearly define what the moral task is and its indexes (stage score, MMS) confound the competence and the attitudinal aspects. In the Konstanz Method of Dilemma Discussion (KMDD) two major innovations were introduced aiming at a more effective use of dilemmas for moral education and a more valid evaluation methodology (Lind, 2003): First, to create an optimal environment for moral and democratic learning, KMDD employs semi-real, 'educative' dilemmas as tasks. Semi-real dilemmas are dilemmas which do not affect anyone participating in a dilemma discussion, but are on the one hand likely to cause real conflicts between the moral ideals of a person, and on the other hand also cause controversies between the participants, thus triggering real moral emotions. 'Educative' is a dilemma, if it triggers moral emotions enough to stimulate learning, but not too strongly to prevent learning. The role of those dilemmas then is similar to a vaccination, in which weakened viruses are injected into the body to stimulate its ability to resist the real virus. Educative moral dilemmas are not so real that they kill of moral-democratic learning, but real enough that they cause the individual to develop his or her cognitive-affective capacities (Lind, 2003; http://www.uni-konstanz.de/ag-moral/ ). Furthermore, KMDD has been designed to facilitate free moral deliberation and discourse through several features: a)sufficient time is allowed

Blatt and Kohlberg's method of moral dilemma discussion (MDD) has been introduced more than 30 years ago and since then it has been tried out in many intervention studies, and become part of teacher education programs in many countries.It rests on three assumptions: a) Moral and democratic behavior is largely dependent on a person ability to apply his or her moral ideals in a conflict situation, that is, on his or her competence to make moral judgments and to enter a moral discourse with opponents.
b) These moral competencies can (and need to) be fostered through out the life-span, that is, from early age through childhood, youth and adulthood, in family, schools, university and at the workplace.
c) Moral competence and other desired outcomes are best fostered through providing an learning environment in which the individual is challenged by a moral task and in which s/he feels safe to freely express his/her moral ideals and arguments and in which s/he also respects others' right of their own opinion.
In contrast to many other methods of moral education, this method is well-founded in the philosophy of education and in psychological and educational research, and its effects have been thoroughly studied in well-designed studies (Enright et al., 1983;Higgins, 1980;Leming, 1985;Lockwood, 1978;Schläfli et al., 1985;Lind, 2002a).From the beginning, the method of MDD has shown to have a substantial effect size; hardly ever were negative effects reported.However, there is still little know about the critical conditions in this method to promote development, and there are still controversies regarding the question of how to make this method more effective, for which population it is best suited, and what design should be used to evaluate its effects empirically.To make the effects of this method more effective and to make this method better teachable and applicable in classroom teaching, we have experimented with various changes of the dilemma discussion and thoroughly studied their effects and the effects of other intervention experiments.
In this paper I will present the Konstanz method of dilemma discussion as it has emerged out of these studies.The Konstanz method is based on a critical review the original Blatt-Kohlberg method and empirical evaluation studies (e.g., Berkowitz, 1981;Berkowitz et al., 1980;Oser & Althof, 1994;Schläfli et al., 1985;Walker, 1983).To create an optimal learning environment, the Konstanz method employs semi-real dilemmas as tasks.Semi-real dilemmas are dilemmas which cause a real conflict within the individuals between their moral ideals, and also cause controversies between the participants, thus triggering real moral emotions in them, but do not really affect any participant in a dilemma discussion nor anyone else.Their role in the stimulation of moral judgment and discourse competence is seen to be similar to a vaccination, in which weakened viruses are injected into the body to stimulate its ability to resist the real virus.Furthermore, the Konstanz method has been designed to facilitate free moral deliberation and discourse through several features: a) as already mentioned, semi-real (rather than real) dilemmas are given; b) sufficient is allowed for the clarification of the dilemma; c) the teacher withholds his/her opinion of the dilemma and even gives up the discussion chair; and d) participants built up solidarity in small group meetings.Finally, to keep the attention and learning motivation at an optimal level, a dilemma discussion session is divided into phases of support and challenge, which rhythmically alternate about every ten minutes.Special features of the Konstanz method are: a) sufficient time at the beginning of the session to bring out and clarify the different perceptions of the presented problem by the participants, b) or dilemma a phase in which participants rank order the arguments of their opponents, and a concluding phase of session evaluation asking the students questions like What have you learned?Was it wasted time or would you like to do it again sometime?etc.

The Core Features of Blatt-Kohlberg Method and its Offsprings
Since Moshe Blatt (1969) used the method of dilemma discussion to stimulate morel development for the first time, his method was not just copied but changes in many ways.Yet some features seem to have been in common, which makes us call this the "Blatt method" (see Blatt & Kohlberg, 1975;Hersh et al., 1979;Gomberg et al., 1980;Kuhmerker et al., 1991).Blatt's intervention looked like this: -Twelve to eighteen dilemma sessions, one every week.
-Hypothetical dilemmas are used for starting of a discussion in class.
-Provision of a non-threatening classroom atmosphere.
-Opportunities for student-student interaction.
-Use of probe questions and of arguments at the students' stage of moral reasoning as well as one stage above theirs ("plus-1-rule").
-Additional methods like role-playing, filmstrips, audiotapes, video clips, short questionnaire, etc. are used to add to the vigor of the discussion and prevent boredom among students.
-Inclusion of dilemma discussion in the mainstream curriculum.
Aside from this, the Blatt-method of dilemma discussion was modified in many ways.At the beginning, the discussion was often lead by a researcher or doctoral student who was interested in trying out this method and studying its effects.From these studies we can gain many valuable insights today, which help us to better understand how this method works and what its critical conditions are.
Later also many teachers started to use this method and learned to integrate it into their curriculum design.In contrast to external discussion leaders they were able to follow-up on the classroom discussions.The original study by Moshe Blatt (Blatt & Kohlberg, 1975) showed that teachers were more effective than external discussion leaders with the younger students and with disadvantaged students.A meta-analysis of over 140 interventions (Lind, 2002a) showed also that teachers had a more sustained impact on the development of moral competencies in their students.

Methodological Notes: Ways of Estimating Impact through Effect Sizes
From the beginning, the effects of dilemma discussion have been very intensively studied, probably more intensively than any other method of moral and democracy education.In the early years, evaluation studies typically consisted of a pretest-posttest design with a control group, and employed Kohlberg's Moral Judgment Interview (MJI) for measurement.A few studies (like Blatt's) conducted also post-posttests to see if the gains could be sustained.Gains through the educational intervention were measured in terms of Stage movement or increase of Moral Maturity Score (MMS), which ranges from 100 to 500.For example, in Blatt's first intervention, 64 percent of the participants had developed one full stage in their moral reasoning as measured by the MJI.On eyear later, the experimental group maintained this lead over the others (Blatt & Kohlberg, 1975).
In agreement with conventional educational research, these measures were accompanied by the calculation of statistical significance, and most summary articles based their conclusions on "significance counts:" How many of the interventions showed "significant" gains?Because, statistical significance testing is highly controversial, Lind (2002aLind ( /1993) ) re-analyzed all available first generation studies (done from about 1969 to 1986) and calculated correlation coefficients r as a measure of effect size.The meta-analysis of the re-analyzed data showed surprisingly high effect sizes of the Blatt method of dilemma discussion: the mean effect sizes was r = 0.50, which is far above the mean effect size of r = 0.30 which we can expect from effective psychological and educations interventions (Lipsey & Wilson, 1993).Lind (2002a) also found that dilemma discussion was, as Blatt and Kohlberg (1975) already noted, most effective in early adolescents, between 10 and 16 years.
A second generation of intervention studies introduced two major changes (Schläfli et al., 1985): Instead of the MJI, these studies used a measure of preference for principled moral reasoning, the Defining Issues Test (DIT) by Jim Rest (1979), and instead of calculating discrete stage scores, they used a continuous scale, the P-index.They also calculated a measure of effect size, namely d, which is similar to r but can become larger than 1.0.The meta-analysis of these interventions brought about a different picture: The overall effect on preference for principled moral reasoning was much lower than could be expected.Only in the group of adult participants, the dilemma discussion showed substantial effect sizes (mean r = 0.28), which were, however, far below the ones in the MJI-studies.The use of the DIT was obviously not the best way to make progress in moral education.The method of dilemma discussion was designed to foster moral judgment competence but not necessarily impose on the subjects a preference for principled moral reasoning.The MJI was not perfect in tapping the competence aspect of morality (cf.Lind, 1989).But at least it did.The P-index is also not sensitive to changes at lower developmental levels, which may partly explain why adolescents did hardly show gains in their P-score during dilemma discussions.
In a third generation of dilemma discussion studies we introduced several innovations aiming at a more effective use of dilemmas for moral education and a more valid evaluation methodology.We redesigned the Blatt-method (see below) and devised a new measure of moral judgment competence, the Moral Judgment Test (MJT; see Lind, 2002b).This new method of measurement proved to have some true advantages over the others: -Aside from indices of moral preferences, the MJT also provides a pure measure of moral judgment competence, which makes it more valid for studying the effects of dilemma discussion on moral judgment competence than a preference test (like the DIT) or a mixed competence-preference test (like the MJI).
-The MJT is much shorter and economical than the MJI and the DIT and thus is better suited for multiple re-testing (to study lasting effects) and large-scale comparison group studies.
-The MJT is more sensitive to change in lower developmental phases than the DIT or another test, the Socio-moral Reflection Measure (SRM) by Gibbs (Lind, 2002).
-The MJT can be objectively scored and requires no subjective rating like the MJI.
However, like all ability tests, the MJT showed to be vulnerable to test-taking fatigue.In several studies, no increase in the MJT's C-score could bee seen.It seemed that a considerable proportion of participants had been annoyed by our request to fill out the MJT twice of more often.Some participants complaint aloud.Interestingly, this disturbing effect was mostly confined to the participants with high C-scores, and to re-tests within half a year or shorter (we did not observe this phenomenon in our longitudinal studies, in which the assessments were two years apart).We hypothesized that the subjects would not find it so annoying to fill out the test twice if we had talked to them more openly about this.In fact, when we used a new instruction for the MJT, in which we explained why we had to use parts of the questionnaire twice, the mean test scores went up.
---------to be added ------  (Blatt & Kohlberg, 1985, p. 143) c.The Plus-One-Rule The Plus-One-Rule or convention, i.e., the rule that the teacher confronts the student with moral arguments just one stage above the student's level of moral reasoning, has shown to be at odds with Kohlberg's own constructivist theory of learning (which says that children need to discover the rightness or wrongness of their beliefs rather than being told what to belief), that it is impractical (a teacher can hardly assess all of his/her students' levels of moral reasoning all the time nor can he/she address all levels which are in her/his class simultaneously), and that it has not shown to be effective or more effective than other approaches.
As to the teacher's role in the dilemma discussion Berkowitz observes that "teachers are there as models and facilitators rather than simply as producers of moral reasoning.For the latter role we may rightly turn to students' peers.Peer reasoning is often more seductive, moral convincing, and more stimulating.... It is the teacher's duty to promote an atmosphere of fairness, i.e., a context in which the democratic interplay of moral ideas promotes an atmosphere conductive to student development" (Berkowitz, 1981, p. 489) d.Counter-Arguments An alternative method of eliciting moral reasoning and thinking in students is to confront them with counter-arguments, not by the teacher but by the peers.This method seems to be more in line with the basic constructivist notions of Piaget and Kohlberg and seems to have at least as strong an effect on moral learning as has the Plus-One-Rule.In a carefully experimental study, Lawrence Walker found "that these moral discussions will be most effective when they incorporate high levels of conflict in both opinion and reasoning."(Walker, 1983, p

Lawrence Kohlberg
Moral judgment competence is "the capacity to make decisions and judgments which are moral (i.e., based on internal principles) and to act in accordance with such judgments" (Kohlberg, 1964, p. 425) Absolut L 1/3 Stage is more effective (Berkowitz et al., 1981).
L Counter arguments are at least as effective (Walker, 1983), more practical and are better in line with cognitive-developmental theory (Lind, 2003).

Intensity of Session:
Several dilemmas in a 45-minute sessions are effective, yet the effects do not last long.
L 90-minute sessions with only one dilemma allow more time for clarifying students' perceptions of the dilemma and more student involvement, and produce more lasting effects (Lind, 2002a).
3. Frequency of Sessions: Every week one or two sessions over several weeks.
L Three to four sessions already produce lasting effects.
L Longer intervals (of two or three weeks between) are as effective and prevent boredom (Lind, 2002a).

Teacher Interventions:
Teacher intervenes frequently in the discussion process.
L Teacher should be a facilitator rather than a moralizer (hidden curriculum).L Teacher should provide opportunities for role-taking rather than role-playing.
L The structure of discourse must be moral, not merely the vocabulary (Lind, 2003).

Structure of Session:
The method consists only of the dilemma and some probing questions; no further didactical aids are given.
L Didactical guidelines based on modern learning theories need to be worked out (e.g., alternating phases of support and challenge, wait-time) L Semi-real dilemma situation: the dilemmas are hypothetical, but the moral feelings which they trigger should be real (Lind, 2003).
6. Evaluation: For evaluating dilemma discussions, preferences for certain kinds of moral reasoning (e.g., principle moral reasoning) are assessed.
L Now tests of moral cognition & competence are available which cannot be faked upward yet are sensitive enough to measure even small effects of educational interventions (Lind, 2002b).
Moral Dilemma Discussion -"The Konstanz Method" a. Creating an effective learning environment i. Developmental moral tasks: Moral dilemmas ii.
Free moral discourse: Government by rules rather than authority iii.Conditions of optimal learning: Support and challenge b.Building support: i.
Clear and simple (hypothetical) dilemma ii.
One dilemma, not many iii.Sufficient time for clarifying the dilemma core iv.Clear and simple rules for controversial discussions v.
Intermitting small group work vi.Opportunities for reflection and meta-cognition vii.Sufficient time for discussion c.Providing challenge: i.
Difficult decision related to a moral dilemma ii.
Voting on a decision in a social context iii.Articulating opinion and reasons in a social context iv.Defending one's stance against opponents v.
Reflecting on counter-arguments d.Reflection and evaluation i.
Reflection on the learning process as an integral part of dilemma discussion ii.
Cognitive and affective criteria for the evaluation of dilemma discussion e.Stumbling blocks for a good results i.
Too much support (little challenge) ii.
Too much challenge (little support) iii.Too much "teaching" (little free discourse) iv.In appropriate observation tools v.
Wrong evaluation criteria e. Controversies: +1, Optimal Age, Direct Teaching, Ceiling Effect (Watson, Oser & Althof), No Effects (Prim, 1989;Uhl, xx) J. Piaget ( ): "From 11-12 years to 14-15 years a whole series of novelties highlights the arrival of a more complete logic that will attain a state of equilibrium once a child reaches adolescence at about 14-15 years.... The principle novelty of this period is the capacity to reason in terms of verbally stated hypotheses..... To reason hypothetically ... is a formal reasoning process.... [p. 3] From the social point of view, there is also an important conquest.Firstly, hypothetical reasoning changes the nature of discussion: a fruitful and constructive discussion means that by using hypotheses we can adopt the point of view of the adversary (although not necessarily believing it) and draw the logical consequences it implies.... Secondly, the individual will interest himself in problems that go beyond his immediate field of experience."(S. 4) M. Gross (1997)."The irony should not be lost here -socialization and moral development are effective only into early adulthood.Unfortunately, by the time the fruits of these efforts can be observed, they probably cannot be changed."(P.87) S. Uhl (1996): "Die Disäquilibrien entstehen hauptsächlich durch die Begegnung mit ranghöheren moralischen Antwortmustern, die das Individuum zwar als intellektuell anspruchsvoller empfindet und deswegen als anziehend findet (Rest 1976, S. 202-205, und 213-216) Aims: -Focus on the interaction of the students rather than merely on moral rethoric -Neither uncommitted play (word games) nor frightening reality -Integrating moral emotions (affects, attitudes) with rationality and reflection -Fostering moral judgment competence rather than changing moral attitudes and values Method: -Careful framing of the moral dilemma; moral = universalizable values or principles; dilemma = conflict between principles -Alternating phases of support and challenge, community and justice; -Priority, yet not exclusiveness, of behavior: the role of reflection and meta-cognition.(Turiel, 1966;Blatt, 1969)."(S. 377) G. Lind (1998): "I agree, there are very little good evaluation studies been done regarding the effects of moral/character education programs though there are many individual studies now which would lend themselves to such evaluations.The problems are manifold.There is little agreement

Developments in the
-on what standards should be used to evaluate these programs: Specific rule conforming behavior?General moral attitudes and values?Moral cognitions and competencies?(My own preference, for reasons I have outlined elsewhere, is the latter), -on the kind of measures for these standards: Should we use self-reports?Teachers' or parents' reports?Projective tests?Open answers?Attitude tests?Competence tests?(My preference, of course, is with competence tests), and -on the method of finding a conclusion from diverse sources of information: subjective summaries?
Preference for certain kinds of measures (like open interviews vs. closed questionnaires)?Counting significance tests?Averaging effect sizes (meta-analysis)?(Given the state of the art, the latter seems the only that can be defended on scientific grounds.) As far as I know, there have been done only six meta-analytic studies done in the field of moral/character education (Enright et al., 1981;Higgins, 1980;Leming, 1981;1992;Lockwood, 1978;Schlaefli et al., 1985).All focused on the cognitive-developmental approach, and one focused also on the values clarification approach.Other approaches, it seems, have not been subjected to scientific evaluation yet.
Four of the these three meta-analyses (Enright et al., Higgins, Leming, Lockwood) have dealt with moral education programs using Kohlbergs Moral Judgment Interview (MJI).They showed that the cognitive-developmental approach has consistent positive effects, while the values clarification approach did not.Unfortunately, these meta-analyses use only significance counting, which can mislead any conclusions, rather than effect sizes (like d, or r).
The sixth analysis used an effect size measure ("d") but provides only a restricted basis for conclusions since it focused only on studies using the DIT for measuring effects.Since the DIT is more a measure of moral attitudes (the subject's preference for principled morality) than moral competencies (as the MJI and Lind's Moral Judgment Test, MJT, claim to measure) its conclusions seem to be rather limited.For example, while most MJI based studies found that moral interventions are most effective in the ages 12 to 16, DIT based studies found substantial effect sizes only in young adults.
So Lind (2002Lind ( /1993;;also Lind, 1996: http://www.uni-konstanz.de/ag-moral/optimal.htm) analyzed the effect sizes of about 70 intervention studies mainly using the MJI (a few also used the MJT).He found a) that, like in Lockwood's and Leming's analyses, cognitive-developmental interventions had a consistent, positive effect on students' moral judgment competence.There is no study that showed a negative effect, and many studies which show high or very high effects, b) that the average effect size (r = .40)is very high as compared to effect sizes found in other areas of educational, psychological or medical intervention, c) the highest effects are reached with children ranging from about 12 to 16 years of age, and d) that only dilemma discussions, as proposed by Moshe Blatt and Lawrence Kohlberg, showed sustained effects after some time.
As to Geri's question about duration: three to four session seem to be necessary to produce a measurable growth of moral judgment competence.More sessions do not seem to have a proportional increase of impact.
For literature no consistent impact on moral development could be demonstrated so far.One reason for this seems to be, as Darcia Narvaez (1998) demonstrates, that the moral message of moral literature cannot be adequately understood by most children.Uhl (1996)."Die empirische Forschung hat immer wieder gezeigt, daß die reinen Diskussionsprogramme für sich allein genommen keinen nennenswerten Ertrag haben" (p.108) Quotations: Edelstein, W. (1985)."Teachers in their institutional roles have serious difficulties here.... It isn't the teachers' job to provide moral instruction.What they are to do is to organize the prerequisites for moral discussion in the classroom."(S. 395) Arbuthnot & Faust (1981): "Die Fähigkeit zur +1-Argumentation -gemessen an der der Schüler -ist notwendig, um bestimmte Aufgaben eines Moralerziehers zu erfüllen, und deshalb erscheint sie wesentlich für die optimale Moralerziehung."(S. 110) Hersh et al., (1978) 1. have conflicting claims 2. focus on a particular stage (sensu Kohlberg) 3. involve life experience that is real to the participants' situations.4. involve clearly defined facts so that the primary focus is ethical and not Effect sizes: Gain of Moral Judgment Competence by Quality of Education or InterventionEstimations of Gains per Year on a Scale From 0 to 100 (for the Konstanz method: gains per semester) Teachers Presents reasoning one stage above students' stage of moral reasoning.

4 .
What is the Impact of Schooling on Moral Competencies 5. What is the Impact of the Blatt-Kohlberg Method on Moral Competencies, and How Can it be Optimized?
a.The Mean Effect Size: Enhancing the Effect of Schooling b.Age Level "[T]he junior high school age was considered as 'optimal period for a moral discussion program, a notion supported by the results of Study 1." What is the Impact of the Blatt-Kohlberg Method on Moral Preferences?7. The Konstanz Method as a Response to Open Problems a. Identifying the Most Effective Features of Dilemma Discussion Principles of a comprehensive moral education: The forgotten 'Orestes effect').In G. Pollak, & R. Prim, eds., Erziehungswissenschaft und Pädagogik zwischen kritischer Refkexion und Dienstleistung.Festschrift zum 65.Geburtstag von Helmut Heid.Weinheim: Deutscher Studienverlag.Penn, W. (1990).Teaching ethics -A direct approach.Journal of Moral Education, 19(2), 124-138.Power, C., Higgins, A., Kohlberg, L. (1989) Lawrence Kohlberg's Approach to Moral Education.New York: Columbia University Press.Rest, J.R. (1979).Development in judging moral issues.Minneapolis, MI: University of Minnesota Press.Rest, J.R. (1986).Moral development.Advances in research and theory.New York: Praeger.Schläfli, A., Rest, J.R. & Thoma, S.J. (1985).Does moral education improve moral judgment?A meta-analysis of intervention studies using the Defining Issues Test.Review of Educational Research, 55, 319-352.Walker, L.J. (1983).Sources of cognitive conflict for stage transition in moral development.Developmental Psychology, 19, 103-110.
Feature One: Mutual Respect, especially moral and intellectual respect of the teacher for the children's views and values.References Berkowitz, M.W. (1981).A critical appraisal of the 'plus-one' convention in moral education.Phi Delta Kappan, March 1981, 488-489.Berkowitz, M.W., Gibbs, J. & Broughton, J. (1980).The relation of moral judgment stage disparity to developmental effects of peer dialogues.Merrill-Palmer Quarterly, 26, 341-357.( , aber auf seiner eigenen Stufe nicht voll verstehen kann.Das schafft den Anreiz, sich näher mit ihnen zu beschäftigen, sie zu begreifen In an extensive meta-analysis of the moral education literature, Schlaefli,Rest and  Thoma (1985)found that moral discussion techniques produced modest effects."(S. 15) Supportive and Challenging Learning Environment: Achieving an Optimal Level of Arousal and Attention b.Designing an Educative Moral Dilemma c. Support: Giving Sufficient Time for Thinking, Reflection, Clarification of the Dilemma on the Level of the Individual and the Group; Reducing the teachers' role in the discussion process (Edelstein, 1985, p. 395) d.Challenge: Creating Moral Emotions by Pressing for Decisions and Votes, by Identifying Opposing Groups (Sherif-and -Sherif-effect), and by Confrontation with Counter Arguments (Walker, 198., Festinger, 195.; ); e. Preventing Anxiety by Clear Rules and by Using Hypothetical Dilemmas f.Optimal Session Length: 80 to 120 Minutes g.Optimal Session Frequency: Every Other Week or Monthly h.Optimal Program Length: From Three to Any Number of Sessions i. Curriculum Integration: Part of Most or All Subjects Rather than Special Class j.Teacher Training: One to Two Semesters 7. Effective or not? S.