The intentional teaching of morals and values was common in American public education for almost three hundred years and had widespread public support until the 1930s. In recent years there has been a call by some educators for a return to a curriculum that explicitly addresses moral or character education (Kilpatrick, 1992; Wynne & Ryan, 1993). The teaching of character is defined as a central educational responsibility (Wynne & Ryan, 1993) and a way to stem the increase in violent and anti-social behavior which is due, in part, to the "de-moralizing" of our schools and society (Himmelfarb, 1995:16).
Those in favor of including moral or character education argue that if schools work cooperatively and comprehensively with parents and other social institutions, the outcome will be an increase in ethical and proscocial values and behavior (Wynne & Ryan, 1993). Curriculum revision is one of the ways proposed to teach moral or character education (Wynne & Ryan, 1993), with the reading of appropriate literature advocated as a useful technique to "help youngsters grow in courage, charity, justice and other virtues" (Kilpatrick, 1992:268). Administrators like Henry Huffman, an assistant superintendent of schools for Mount Lebanon, Pennsylvania, have put this philosophy into practice by making character education part of the mission and curriculum of his school system for the past six years (Bates, 1995). According to the Character Education Partnership, a Virginia-based group that promotes the teaching of values in schools, about one out of five schools in the nation offers some sort of character education program (Hart, 1995:1).
Advocacy of the use of a literature-based approach has been growing for the past twenty years in the clinical literature of educational psychology and social work. This technique to enhance self-awareness and reinforce positive attitudes is sometimes called bibliotherapy and more specifically, educational/humanistic bibliotherapy or developmental bibliotherapy (Hynes & Hynes-Berry, 1986) when the facilitator is not a therapist and the participants are students. Although one scholar has cautioned that bibliotherapy has been awarded a "scientific respectability it does not have," and that the application of the technique "far outstrips the tight validating studies supporting its use" (Riordan, 1991:306), numerous articles advocate it. Dana and Lynch-Brown (1991) studied gifted children and concluded that books are helpful in teaching them to relate moral principles to real-life situations. Lindsey and Firth (1981) suggest that books can help students identify with and model appropriate behaviors of real and fictitious characters. Calhoun (1987), Lenkowsky and Lenkowsky (1978), Hebert (1991), Miller (1993) and others advocate using literature as a guide to self-under-standing and a way to improve adolescent self-concept. Others recommend bibliotherapy as an effective deterrent to substance abuse (Bump, 1990; Pardeck, 1991) or as a way to help children deal with parental divorce (Early, 1993) or their own physical abuse (Pardeck, 1990). Each of the articles reports anecdotal evidence about the technique or changes over time in one group of students. Our review of the literature found no studies of bibliotherapy or other literature-based approaches to teaching attitudes, values or character traits which use an experimental design to evaluate the effectiveness of the method. It is this additional data which our study seeks to provide.
In order to evaluate the effectiveness of a literature-based approach on the extent to which students develop values, one specific value, caring for others, was selected as the dependent variable and an experimental project was designed. Our hypothesis was that reading and guided discussion of books which stress the theme of caring will have an effect on the extent to which students support this value. Because current developmental theories argue that boys and girls typically proceed through adolescence in distinct ways and learn different moral imperatives relative to caring, responsibility, and personal identity (Gilligan, 1982; Bernstein & Gilligan, 1989), a second focus of the research was to explore the differences between male and female students in caring attitudes and in their responses to reading literature with a theme of caring.
The study was conducted in an industrial New England community located between Boston and Providence, with a population of approximately 40,000. At present, 94% of the population is non-Hispanic and white, 3% is Hispanic, 2% is Asian, and 1% is African American. A team-taught seventh grade in one of the three middle schools in the community was selected. The class had two teachers, 57 students, and a room which was often divided into two rooms using a movable partition. The normal class structure varied between two teachers for all 57 students, divided into two classes, each with its own teacher, and two teachers working with many small groups. The students were between 12 and 14 (with an average age of 13.1) when the study began; 31 of the students were girls and 26 were boys; 51 were white, and 7 were Asian. For this study, the students were divided into two groups, matched on the demographic characteristics of gender, race, age, and teacher's assessment of their academic ability. One of the groups was randomly selected to be the experimental group while the other became the control group.
The Independent Variable: The Reading Project
The independent variable was the reading project designed for this study. Students in the experimental group worked with one of the teachers for ten weeks during the class time designated for reading. The project consisted of reading and discussing three books, chosen by the researchers in consultation with the teachers on the basis of the quality of writing and theme. In each book, the theme was the importance of caring for others as the main character makes choices regarding his or her behavior toward people who began either as strangers or friends. In each instance, the main character ultimately decides to care for others, even if it means sacrificing personal goals and gains. In all three books the main characters are adolescents, two are female and one is male, two are white and one is Asian. (See Appendix 1 for brief descriptions of the three books.) The students in the experimental group participated in classroom discussions and exercises designed to reinforce the theme presented in the books. (See Appendix 2 for a description of the classroom exercises and the training the teachers received.) During the same class time, students in the control group worked with the other teacher in the other half of the classroom space, reading and discussing the books normally read in this school system in seventh grade (specifically, books with animals as the main characters, such as Call of the Wild).
The Dependent Variable: Caring for Others
The dependent variable, support for the value of caring for others, is defined as the extent to which students "see and respond to need" and advocate "taking care of the world by sustaining the web of relationship so that no one is left alone" (Gilligan, 1982:62). The extent to which students support this value was measured both before and after the reading project by a set of three essays on caring for others. Students in both groups completed the essays as part of their in-school work during the two weeks before and the two weeks after the ten-week reading project (in January/February and May/June of 1994). The teachers asked the students to write the essays as part of their work for the language arts parts of the curriculum, instructing them that there were no right or wrong answers. Students were asked to write their opinions or make up a story without worrying about what they were "supposed to say." The researchers were given copies of the student essays identified by code number. Due to absenteeism, some students did not do each of the essays twice.
The three essay questions that were used to measure caring attitude were(1):
1. John and Bill both have jobs in a movie theater. The manager is deciding on the work schedule for the weekend. Both John and Bill want Friday night off. John wants to spend the evening visiting his grandmother who is sick and in the hospital, and Bill has plans to go out with his friends. The manager tells them that he needs one of them to work on Friday night, but that they can either decide between themselves who will work or he'll flip a coin to decide. What do you think should happen and why? (Grandmother)
2. Susan reads-in the local newspaper about a family that lost all their belongings when there was a fire in their home. She reads that they didn't have much insurance to replace their belongings. She decides to help out by sending them her whole allowance of $5.00 a week for the next three weeks. Describe how you feel about Susan's behavior. (Allowance)
3. Sometimes people have small families. Do you think that friends can be like families? In what ways? In what ways are they different from families? Write about your opinions. (Friends)
Each essay was read for content by both researchers and coded into one of three categories. For Grandmother and Allowance, each was coded from most to least supportive of caring for others.
For the essay Grandmother, answers that defined John's and Bill's needs as equivalent were defined as least supportive of caring (i.e., "Bill and John should split the hours in half that Friday"; or "The manager should flip a coin since it's fair and quicker"). Answers that gave some priority to John's visiting his grandmother were coded as moderately supportive of caring (i.e., "I think Bill should work because he should understand that John really needs to see his grandmother this week. Next week John will work and let Bill have the time off'). Essays that identified John's desire to visit his grandmother as having higher priority than Bill's spending time with friends were classified as most supportive of the value of caring. Two examples are:
If Bill has any decency, then he should work instead of John. John seeing his sick grandmother is more important than going out with friends. Friends last forever, but grandparents don't.
I think John should get the day off because his grandmother is sick. It's a better cause for a day off. Bill only wants to spend time with his friends. Bill should understand why John should get the day off.
Inter-coder reliability was 95% for this essay. At the pretest, 63% of the sample (57% of the experimental group and 68% of the control group) were classified as most supportive.
Essays in response to the Allowance question were judged to be least supportive if the student disagreed with Susan's behavior. Examples include: "I wouldn't do that. Susan's five dollars wouldn't help very much anyway"; "I don't care. I think she was nice, but stupid to do that"; "Susan's behavior is not logical. If she read the newspaper I read, then she would be broke. I doubt $15.00 will help the poor family anyway." Essays that were partially enthusiastic about Susan's behavior or felt she might get some sort of reward for it were coded as being somewhat supportive of caring. The student who said, "She is very kind, but I don't think she should give all her money away. She should get a group of people and start a collection for the family so Susan won't have to give all her money," was placed in this category. If the student felt that Susan had done the right thing and expected nothing in return, the answers were classified as most supportive.
I think Susan is doing a great thing because most people wouldn't care. I also think she has a big heart to give up her allowance for three weeks. That is a really great thing to do.
I feel that Susan is a great person. I think that Susan giving them 15 dollars is better than a millionaire giving them 1,000 dollars, because that's all she had. To me that's very unselfish.
Inter-coder reliability was 95% for this essay; 75% of the sample (78% of the experimental group and 71% of the control group) was classified as most supportive of caring at the pretest.
For the essay Friends, student answers were coded into three categories: friends and families seen as very different from each other, friends and families seen as somewhat similar, and friends seen as serving the same emotional and support functions as families do, even if the people live in different households. An example of the first category is the essay that said, "I don't think friends are like family. Friends are people to hang around with and have fun with. You can have fun with family but it's different. I wouldn't consider my friends my family." An example of the second category, seeing some similarities between the two units, is from the student who wrote, "In some ways friends can be like family, sometimes you can trust them and they can care for you. In other ways, they aren't like family because they can't know you as well as family members do." Examples of essays that were coded as seeing the units as serving the same emotional and support functions include the following:
I think friends can be like family. You can like them as well as family and they can like you in turn. You can trust them even if you haven't grown up with them.
I think friends can be families. I think that friends can be families in the way that you can share secrets and help each other out. You can love each other in a way.
I do think friends can be like families. They can do special things for you, or just show you they care. The only way they are different is that just are not blood related. There could be other differences, but it doesn't matter as long as they care for you.
The inter-coder reliability for this essay was 94%. At the pretest, 59% of the sample (50% of the experimental group and 71% of the control group) thought that friends can provide the same kinds of emotional and social supports as do families.
Gender and Caring
In analyzing the values of the students at the beginning of the study, some differences were found between the girls' and boys' essays, but they were not statistically significant. The girls' responses to two of the essay questions were more supportive of caring, with boys' answers more supportive of caring on the other essay. Eighty-one percent of the girls and 62% of the boys gave the most caring response to the Allowance essay at time 1 ([[Chi].sup.2] = 2.92, df = 2, p = .23); 55% of the girls and 70% of the boys gave the most caring response to Grandmother ([[Chi].sup.2] = 1.3, df = 2, p = .52). On the essay Friends, 72% of the girls and 61% of the boys said that friends could be as supportive as families ([[Chi].sup.2] = 1.5, df = 2, p = .47).
Results of the Experiment
To determine the effect of the reading project, the essay each student wrote after completion of the reading curriculum project was compared to the one written ten weeks earlier. Each student was classified as becoming more supportive of caring for others, less supportive, or showing no change in values. The experimental and control group results are analyzed in the crosstabulation presented in Table 1. The majority of all students in both groups were consistent in their values at both time 1 and time 2, and showed no change in support for the value of caring. For the students who do show change, the results for two of the essays are in the hypothesized direction. There was a statistically significant difference between the two groups for the Friends essay; 30% of the experimental group and none of the control group became more supportive of the belief that friends can be like family in providing emotional and social support. For Grandmother, the results approached statistical significance, with 20% of the control group and none of the experimental group becoming less supportive of caring for others after the reading project and slightly more (3%) of the experimental group becoming more caring. For Allowance, the results, which were not statistically significant, were nevertheless not supportive of the hypotheses; 23% of the control group and only 8% of the experimental group became more supportive of caring values after completion of the reading project. Results of analyses controlling for gender are very similar to those with no control variable, and also demonstrate some support for the hypothesis that a literature-based curriculum can have an impact on caring values.
COMPARISON OF EXPERIMENTAL AND CONTROL GROUPS: CHANGE IN SUPPORT
FOR THE VALUE OF CARING FOR OTHERS
EXPERIMENTAL GROUP CONTROL GROUP
FRIENDS SEEN LESS 0 (0%) 2 (10.5%)
LIKE FAMILY FOR
EMOTIONAL AND SOCIAL
NO CHANGE 16 (69.6%) 17 (89.5%)
FRIENDS SEEN MORE
LIKE FAMILY FOR 7 (30.4%) 0 (0%)(*)
EMOTIONAL AND SOCIAL
(*N = 42, [[Chi].sup.2] = 9.7, d.f. = 2, p = .013)
LESS SUPPORT 0 (0%) 4 (20%)
NO CHANGE 18 (81.8%) 13 (65%)
FOR CARING 4 (18.2%) 3 (15%)(*)
(*N = 42, [[Chi].sup.2] = 4.8, d.f. = 2, p = .09)
LESS SUPPORT 1 (3.8%) 2 (9.1%)
NO CHANGE 23 (88.5%) 15 (68.2%)
FOR CARING 2 (7.7%) 5 (22.7%)(*)
(*N = 48, [[Chi].sup.2]= 2.99, d.f. = 2, p =.22)
DISCUSSION AND CONCLUSION
The strongest support for the hypothesis that a reading project can influence values is the finding that the experimental group became significantly more committed to the view that friends can fulfill many of the same emotional and social support functions as can families. On one of the other two essays, the change was in the hypothesized direction. The experimental group became more supportive of the caring value for the Grandmother essay, giving weak support for the hypothesis. However, for the Allowance essay, the control group became more supportive of caring than did the experimental group, a finding that does not support the effectiveness of the reading project.
Several issues need to be considered when interpreting the results of this research. The indicators of the value of caring might be questioned. The fact that a large percentage of the essays for students in both groups were supportive of caring values at time I is a particular source of concern. First, a question of validity can be raised. In a school setting, students might write the essays they think their teachers want to read rather than revealing their own beliefs. Second is the issue of variability; with only a relatively small amount of improvement possible, interpretation of the results is difficult.
Another methodological concern is the amount of missing data. Since a large number of students did not complete the essays both times, as many as sixteen cases were omitted from each analysis.
One interpretation of the results is that even with their methodological weaknesses, the findings are real, and there was some change in caring values as a result of reading and discussing selected literature. The changes were small, a reminder that values are not quickly or easily changed. Knowing that these students, like other American children and teens, watch an average of three hours of television per day (Time, 1995:74), spend time on many out-of-school activities, and have other curricula materials, with both implicit and explicit values presented, it may not be surprising that the ten-week reading project in this study had only a modest effect. Teaching values and developing character is a complex and time-consuming task. If the value system presented in much of the popular culture designed for young adolescents does not include a great deal of support for positive values and prosocial behavior, and if a school curriculum must cover many other topics, it may not be possible to do substantial character or moral education with small, discrete projects. Schools may need an entire curriculum, beginning in the early grades, where values such as caring for others are introduced early and reinforced in later grades. Such reinforcement could include the reading and discussion of books with specific values as well as other techniques, such as community service projects, the development of rules within the school, and the inclusion of families in curriculum development. Small reading projects may be able to affect values to some extent, but only when part of a larger framework will they be able to provide the kind of socialization the proponents of moral and character education advocate.
Reading Project Books
Patricia Hermes, Friends Are Like That, New York: Scholastic Inc., 1984.
The major character, Tracy, an eighth grade student, has to weigh being popular versus supporting her best friend, Kelly, who is a nonconformist and a social pariah. At first Tracy is enthralled with the attention she receives from the school's "popular" crowd including invitations from her first potential boyfriend. Most exciting is the invitation that she, but not Kelly, gets to come to Angie's party, complete with a promise of a blind date with a popular boy from a private school. The pulls on Tracy almost end her friendship with Kelly, but in the end she chooses her over the popular crowd. Tracy realizes that Kelly will not change, but that their friendship is very important to her. Tracy tells Kathy, "I like you just the way you are. Different."
G. Clifton Wisler, Red Cap, New York: Lodestar Books, 1991.
This book is based on the true story of Ransom Powell, who in 1862, at age thirteen, leaves his home in Maryland to join the Union Army. Standing just four feet tall, Ransom is inducted into the army as a drummer boy. Two years after joining the 10th Virginia Regiment, he and others in his company are captured and sent to Camp Sumter, a Confederate prison at Andersonville. Conditions are horrid - men are robbed of clothes and possessions, poorly fed, and receive almost no medical care. Through all of this, the prisoners in Ransom's company help and support each other in ways a family would.
After months of working on the burying detail, a confederate soldier offers him the job of drummer boy for the confederates while in the camp. With the encouragement of his friends, he accepts and gets better living conditions among the confederates, who adopt him as their friend. When the soldiers return to the field, they find him a position as an orderly for the camp commandant, Captain Wirz. Ransom finds a way to help his Union comrades by smuggling food for them and making maps for them to use in their escape attempts. When he is discovered, he is sent back to the stockade where he watches all of the rest of his company die of various diseases. At the end, with the support of other prisoners, he is included in a naval prisoner exchange and freed.
Minfong HO, The Clay Marble, New York: Farrar Straus Giroux, 1991.
Twelve-year-old Dara and her family are among the thousands who are forced to flee from their villages in war-torn Cambodia during the early 1980s. With her mother and older brother, she travels in an oxcart to Nong Chan, a refugee camp on the border of Cambodia and Thailand. It is at this camp that Dara and her family meet another family, with a daughter, Jantu who soon becomes her friend. They all become emotionally attached to each other, even though as her friend Jantu says, "What I have, and what you have, are leftovers of families. Like fragments from a broken bowl that nobody wants. We're not a real family."
Bombing forces them to move again. During this move Jantu's younger brother is wounded and must be taken to a hospital. Dara does not look for her mother, but goes with Jantu and her brother until they find an ambulance which will take them to the hospital. Data is afraid she will never again see her family, but Jantu tells her she must find them. She gives Dara a marble she has made from the mud, telling her that it is magic and will make her strong, brave and patient.
After a long search Dara reconnects with her mother and Jantu's family. It is clear that this unit is now one family. Jantu's sister, Nea, and Dara return under perilous conditions to the hospital. Each time, she is successful, which Data attributes to the magic marble. They all return to the camp together, but in the dark, Jantu is accidently shot by Dara's brother who is standing watch. After Jantu's death, the new unit of Nea, Nea's brother and grandfather and Dara, Dara's mother and brother decide to stay together. They set off to Dara's home where they intend to plant rice and farm. As they begin their journey, Dara realizes that there was never any magic in the marble, but rather that she had made her own magic.
Reading Project Training and Exercises
The teacher training consisted of a review of the basic techniques used in Webbing With Literature by Karen D'Angelo Bromley (Boston: Allyn and Bacon, 1991). These techniques involve selecting a core concept and using webs or lines to connect information related to the core. The lines are then connected to each other to represent supporting details between the information already connected to the core concept. The major concept emphasized was "caring" and a web was drawn around this theme for all three books.
Group facilitator and trainer, Paula Foster met with the teachers for two 2-hour sessions during which the structure of the classroom discussions was designed and activities to aid in the discussions were developed. Because both teachers were very comfortable with theme teaching and because their classroom was already literature-based, the sessions primarily centered around how to facilitate discussion that would emphasize the theme of caring (in the experimental group) and how to work with this theme during the group activities.
Activities for students in the experimental group included writing favorite quotes and feelings in a journal, making a group collage that expressed the theme and the feelings in each novel, and webbing how the characters in the books were connected to each other (e.g., by marriage, birth, friendship).
At the end of the research Paula Foster again met with the teachers to debrief them and to answer questions. The teachers thought that Red Cap was the least successful book in capturing the students' attention while The Clay Marble was the most successful. The teachers reported their attempts to connect the students' feelings and reactions to current events, which often brought up the "micro" versus "macro" issues of caring (i.e., friends versus families, individuals versus groups, countries versus families).
This research was supported in part by a grant from the Rhode Island College Faculty Research Fund. The authors thank the administration of Versa-Care Incorporated for their support. They are also grateful to Roger Clark, Kathleen Dunn, Michele Hoffnung, Tobi Krasnow, and the teachers and principal of the participating school for their assistance.
1 For example, an essay question we pretested asked "Sometimes in unusual situations, like during wars or disasters such as floods or earthquakes, people put themselves in danger to help others. Do you think that this is foolish or not?" Student answers tended to focus as much on how willing they were to put themselves in danger as they did on their desire to help others (i.e., "If I got hurt, then I couldn't help anyone else"). Since we were not interested in the variable of bravery or willingness to put oneself at physical risk, we did not use this essay question as one of the indicators of caring.
Bates, S. (1995). A Textbook of virtues, New York Times Education Life (Section 4A), January 8, 1995, 16-19, 43-45.
Bernstein, E., & Gilligan, C. (1989). Unfairness and not listening. In C. Gilligan, N. P. Lyons, & J. T. Hanmer (Eds.), Making connections: The relations worlds of adolescent girls at Emma Willard School. Troy, NY: The Emma Willard School.
Bump, J. (1990). Innovative bibliotherapy approaches to substance abuse education. Arts in Psychotherapy, 17, 355-362.
Calhoun, G. (1987). Enhancing self-perception through bibliotherapy. Adolescence, 22, 939-943.
Dana, N. F., & Lynch-Brown, C. (1991). Moral development of the gifted: Making a case for children's literature. Roeper Review, 14, 13-16.
Early, B. P. (1993). The healing magic of myth: Allegorical tales and the treatment of children of divorce. Child and Adolescent Social Work Journal, 10, 97-106.
Gilligan, C. (1982). In a different voice. Cambridge, MA: Harvard University Press.
Hart, J. (1995). Reading and righting: Hunting lost values; Boston schools eye adoption of a "character curriculum." The Boston Globe, February 16, 1, 28.
Hebert, T. P. (1991). Meeting the affective needs of bright boys through bibliotherapy. Roeper Review, 13, 207-212.
Himmelfarb, G. (1995). A demoralized society: The British/American experience. American Educator, 18, 14-21.
Hynes, A.M., & Hynes-Berry, M. (1986). Bibliotherapy - The interactive process: A handbook. Boulder, CO: Westview Press.
Kilpatrick, W. (1992). Why Johnny can't tell right from wrong. New York: Simon & Schuster.
Lindsey, J. D., & Firth, G. H. 1981). The Clearing House, 54, 322-325.
Lenkowsky, B. E., & Lenkowsky, R. S. (1978). Bibliotherapy for the LD adolescent. Academic Therapy, 14, 179-185.
Miller, D. (1993). The literature project: Using literature to improve the self-concept of at-rick adolescent females. Journal of Reading, 36, 441-448.
Pardeck, J. (1990). Bibliotherapy with abused children. Families in Society, 71, 229-235.
Pardeck, J. (1991). Using books to prevent and treat adolescent chemical dependency. Adolescence, 26, 201-208.
Riordan, R. J. (1991). Bibliotherapy revisited. Psychological Reports, 68, 306.
Time. (1995). Special report: The state of the union. Jan 30, 1995, 74.
Wynne, E. A., & Ryan, K. (1993). Reclaiming our schools. New York: Macmillan.
Paula Foster, MSW, LICSW, Director of Outreach, VersaCare Incorporated, Attleboro, MA.