The 1990s is not the first time in our history that character education has captured the attention of American educators. In the first three decades of this century, character education became a major preoccupation. Such factors as increasing industrialization and urbanization, the continuing tide of immigration, World War I, the Bolshevik Revolution, and the spirit of the Roaring '20s, contributed to a mood among the population and among educators that social stability was being threatened and that moral standards needed to be strengthened. During the 1920s and 1930s, virtually every school in America was responding in some way to the educational goal of developing character (McClellan 1992, Yulish 1980).
The current revival of interest in character education, if it is to succeed, will have to successfully address the question of the assessment of program effectiveness. Today, unlike the beginning of this century, a body of research exists related to the topic of educating for character that can, if utilized and expanded, inform practice and assist in the development of effective programs.
The Early Character Education Movement
The character education movement of the first three decades of this century utilized elaborate codes of conduct and group activities in school clubs as the primary means to teach character (McClellan 1992, McKown 1935). A widely used code of conduct was the “Children's Morality Code,” which emphasized “ten laws of right living”: self-control, good health, kindness, sportsmanship, self-reliance, duty, reliability, truth, good workmanship, and teamwork (Hutchins 1917). Schools attempted to integrate such codes into all aspects of school life, and student clubs that harnessed the power of peer influence were created to encourage students to practice the virtues.
Between 1924 and 1929, the Institute of Social and Religious Research, funded by John D. Rockefeller and housed at Teachers College, Columbia University, undertook the Character Education Inquiry, the most detailed and comprehensive inquiry to date into the nature of character and the school's role in its development (Hartshorne and May 1928–1930). The study assessed the character-related behavior of 10,865 youths, mostly in grades 5 through 8, in 23 communities across the United States, focusing on student deceit and service. In schools where character education was taking place, the researchers created classroom situations that provided students with opportunities to cheat and to voluntarily engage in helping behavior. They found that the incidence of deceit varied widely in classrooms and schools and that deceit was situationally specific; honesty in one situation did not predict well to another. Furthermore, they found no relationship between membership in organizations that purported to teach honesty and honest behavior. Among the many disturbing conclusions within the 1,782 pages of text is the following: The mere urging of honest behavior by teachers or the discussion of standards and ideals of honesty, no matter how much such general ideals may be “emotionalized,” has no necessary relation to conduct.... there seems to be evidence that such effects as may result are not generally good and are sometimes unwholesome.... the prevailing ways of inculcating ideals probably do little good and do some harm.
The Character Education Inquiry raised serious questions regarding the effectiveness of heavily didactic approaches to character education.
Moral and Values Education
By the 1950s, character education curriculums had all but disappeared in American schools. The year 1966 signaled the beginning of a new period of interest. In the School Review, Lawrence Kohlberg for the first time linked his cognitive-developmental theory of moral reasoning with the practice of moral education in schools (1966). Louis Raths, Merrill Harmin, and Sidney Simon coauthored Values and Teaching, the highly influential first statement of the theory and technique of values clarification (1966). Although not the only approaches advocated, moral dilemma discussion and values clarification dominated the field of moral education for the next 20 years and were extensively researched.
How much impact these two approaches had on educational practice is difficult to judge. The values clarification approach was by far the more popular with teachers; one handbook of strategies for values clarification sold 600,000 copies (Kirschenbaum 1992), an almost unheard of figure for an education methods textbook.
Although the two approaches were different in many ways, they both emphasized that teachers were not to moralize. In Kohlberg's moral dilemma discussion approach, the teacher facilitated student reasoning, assisted students in resolving moral conflicts, and ensured that the discussion took place in an environment that contained the conditions essential for stage growth in moral reasoning. Values clarification sought to have each student clarify his or her values by following the prescribed seven-step valuing process. The teacher only facilitated the valuing process and, for fear of influencing students, withheld personal opinions. The teacher was to respect whatever values the students arrived at.
All of the reviews of the moral discussion research program have reached similar conclusions (Enright et al. 1983; Lawrence 1980; Leming 1981, 1985; Schlaefli et al. 1985); namely, that in approximately 80 percent of the semester-length studies a mean upward shift in student reasoning of one-fourth to one-half of a stage will result when students are engaged in the process of discussing moral dilemmas where cognitive disequilibrium and exposure to examples of the next highest state of moral reasoning are present.
The achievement of the predicted results of the moral discussion approach must be interpreted cautiously. First, the stage growth found as a result of the moral discussion approach is in stages 2, 3, and 4 and is small—usually less than one-third of a stage for interventions of at least one semester. Second, none of the moral dilemma studies reviewed used any form of social or moral behavior as a dependent variable. Kohlberg and his associates did argue that moral reasoning and moral behavior were related at the highest stages, the stages of principled reasoning (Kohlberg and Candee 1984); however, analyses of the evidence have detected only weak associations (Blasi 1980). One study found that among 4th and 8th grade students, stage 1 and stage 3 of moral reasoning are associated with fewer conduct problems than stage 2 reasoning (Richards et al. 1992). This finding raises the interesting possibility that raising students' reasoning from stage 1 to stage 2 may be associated with a deterioration in student conduct. Thus, even though the moral dilemma approach “works,” it appears to be of little practical utility in influencing students' behavior.
Research findings of the values clarification approach are also highly consistent: they show no significant changes in the dependent variables (Leming 1981, 1985, 1987). Whereas research on the moral dilemma approach involves only a single dependent variable (stage of moral reasoning), the values clarification research program contains a wide range of dependent variables such as values thinking, self-concept, attitudes toward the subject matter and the school, dogmatism, and value-related behavior. While the percentage of the studies finding the predicted results varies from dependent variable to dependent variable, the predicted change in a given variable is seldom found in more than 20 percent of the studies (Leming 1987). The research base for the moral and values education curriculums of this period offers little assistance in planning for character education where changes in student behavior is a central objective.
Research on Sex and Drug Education
Two research programs that have evolved over the past 30 years, although not typically described as moral or character education, have focused on character-related student behavior: sex and drug education. In his seminal review, Kirby (1980) concluded that sex education programs generally increase student knowledge about sexuality, change some attitudes—students become more tolerant of the sexual practices of others—but do not change students' values or sexual behavior. Most researchers in this field are in unanimous agreement (Dawson 1986; Furstenberg et al. 1985; Hofferth and Miller 1989; Kirby 1980, 1984; Marsiglio and Mott 1986; Reppucci and Herman 1991; Stout and Rivara 1989).
The most recent development in the field of sex education is value-based programs. These programs place sexuality within the context of human relationships and emphasize values such as respect for others, personal dignity, commitment, self-control, and abstinence. Two recent programs, one short-term and one long-term, illustrate the nature and level of effectiveness of these programs.
The Responsible Sexual Values Program in Franklin County, Ohio, presents a three-day instructional unit at the middle school level that integrates information about human sexuality, marriage, and parenting skills with group activities (Adameck and Thoms 1991). After the curriculum has been presented, volunteer student organizations provide peer support for the norm of abstinence. In addition, parents attend a two-session workshop on the goals of the program and participate in students' homework assignments. The first-year evaluation of the program indicated that student knowledge of sexuality increased, as did attitudes supportive of the abstinent lifestyle; however, by the fifth year any effects on attitudes and behavior had disappeared (Adamek 1993). Problems with faithful implementation of the program, as well as the increasing power of peer group norms, apparently combined to overwhelm any possible program effects.
A similar, but broader and more intensive value-based program is the School/Community Program for Sexual Risk Reduction Among Teens, developed at the University of South Carolina (Vincent et al. 1987). The K–12 program integrates value-based sex education information and activities within regular school subjects. Implemented on a countywide basis, the program involved not only parents but also clergy, church leaders, local newspapers, and radio stations. Special events raised community awareness. Two years following the implementation of the program, estimated pregnancy rates for females aged 14–17 had declined.
Value-based sex education that involves schools, parents, and the community in a common effort to encourage responsible sexual behavior appears to have some potential for changing adolescent attitudes and sexual behavior; however, due to the relatively few number of evaluations, caution should be used in attempting to draw generalizations from the data.
Three Waves of Drug Education
Both the public and teens see substance abuse as one of the most important problems facing America today. In the past 30 years there have been three broad shifts in the approach to drug education. Throughout the 1960s, drug abuse education largely provided information regarding the deleterious effects of drugs and used scare tactics to deter students from substance abuse. The “affective” or “humanistic” strategies of the early-to mid-1970s focused on teaching students personal skills such as problem solving and decision making and sought to develop positive health-related attitudes.
Reviewers of the research on these first two waves of drug abuse curriculums uniformly concluded that these programs tend to be successful in increasing knowledge, but less successful in changing attitudes; they have little or no effect on drug and alcohol abuse (Berberian et al. 1976, Kinder et al. 1980, Schapps et al. 1981).
Drug education in the 1980s shifted to the “social influences” strategy. This peer-centered approach tried to make students aware of the social factors that create pressures to use drugs and to help students develop the skills to resist those pressures through role-playing in class. Group activity and the discussion of personal experiences seek to develop group norms against drug abuse.
Three recent program evaluations conducted with upper elementary or junior high school students suggest that this approach has promise. Flay and associates found that a six-lesson smoking unit resulted in reduced smoking behavior (1985). Dielman and associates found that a four-lesson social influence program reduced the rate of increase of alcohol use among 6th graders who had already begun to use alcohol (1989). Ellickson and Bell found that their eight-lesson program reduced alcohol and tobacco use among junior high school students but did not sustain the reduction over time with regard to alcohol use (1990).
Reported data from the first two years of the Midwestern Prevention Project also support the social influences strategy (Pentz et al. 1989). A longitudinal trial of a primary prevention program of cigarette, alcohol, and marijuana use, this comprehensive, community-based program uses school study, parent involvement with homework, the mass media, community organization, and health policy programming to combat adolescent drug use. The entire adolescent population from 15 midwestern communities constitutes the sample. In the first two years, 22,500 6th and 7th grade students were exposed to the program. By the second year of what is to be a six-year evaluation, the use of all three target drugs was lower among students in the program for one year than among students just entering the program.
From the review of drug education programs, we can conclude that (1) gains in knowledge were common to all such programs; (2) lecture appears to have the smallest effect on attitudes, while peer programs have a greater influence and; (3) social influence programs appear to be the most effective in reducing the incidence of drug usage. Although effects declined in all the programs over time, such attrition was generally less substantial in the social influence programs.
The Impact of School Atmosphere
Introducing another angle on character education, a number of school-based research projects have investigated the relationship between school atmosphere and student behavior.
One of the major educational success stories over the past decade is the use of cooperative learning strategies. In cooperative learning, students are placed in small groups where the group learning assumes central importance and students are responsible not only for their own learning but also for the learning of others.
This type of learning environment organization has resulted in impressive student achievement and positive social values and behavior. Reviews of the extensive literature on this topic have found that in addition to increasing academic achievement, students learned to get along better with students of other races and ethnic groups, were more accepting of main-streamed students, demonstrated greater mutual concern for one another, and were more likely to engage in prosocial behavior (Johnson et al. 1981, Slavin 1990).
Another approach that emphasizes student responsibility is the just community approach developed by Lawrence Kohlberg and his associates. In the late 1970s, Kohlberg revised his perspective on moral education, emphasizing collectively derived social norms rather than individual values as a goal of moral education (Kohlberg 1978, Power et al. 1989). This new emphasis grew out of Kohlberg's experiences chronicling the democratic development of norms required to organize the social environment in small alternative schools.
In the just community approach, students confront real problems related to the social organization of the school. Within a democratic context, students discuss group problems and develop the norms by which group life is organized. For example, in a four-year study of the Cluster School in Brookline, Massachusetts, Power and Reimer identified four problematic areas in the life of the school: race relations, stealing, drug usage, and absenteeism (1978). Through a process of collective deliberation, students and teachers proposed and agreed on norms for behavior. The group then enforced compliance. Because this approach harnessed strong peer pressure within a democratic context, students eventually modified antisocial behavior in three of the four normative areas. In the case of drug usage, no collective norm emerged because the students did not share the teachers' perception that such a norm was needed.
While the just community research is based on an atypical educational setting, there are encouraging data from research on school climate in more typical school settings. Several studies have shown that schools that seem to have an impact on student character respect students, encourage student participation in the life of the school, expect students to behave responsibly, and give them the opportunity to do so. Discipline is not always imposed, but within the framework of shared group norms, students accept discipline as legitimate and change their behavior accordingly (Minuchin et al. 1969, Boesel 1978, Coleman et al. 1981, Rutter et al. 1979).
Recently, community service programs have gained popularity as a potential means for shaping youth character (Nathan and Kielsmeier 1991). Such programs place students in activities designed to exert a positive influence on the community. Studies of earlier but similar programs have detected only small effects on students' sense of civic responsibility as a result of these programs (Conrad and Hedin 1982, Holland and Andre 1987, Rutter and Newmann 1989). This pattern of results may be due in part to the inability to control the nature of the students' experiences in the field setting, resulting in uneven and less than desirable experiences for many students. In addition, the programs were electives, and student commitment varied.
Contemporary Character Education Programs
With striking similarity to the 1920s, the late 1980s and early 1990s have been a time of feverish activity with regard to character education (Lickona 1991, National School Boards Association 1987). Also like the 1920s, few of these new character education programs have systematically evaluated their effects on children.
Two approaches exist to the evaluation of contemporary character education programs. The first relies on informal evaluations that collect anecdotal evidence or survey teachers and administrators; it does not attempt to control for potential bias in information on student behaviors, nor does it compare students within the programs with nonprogram students.
One program that uses informal evaluation is the American Institute of Character Education in San Antonio, Texas, which produces character education curriculum materials for grades K–6. The curriculum consists of posters and a set of stories that illustrate values such as honesty, kindness, and generosity. Teachers discuss the stories with the students and make the students aware of the relevance of the values in their lives.
By the late 1980s, the Institute claimed to have reached as many as 18,000 classrooms in 44 states with its curriculum. Supporters assert that the program has reduced alcohol and drug abuse, encouraged school attendance, and discouraged vandalism, but these claims are supported entirely by testimonials (Goble and Brooks 1983).
A second character education program that has only informal data on its effectiveness is that of the Jefferson Center for Character Education. The Jefferson Center program attempts to teach the language of values such as honesty, perseverance, respect, and tolerance. Teachers help students learn the words, understand the concepts, and finally recognize and practice the appropriate behavior. The schoolwide program utilizes examples from the students' experiences and the school environment to reinforce the value behaviors.
Beginning in fall of 1990, 31 elementary and middle schools in Los Angeles, California, implemented the program. In the fall of 1990 and in the spring of 1991, 20 elementary and 5 middle school administrators were asked by phone to assess the program's effectiveness. The administrators stated that all forms of reported discipline problems had decreased, student morale had increased, parents had become more involved in the life of the school, and students were acting more responsibly (Satnick 1991).
The second evaluation approach utilizes experimental designs, focuses on student behavior, compares program students with nonprogram students, and attempts to control for potential sources of bias. Two recent character education programs that have assessed their effectiveness through controlled research efforts are the Weber County Character Education Project and the Child Development Program.
The Weber County K–6 program in Utah is based on a five-step teaching model that (1) stimulates interest in a principle, (2) models the principle, (3) integrates the principle with prior knowledge, (4) involves parents with homework, and (5) extends principles learned into real-life situations. The program currently involves 3,000 students and 109 teachers.
A grant from the Thrasher Foundation has enabled the Institute for Research and Evaluation of Salt Lake City, Utah, to undertake a longitudinal study of the effect of the Weber County program on the character of youth (Weed 1993). Over a two-year period, teachers in the program classrooms reported a statistically significant two-and-one-half-times reduction in problem behavior in students. In contrast, problem behavior in control schools actually increased. Additionally, it was found that students who had been in the program for two years and had moved into 7th grade scored significantly better on attitudes against substance abuse and in favor of positive school conduct than did nonprogram students.
A second controlled evaluation of a character education program is that of the Child Development Project in San Ramon, California (Solomon et al. 1987). Supported by an initial grant from the Hewlett Foundation, this project implemented a K–6 character education program in three elementary schools. The curriculum consists of five components: (1) teacher highlighting and exposing students to prosocial examples; (2) cooperative learning activities; (3) use of children's literature and classroom incidents to develop respect, sensitivity, and understanding toward others; (4) involving children in helping relationships; and (5) fostering moral reasoning and self-control through student-centered developmental discipline.
The program has been formally evaluated using a longitudinal design that tracked students over a seven-year period from kindergarten through 6th grade (Benninga et al. 1991). The evaluation collected data using hypothetical-reflective interview measures of social problem-solving skills and observational data on four types of classroom behavior: (1) supportive and friendly behavior, (2) negative behavior, (3) spontaneous prosocial behavior, and (4) harmoniousness.
The study found that after five years in the program, students scored significantly higher on measures of sensitivity and consideration of others' needs. In addition, they preferred conflict resolution strategies that were more prosocial (Battistich et al. 1989). The program's impact on character-related student behavior showed mixed results. No difference was detected between comparison students and program students with regard to the incidence of negative behaviors. On the other three behavioral variables, the evaluation did detect a significant difference favoring the schools using the program when data from all five years were combined. These differences, however, were not consistent across all grade levels (Solomon et al. 1988), and when cooperative group activities and teacher competence were used as covariates, no program effect was detected on the variable of harmoniousness.
The evaluation also followed a group of program students into junior high school, obtaining positive results on a wide range of variables such as democratic values, social understanding, feelings of loneliness, social anxiety, and higher-order reading comprehension. Out of a total of 105 program versus comparison tests, however, only in 11 cases did these tests favor the program students (Developmental Studies Center 1993).
The original six-year evaluation found that the detected prosocial behaviors did not generalize outside of the program classrooms and classmates (Solomon et al. 1987). The junior high extension data found that teachers did not rate the prosocial behavior of program students any different from comparison students (Developmental Studies Center 1993). In this regard, the results from the Child Development Program are consistent with the findings of Hartshorne and May (1928–1930); namely, character-related behavior and moral behavior tends, to a large extent, to be situationally specific.
What We've Learned So Far
- Didactic methods alone—codes, pledges, teacher exhortation, and the like—are unlikely to have any significant or lasting effect on character.
- The development of students' capacity to reason about questions of moral conduct does not result in a related change in conduct. Apparently, one cannot reason one's way to virtuous conduct.
- Character develops within a social web or environment. The nature of that environment, the messages it sends to individuals, and the behaviors it encourages and discourages are important factors to consider in character education. Clear rules of conduct, student ownership of those rules, a supportive environment, and satisfaction resulting from complying with the norms of the environment shape behavior.
- Character educators should not expect character formation to be easy. Schools that expect easily achieved and dramatic effects will be disappointed.Several limitations of the current research are worth noting.
- The majority of programs have been limited to elementary schools. Given that the rise of the current character education movement has been prompted largely by adolescent risk-taking behaviors, this focus is somewhat puzzling. Because research with adolescent samples has shown that it is difficult to sustain program effects over time, it is important that research on character education for this group receive intensive attention.
- Few carefully controlled evaluations of character education programs exist. Research cannot inform practice with only informal evaluations of low generalizability.
- All studies that have utilized multiple classrooms have detected considerable variations in program effects between classrooms. One possible explanation is differences in program implementation. Another possible explanation is the nature of the teacher and the climate established in that classroom; some exemplary teachers may naturally establish classrooms climates that promote good behavior. As in the Character Education Inquiry, we need to recognize that classroom differences are the rule, rather than the exception (Hartshorne and May 1928).
Finally, those interested in character education have long believed that morally inspiring literature should be a part of any character education program. Surprisingly, not one research study has attempted to assess whether reading such literature has the expected effect on character.
Future Directions
If character education in the 1990s is to avoid the “confusion and lack of clear knowledge” that beset character education efforts in the 1920s (Yulish 1980), a research base to inform practice must be developed.
A useful theoretical perspective that has some potential for unifying the field of character education is found in the works of individuals as diverse as Aristotle, Emile Durkheim, John Rawls, and Lawrence Kohlberg. These authors believe that three levels of development related to the formation of character exist. At the lowest level, rules are external to the child and behavioral conformity is assured through discipline and self-interest. At the next level, rules are embodied in social groups, and compliance with the rules is the result of youths' desire to gain acceptance within that group. At the highest level, rules are interpreted in terms of self-chosen principles.
Existing research supports this framework. That character is fostered by clear rules—fairly enforced—and by orderly classroom and school environments suggests that discipline is an essential element of moral education. The influence of cooperative learning methods and just community environments on student character suggests mechanisms by which schools can utilize the dynamics of attachment to groups in a positive pro-character manner. In addition, the newest wave of sex and drug education programs that elicit help from peers, parents, and community in defining and supporting appropriate behavior have been found to be the most effective to date in changing student behavior. Finally, the moral dilemma discussion methodology provides a means by which schools can assist youths in the development of their moral reasoning so that, as adulthood approaches, they develop the capacity for principled moral reasoning.
The development of a “grand theory” of character education and research based on that theory is a crucial next step in the future of character education. At present, atheoretical thinking and research on character education hampers the effort to develop effective programs. The current research in the field consists of disparate bits and pieces of sociology, philosophy, child development research, sociopolitical analyses, and a variety of different program evaluations. While such diversity is inevitable, character education needs to develop a more coherent view that can integrate the available research, provide focus to the movement, and guide the curriculum planning and research in a way that yields cumulative knowledge regarding the schools' role in fostering character.
