A DEVELOPMENTAL STUDY OF CHILDREN’S BELIEFS ABOUT PUNISHMENT Thesis for the Degree of Ph. D; MICHIGAN STATE UNIVERSITY JEANNETIE MARIE HAVILAND 2 . ,... . V 2... ,. ,...a:..'ur....;re;. I 1. IV 1.: . "1.1.9.7.. :31: .. .2 :4. 197.0 I I \. v 13L} Fla. ..4. uh: 2,5. 5’... :2, 3.7: ; .Lvfis. :. . 3.22.5. ; g , $5.34.?! \\ .......,....... IMIIIIIIIIIIIIIIIIIILIIIII [If] lldichigan State University _ v—g— w. This is to certify that the thesis entitled A Developmental Study of Children's Beliefs about Punishment presented by Jeannette M. Haviland has been accepted towards fulfillment of the requirements for Ph.D. Psychology degree in ‘50 ,4 ‘a'agazyw + Date fife-we M 7/ 0-7639 ABSTRACT A DEVELOPMENTAL STUDY OF CHlLDHEN'S BELIEFS ABOUT PUNlSHMENT BY Jeannette Marie Faviland This study examined changes and differences among first, third, and fifth grade children's beliefs about the justice and efficacy of punishments. Situational influences on the develOpmental changes were examined. These included the indirect influence of the children's teachers and friends and differences in responses evoked by systematically different interview stories. Children and teachers in the first, third and fifth grades at three schools were given the Punishment Judg- ment Interview at the beginning and the end of their school year. at the end of the year teachers also responded to a Pupil-Interaction Scale and.students also responded to a sociogram (used to rate peer pepularity). The first prediction was that developmental changes would occur between first and fifth grade. This pre- diction was supported and used as a basis for hypothe- sizing cognitive content in beliefs about punishment. Within each class, each grade and over all grades, there Jeannette Marie Haviland were very significant changes from punitive, authori- tarian responses to restitutive, egalitarian responses. This change during the school year occurred to some degree even in classes whose teacher consistently gave punitive reaponses. Differences in belief changes due to situational variation were hypothesized to be exceptions to the general change from punitive to restitutive beliefs. First, there was a significant difference between the children's responses to interview stories which involved an adult and a child and those which involved two children. At each grade level, the more mature (restitutive) res- ponses were given when two children were involved, rather than a child and an adult. Secondly, the teachers had an effect on the change in the Judgment of children in their classes. In the first and third grades, the differences in response changes between the class of the most restitutive teacher and the class of the most punitive teacher was signifi- cant. Classes whose teachers gave restitutive responses were more likely to give restitutive responses than classes whose teachers gave punitive responses. How- ever, there were no significant differences among the fifth-grade classes. The fourth prediction was concerned with the Jeannette Marie Haviland influence of popular children on the beliefs of their peers. However, contrary to eXpectations, there were no differences between the responses of pOpular (chosen as "best friend" by three or more of their classmates) children and other children in the same class. Therefore, the effect of peer beliefs could not be tested in this way. The general conclusion of this study was that beliefs about punishment relate to other cognitive structures and change deve10pmentally in a similar manner. However, it was possible to retard or accel- erate the general response or response change by mani- pulating the interview or, more importantly, the adult teacher with whom the child associated for a school year. A DEVELOPMENTAL STUDY OF CHILDREN'S BELIEFS ABOUT PUNISHMENT BY Jeannette Marie Haviland A THESI 8 Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1970 A CKN OWLEDGMEN 'l‘ S I thank the members of my committee for their advice and encouragement, particularly Ellen Strommen and Lauren Harris, who were available and even consi- derate evenings and Saturday mornings. The administrators, students, and teachers of the Holt Sohool System, Holt, Michigan, and of the Arlington School System, Arlington, Virginia, were most gracious and helpful. I thank them for their patience. The members of my family need special thanks for help far beyond even what was requested. My husband Randolph for his time at all hours, my mother, Helen Jones, for assisting as an interviewer, and.my mother- in-law, Mary Elizabeth Haviland, for editing and typing. All my family and friends, as babysitters and bottle washers, deserve many thanks. ii TABLE OF CONTENTS IntrOductj-on O O O O O O 0 0 O O O O O O O Punishment Beliefs . . . . . . . . . . The DevelOpmental-cognitive Structure of Punishment Beliefs . . . . . . . . . Teacher and School Effects on Punishment Beliefs O O O O O O O I O O O O O O 0 Peer Effects on Punishment Beliefs . . Summary and Hypotheses . . . . . . . . Method . . . . . . . . . . . . . . . . . . Subjects . . . . . . . . . . . . . . . Schools and Teachers . . . . . . . . . Instruments . . . . . . . . . . . . . . Punishment Judgment Interview . . . . Pupil-Interaction Scale . . . . . . . Sociogram . . . . . . . . . . . . . . Procedure . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . Changes and Variations in Beliefs . . . Teachers' Beliefs and Reported Teaching PraCtices O O O O O O O O O O O O O 0 Peer POpularity and Punishment Beliefs iii 13 19 20 25 25 25 27 27 29 31 32 36 39 52 53 Conclusions . . Changes and Variations in Beliefs . . . . . Teachers' Beliefs and Reported Teaching Practices Peer POpularity and Punishment Beliefs . . . References . . Appendices . . Appendix A. Appendix B. Appendix C. Punishment Judgment Interview . Pupil Interaction Scale . . . . Punishment Judgment Categories. iv 55 55 6h 66 69 73 73 78 79 Table 10 Table 2. Tabla 3 0 Table ‘40 Table 5. LIST OF TABLES Multivariate 'X 2: Punishment judgnent differences and interactions for grade, sex, story, and response change . . . . Percent of types of response changes in each grade to teacher and peer punishinent interVieW e e e e e e e e e Chi-squares of differences in responses to the teacher stories for school systems D and H vs A by grade . . . . Comparison of a first grade taught by a teacher giving consistent restitutive responses with averaged fifth grade respOHSGSeeeeeeeeeeeeeee Comparison of response changes between classes of most and least mature teachers at each grade level . . . . . hO LLB h6 h8 SO INTRODUCTION This is a study of children's beliefs about the value of different types of correction or punishment and how these beliefs change during the school year relative to the be- liefs of the children's teachers and class peers. The study investigates the generality of these beliefs in various situations and relates the develOpment of such beliefs to the deve10pment of "moral judgments". There is a severe shortage of information about pun- ishment beliefs. The only references to such studies are in the literature on moral judgment, since discussions of retribution and correction have been linked historically to discussions of religious morals and ethics. In addition, there are few studies (Parke & Walters, 1967; Sears, Rau, & Alpert, 1965) of the actual effectiveness of different types of correction or punishment, but none of these are deve10pmental studies. Although beliefs and practices are not necessarily related, many theories of develooment and several studies of behavior indicate that incongruences between belief and behavior often lead to changes in one or the other to make them.more congruent (Piaget, 1932; Festinger, 1957; Kohlberg, 1969). The current study does not examine the relationship between belief and practice, 1 2 but reference to this literature is made to indicate the general paucity of information in this area. Given the general emphasis in the philosOphic litera- ture (e.g. Rousseau, 1911; Dewey, 1957) on the importance of correcting children's behavior, and the general emphasis in the psychological literature on child-rearing methods, it is surprising that there has been so little research on beliefs about punishment. We only emphasize its importance by quoting Dewey: Now it is a wholesome thing for any one to be made aware that thoughtless, self-centered action on his part exposes him to the indig- nation and dislike of others....But these influences are immensely overdone in compari- son with the assistance that might be given by the influence of social judgments...which enable an individual to see for himself what he is doing....we need a permeation of Jidg- ments on conduct by the method and.materials of a science of human nature. (1957: p. 321) Punishment Beliefs The available information on punishment beliefs comes from early studies by Barnes (1902) and Piaget (1922) on "moral Judgments". Piaget was pursuing the study of the de- velopment of ethical thought in man by analyzing its deve- 10pment in the individual child Just as philosOphers had attempted to analyze its development through reasoning or through raflher superficial cultural camparisons. Piaget, himself, discusses how the results of his studies are re- lated to the theories of J. M. Baldwin, M. Bovet, and, 3 principally, E. Durkheim. Durkheim in particular argued.that society imposes or reveals constraints on the child and that the child has no natural tendency toward morality. The implication is that the child is natively amoral and would remain so if he were not constrained to obey and later respect the moral- ity of his elders. Thus "good" becomes "duty". Equating good with duty offers many problems to Durkheim, For example, it leaves unanswered the questions about the deveIOpment of morality beyond conformity. 0n the other hand,rBovet, a French educational psycho- logist and philosopher, claimed that "duty" and "good" are separable, and that "good" evolves from respect. When indi- viduals mutually respect each other as equals, "commands vanish and turn into mutual agreement, and rules that have been freely consented to lose their character of external obligation. Nor is this all. For since the rule is now subjected to the laws of reciprocity, it is these same rules, rational in their essence, that will become the true norms of morality," (The Moral Judgment of the Child, p. 372) Piaget takes up Bovet's argmment about two kinds of "morality” -- "duty" and "good" -- and extends it with the results from his studies. Piaget found that very young children's morality was based on duty; they believed that the "good" was being obedient. Older children's morality was based on equality or Justice; they believed that it is h good to make things equal. Piaget illustrates this with examinations of children's discussions of the right thing to do in various situations -- stealing, lying, collective responsibility, punishments and so on. With reference to the punishment situations, Piaget found that there were two kinds of punishments, each related to one of the two kinds of morality. "There are, in the first place, what we shall call expiatory punishments, which seems to us to go hand in hand with constraint and the rules of authority." (p. 205) In this case the punishment is arbitrary and not related to the crime. As Piaget says, the punishment could be corporal or the taking away of toys or a hard task so long as it is irrelevant to the crime. ”And there are, in the second place, what we shall call pgnishments,py_reciprocity in so far as they go hand in hand with c00peration and rules of equality." (p. 205) In the case of reciprocity the response to the crnme is a purely social one in which the "bonds of solidarity" are restored by repairing, in some sense, the crime. The range of responses outlined by Piaget include that the transgressor (a) be isolated, (b) suffer the material consequences of his acts, (c) be deprived of the thing he misused, (d) suffer the same crime to be committed towards him, (e) offer retri- bution, or (f) be censured. Piaget observes that the number of reciprocity responses offered by Swiss children increases from about 28% at ages 6-7 to h9% at ages 8-10 to 82% at 5 ages 11-12. Piaget places little emphasis on these figures, offering them only to illustrate a deveIOpmental change in preferences for reciprocity. He does not mention whether there are more or fewer expiation responses than recipro- city responses, nor whether there are sex differences or IQ,differences. Barnes found the same deve10pmental differences as did Piaget and later replications by Lerner (1937), MacRae (1952), Johnson (1962), and Kohlberg (1963). Three of the later studies (MacRae, Johnson, and Kohlberg), however, also pointed out that beliefs about punishment, collective responsibility, objective responsibility and so on were not necessarily related on a two point scale of authority- oriented vs. equality-oriented. MacRae showed that an equal- ity-oriented, "mature" response in one situation did not predict a similarly “mature" response in another situation. MacRae's and also Johnson's studies also showed that "moral- ity" was not a unitary concept in the sense that Piaget had anticipated. A study by Haviland (1968) indicated not only that different areas of moral judgment were unrelated, but what within the area of punishment beliefs alone, the judg- ments differed from situation to situation. In adult-child situations adolescents believed that severe restrictive and physical punishments should be used. In peer situations adolescents were more likely to suggest reciprocal punish- ment as . 6 Because of differences in responses to stories of punishment, collective reaponsibility, immanent justice and so on, Kohlberg's further studies of moral judgment have not been studies of responses to these various situa- tions, but instead have been studies of moral dilemmas. Each response to a dilemma is classified according to a developmental, hierarchical scheme of reasoning about jus- tice. Kohlberg claims that this justice factor of moral judgment is a develOpmental-cognitive structure. (This concept will be discussed below.) As such it should be relatively free of situation specific responses (which Kohlberg equates with responses learned through reinforce- ment training). However, it must be noted that stage 1, the lowest (i.e., least mature) in the hierarchical schema of Kohlberg's six stages of moral judgment, requires that the justice of a solution be related to material or physical gain or loss. In this sense, Kohlberg is still relating his research to punishment. Neither loss of esteem nor isolation (stages 2 and 3) appear to Kohlberg to be a "pun- ishment", although Sears and Han considered them to be "punishments" in their studies of conscience. Later stages in Kohlberg's schema involve legal reciprocity or innate equality and.may be related to more advanced stages of reciprocity in Piaget's original description of it. If Kohlberg's results could be construed in this fashion as studies of punishment, they must imply that punishment 7 beliefs develOp according to Piaget's hypotheses about cog- nitive development. This interpretation, of course, con- flicts with the conclusions of Haviland's research. How- ever, Haviland did not offer develOpmental data; all her subjects were about the same age chronologically and intell- ectually. 0n the other hand, all of Piaget's (and possibly Kohlberg's) data were obtained from adult-child situational dilemmas. Neither investigator studied the situational variable systematically, that is to say, neither varied the type of transgression and the relationship between the criminal and his victim. Clearly, a developmental study using variously structured situational dilemmas is needed to resolve this issue. The Developmental-cognitive Structure of Punishment Beliefs It is necessary to consider the hypotheses concerning the develOpment of cognitive structures in order to under- stand why different beliefs about punishment are expected to occur at different ages or in particular situations. This discussion will examine why some responses might be labeled.more mature or comprehensive than others. The Piagetian hypothesis is that at every age the child has a concept of his world and his position in it. The hypo- thesis predicts that a child's response to any situation is determined primarily by his concept of how that response will affect his world and himself (if they are differentiated). 8 Piaget has shown that the child's concept changes in a pre- dictable manner with the child's ago. From studies of child- ren's beliefs about dreams, about justice, about conservation, about causality, Piaget has develOped a schematic description of the concepts that a child uses as he deveIOps. Each of these different concepts is characteristic of a particular stage. The successive stages are called sensorimotor, pre- Operational, concrete-Operational, and formal-Operational. The description of each stage attempts to examine both how the child gets information (assimilation) and how he uses it (accommodation). Piaget means that one should consider these stages as "real", not simply as useful linguistic tools. The stages are not stages in the sense that Gesell means -- a classi- fication of behaviors common to various age groups, nor are they stages in a Freudian sense -- a point in time when one's energies are concentrated on a particular sort of psychic problem. They are stages in which a child thinks of himself and his world in a particular way, which type of thinking may be described as a cognitive structure. Kohlberg (1969) has elaborated this description of cognitive stages by designing "rules" for defining a cog- nitive stage: "1. Stages imply distinct or qualitative differ- ences in children's modes of thinking or solving the same problem at different ages. 9 2. These different modes of thought form an in- variant sequence, order or succession in Ifidivi- duaI deveIOpmefit.... 3. Each of these different and sequential modes of thought form a "structured whole". A given stage-response...represents an underlying thought-organization.... h. Cognitive stages are hierarchical inte rations. Stages form an order of increasineg dIfferent- iated and integrated structures to fulfill a common function...." (1969, p. 352-353) Each of these aspects of the cognitive stage offers an empirical problem. Since Kohlberg assumes an inter- actional (genetic-environmental) process of develOpment, he discusses each aspect of this definition in relation to the interactional process. According to Kohlberg, the first statement implies that cognitive learning cannot be simply a process, inaccurate at first, of imbibing adult teachings, since then only quantitative differences, and not qualitative differences, would occur. Actually, this is not clearly necessary to invoke genetic responsibility. Insufficient information by itself might lead to differing reasoning processes, as a chimp not allowed to see its own body may be expected to have a qualitatively different way of thinking about and using his body. As for the second aspect, invariant sequence, it is logically tied to the fourth aspect of hierarchical i322- grations. Again these aspects are not logical necessities in the develOpment of cognitive stages, but are empriically testable notions about them. Some studies have shown that 10 certain aspects of thinking follow the rules of integration and differentiation. Kohlberg's analysis of children's developing beliefs about dreams does, for example. But his analysis of moral judgments leaves many unanswered questions. It is not entirely clear how each of his stages of moral judgment is more differentiated or integrated than its predecessor. However, Turiel (1966) has shown that the stages are relatively invariant. Turiel found that children preferred responses one or two levels, but not more, above their own, and could learn to use these res- ponses; on the other hand, they could use and apparently comprehend all responses below their own level but pre- ferred not to use them. The developmental process through which these levels are attained is not known. For example, Harrower (l93h) could not find evidence to show that upper- class British children use an authority-oriented stage of morality at any age. Harrower thought that his results reflected class differences; however, the impact of class differences on cognitive development has not been explored fully in the area of moral judgments. Therefore, one can see that this "requirement" for cognitive stages is part- ially explored, but empirical questions remain. The third aspect of the definition of stages -- con- cerned with the structured whole of the stage -- is in fact the main requirement of the Piagetian stage and it requires Y .38 820 ll prhmarily that the person studying data be able to find categories in which to put several kinds of reaponses. In an empirical sense it means that each response must have a component in common with other responses. It is left to researchers to find these components, although the compon- ents should not be arbitrarily chosen, but logically derived from the data. This, of course, leaves the hypotheses Open to constant re-evaluation and criticism. This discussion intends to show that Kohlberg's "rules" for defining cognitive stages are quite arbitrary. As we have shown, there is no logical necessity for a develOpmen- tel-cognitive schema to follow these rules, as long as deve- 10pment and cognition are defined independently of the rules and the rules are testable. Kohlberg agrees implicitly with both these conditions when he attempts to justify these rules by illustrating their usefulness in describing the develOpment of beliefs about dreams and of moral judgments (Kohlberg, 1969). Now, with respect to beliefs about punishment, we know from studies by Barnes and Piaget and replications by Lerner, MacRae, Johnson and Kohlberg, that beliefs do change from authority-oriented at about 6-7 years of age to more equal- ity-oriented around 11-12 years. However, no evidence is available to show whether this change is related to the "rules" quoted from Kohlberg as necessary if we are to accept 12 punishment beliefs as a cognitive-deveIOpmental phenomenon. Perhaps a more thorough study may indicate that the deve10p- ment of punishment beliefs is related to the development of other cognitive structures. A recent study of cognitive stages (pre-, concrete, and formal Operational) and moral judgment (justice) stages has shown that they develOp con- comitantly (Lee, 1968). Kohlberg (1969) also claims that acquisition of a particular mode of thinking, e.g. formal logic, is necessary to, but not sufficient for, the devel- 0pment of the corresponding moral judgment stage. These hypotheses may hold for punishment beliefs as well as for justice beliefs. On the other hand, perhaps more thorough study will indicate punishment beliefs, as Opposed to justice beliefs, are more directly linked to cultural beliefs, that is, they will be easily conditioned, non-hierarchical, and may not follow invariant sequences, or some combination of these factors. In so far as punishment beliefs are related to moral judgments (justice), they may be related to cogni- tive stages and follow invariant sequences; in so far as punishment beliefs are related to distinct cultural practices (parental and other adult's use of punishment, Opportunity for peer coooeration, direct verbal training, etc.) they may be easily conditioned, non-hierarchical, follow variant sequences, and so forth. More likely, neither of these extremes will be entirely the case. 13 Teacher and School Effects on Punishment Peliefs Since the studies of Hartshorne and May in the late 20's, a generally pessimistic feeling about the effective- ness of teachers and teaching in influencing children's morality has pervaded the educational and psychological literature. Hartshorne and May observed that "character training", among other socializing influences, had no effect on children's cheating. Cheating seemed to be largely de- termined by the situation, so that children could not be characterized as generally "honest" or "dishonest". Kohlberg (1967) claims that Hartshorne and.May's results are valid for children. Ethical considerations about honesty which apply across situations do not occur until the last stages of moral develOpment. Before then, children justify cheat- ing by the results it has on others; therefore, if there is a good possibility that the results will be pleasant, cheating is condoned. Nevertheless, two studies (Blatt, 1970; Turiel, 1966) have shown that direct teaching may affect the level of moral judgment. Exposure to levels of judgment one or two above that usually used tended to change that levelto one above it. Exposure to levels of judgment one or two levels below that usually used had less effect on moral judgment and the effect did not last for long. In each of these studies the changes affected in children's responses were brought about by intended training. 1h A single study by Graham (l9h6) indicates that teachers' political beliefs may influence students' political beliefs without any direct training in these beliefs. This small but important study has never been replicated. The only extensive study of indirect training is Rosenthal's (1966). In this study the teachers' beliefs about the children's ability (I.Q.) affected children's responses (and no doubt teachers' responses). The interactions in the Rosenthal study are too complex to admit examination of how teachers' beliefs affected teachers' behavior or how the teachers' beliefs or behavior affected children's beliefs or behavior. Perhaps in a real psychological sense these beliefs -- be- haviors dichotomies are inseparable (Festinger, 1957). There is no evidence that teachers' beliefs about what sorts of punishment are apprppriate and effective will or will not affect their students' beliefs about punishments. Neither do we know whether or not teachers' punishment be- haviors will affect their students' beliefs. Of course, this raises some important questions. If a teacher believes that the best way to influence an erring child's behavior is through physical punishment and restriction, will her pupils come to have the same belief? If her beliefs are related to her behavior will the punished pupils believe that their teacher is acting apprOpriately or inapprOpriately? And on the other hand, how will the teacher who believes I an 1 fies . i do" 38. \n ‘14“ ‘I 15 in using censure or social isolation on erring pupils in- fluence her pupils? This becomes particularly complex if belief in different sorts of punishment is a hierarchical develOpmental phenomenon, in which case a teacher might be cognitively mature or immature with respect to her pupils. For example, Kohlberg argues (seminar, Michigan State Uni- versity, 1969) that a possible contributor to some current student-administration clashes is that the students are Operating on a higher cognitive beliefs level than are the administrators. Although Kohlberg and his students have assessed the moral judgment of student demonstrators, they have not assessed school administrators' moral judgments, so the hypothesis remains untested. There are two possible conflicting hypotheses about the effect of teachers' beliefs, based on studies of jus- tice by Kohlberg's students. In one case Holstein (1969) found that children at age 12 who had parents using stage- 5 or 6 levels of moral judgment were more advanced in moral judgment beliefs than children whose parents used stage 3 or u levels. The level of moral judgment was not equi- valent to the parent's in either case, however. These re- sults lead to the hypothesis that the higher the teacher's level of moral belief (or punishment belief) the higher will be the level of her students' belief. 0n the other hand, Turiel's and Blatt's studies suggest that a.moral 16 judgment statement more than two levels above the level usually used by a child has little or no effect on his thinking. That is, a child does not comprehend levels of moral judgment far beyond his own and so rejects them. These results lead to the hypothesis that teachers who Operate on a level near that of their pupils would be most influential. There are then many possible interactions. An indi- vidual student or a class of students may at the beginning of the school year manifest punishment beliefs that are hierarchically lower than, equal to, or higher than the teacher's. Depending upon the degree of discrepancy, the direction of the difference, and the actual hierarchical nature of the beliefs, there are several possibilities. First, of course, beliefs about punishment may not be related to stages of cognitive growth. In this case, one would predict that most students in all classes, regard- less of the age or stage of the students or the beliefs of the teacher, would be reinforced directly and indirectly for adopting the teacher's beliefs. Furthermore, if these beliefs are related very strongly to particular situations, then one would predict that children's beliefs would best concur with their teacher's beliefs about school-related situations rather than home or play situations. If beliefs about punishment do seem to be strongly 1? influenced by an adult, regardless of the original discre- pancies in belief, then how will we account for the original findings of Piaget and Lerner, etc. that beliefs in punish- ment ordinarily change in a particular direction? One poss- ibility (also mentioned by Bandura, 1963) is that adults' beliefs about punishments for inferiors (e.g., children) and peers (e.g., other adults) differ. Children may be reflecting this difference in learning first about the sorts of punishments appropriate for children, and as they grow older and are more often accepted as "peers", learning about other sorts of punishments apprOpriate for "peers". Piaget suggests that the belief in reciprocal punishment may arise in part through peer relations in middle childhood. This hypothesis does not negate the possibility of changes in adult-child relations occurring concurrently. 0n the other hand, it may be that punishment beliefs are related to many other kinds of beliefs and that it, therefore, will be easier to change beliefs in a direction compatible with other cognitive changes. In this case the teacher who has punishment beliefs which are more, rather than less, mature than her students', should be more influ- ential in changing those beliefs. As mentioned above, an extreme discrepancy in beliefs in either direction may or may not be related to changes (Turiel, Blatt). To test the hypothesis that teachers whose beliefs 18 are more mature rather than less mature than their students have greater influence on studentsI beliefs, it is necessary to sample teachers with widely different beliefs. kohlberg claims that widely differing stages of moral judgment occur among young adults (college age). No information is avail- able on older adults. Several studies of teaching methods (e.g., Bierstadt, 1969) have suffered from homogeneity among teachers. Other studies have indicated that teachers as a group are rather authoritarian (e.g., Mohandy and Guha, 1967), and that they use direct (authority oriented) rather than indirect (peer oriented) methods of instruction (e.g., Bjerstedt, 1968; Measel, 1968). Many investigators, there- fore, have had difficulty in studying differing teachers' effect, since it was difficult to find a heterogeneous sample of teachers. This is a problem for a study of teacher effects on punishment also. This authoritarian trait in teachers, especially if it implies a high degree of punitiveness, may lead to the hypo- thesis that teachers in general inhibit the deve10pment of moral judgment. In this case it is possible that children will have one sort of beliefs about school situations and another about home and/or peer situations. 0n the other hand their beliefs might not be situation specific, thus leading them to accept the teacher's beliefs at one stage and reject these same beliefs at another stage. 19 Peer Effects on Punishment Beliefs Children's effects on one another have scarcely been studied. Sullivan, as an example from the psychoanalytic literature, suggested that children and children's insti- tutions such as the school exert a strong influence tOWard correcting personality "warps" during middle-childhood. Piaget suggested that cooperation among peers in middle- childhood contributes to the develOpment of reciprocal moral judgment. There is no published research on these issues, however. Kohlberg (1969) makes reference to strong peer influences on the develOpment of moral judgment, and perhaps further research will offer support for his claims. He also claims (1966) that children who are social "isolates" are morally immature on his scale when compared with their classmates. Research on conformity reveals that in both ambiguous and.unambiguous perceptual situations children will conform to a group judgment of the perception. But in the unambi- guous situation, the effect changes with age, increasing from ages 7 to 13 and decreasing thereafter (Hamm and Hov- ing, 1967). Berenda's (1950) study of peer and teacher influences on children's perceptional judgments in unambiguous situa- tions revealed striking differences. Peer's influence brought about a change in judgments in 8h.5 per cent of 20 the children tested. Only 37 per cent of the children fol- lowed their teachers' incorrect judgments. No child follow- ed the teachers' incorrect judgment in more than half their judgments, although almost half the children followed their incorrect peers in more than half their judgments. The teachers' influence was significant only for younger (7-10) children. In general, hypotheses about adult (principally parent) and peer influences on children may be summarized by (a) adults are more influential for younger children than for older and (b) conformity to group Opinion increases through- out childhood and decreases during adolescence. With re- ference to moral judgments or punishment beliefs, this leads to the hypothesis that young children will be more influenced by their teacher's beliefs than will older children, while older children will be more likely to conform to group or class beliefs than will younger children or adolescents. SummarLand Hypotheses Punishment beliefs have been studied as one aspect of beliefs about moral judgment. Piaget and others have shown that young children believe that authority-oriented punishments are most effective. That is, young children say that~authority-oriented punishments are most just and most effective in teaching the wrongdoer the evil of his 21 deed. Older children tend to believe that restitutive pun- ishments are best. In this case the criminal is punished in order to repair, as much as possible, the effect of the crime. This sort of response ranges from "an eye for an eye" response tO an eXplanation of the effects to the crim- inal so that he would understand himself and help repair the damage he has caused. This study examines this change in detail, asking questions about the influence of author- ities and peers and their beliefs on this change in child- ren's beliefs. The first hypothesis is concerned with the nature of punishment beliefs. It is possible that the development of children's beliefs about punishment corresponds to the develOpment of other cognitive factors ordinarily related to age, such as St On the other hand, they may depend upon situational variation related to the children's relation- ships with adults and other children. As pointed out in the discussion preceding this section, there is evidence for both points of view. Hypothesis I combines these two points of view. Hypgthesis I: Beliefs about punishment will change deve10pmentally in general, but there will be variations due to the status of the children and to the situation. (Is) One variation is hypothesized to be a difference in response to situations involving two children and situations 22 involving a child and an adult. Since the response given by young children is authority-oriented, it seems likely that situations which involve an authority (an adult) are more likely to evoke this kind of response than situations which involve peers. The second variation examined is the relationship between the authority and the child and its effect on the child's belief. (Ib) If teachers believe that authority-oriented punishments are best and if they use their authority to instruct their students about moral- ity, their students will have authority-oriented beliefs. However, if teachers believe that restitutive punishments are best, their students will have restitutive beliefs. Since Berenda found that younger children most often conform- ed to their teachers, the teacher effect is hypothesized to be greatest for first graders (Berenda, 1950). This hypothesis gives more weight to the theory that they are related directly to situational variables and modeling; in this respect it is consonant with Bandura's (1963) theory, not with Kohlberg's (1969). Hypothesis II: Hypothesis II is related to Hypothesis I in predicting changes related to variations in the teacher- student beliefs and relationship. At the end of the school year, children's beliefs will have become more consonant with their teacher's beliefs than they were at the beginning of the school year. This 23 hypothesis does not predict differences dependent upon hier- archical degrees of discrepancy between the teacher's and student's beliefs. That is, it eschews the differences found by Turiel (1966) for moral judgments. Hypothesis III: The third hypothesis is concerned with an examination of how teacher's beliefs in this area are related to their teaching of punishment beliefs. The teaching of punishment beliefs was measured by scripts written by the teachers of their behavior in situations where a class would be concerned with punishment. Teachers with restitutive punishment beliefs are more likely to ask their classes for discussion of punishment issues than to tell the class what the correct beliefs are. Teachers with authority-oriented beliefs are more likely to lecture their classes on punishment beliefs than to ask for class discussion. This hypothesis is based on the phil- osOphical-logical distinction between restitutive and auth- oritative beliefs which requires that-the behaviors assoc- iated with each type of belief involve either authority- subordinate behaviors (authority-oriented beliefs) or peer- oriented behaviors (restitutive beliefs). Hypothesis IV: Hypothesis IV is concerned with the influence of peer beliefs upon children's punishment beliefs. In accordance with Piaget's, Kohlberg's, and Berenda's studies and hypotheses the pOpularity of children is hypothesized 21+ to be related to their judgments. This is based upon the assumption by Kohlberg that pOpularity implies greater peer interaction for the pOpular child than for less pOpular children. Children who are pOpular (chosen as "best friend" by three or more of their classmates) are more likely to ad- vocate restitutive punishments than are their peers. METHOD First, third and fifth grade children at three schools were interviewed about their beliefs about punishment early in September and early in June. Teachers were rated on a pupil-interaction scale and on a punishment judgment inter- view similar to the one administered to the children. Changes from September to June were examined and compared according to grade, school, and, most importantly, teacher. A socio- gram was administered to the children in June, and students who were deemed most popular were compared with their class- mates on punishment judgment to determine the relationship between pOpularity and judgment. Sub ects The classes consisted of six 5th grades, four 3rd grades (one a combination of two 3rd grades and one 2nd grade with three teachers team-teaching), and four and one-half 1st grades (the half was 16 children assigned to a remedial teacher part of the day and in the regular first grade part of the day). schools and Teachers In order to obtain many teachers and, it was hOped, different kinds of teachers, three different public schools were chosen. 25 26 School D is located in a very small Michigan town. Most families in this town work in a nearby automobile factory or in agricultural enterprises. About half the teachers come from the community, while the other half commute from a university where they or their spouses are students. School H is located in a moderate-sized town in mid- Michigan. Schools H and D bus pupils, so the pupil pOpu- lation is similar. All the teachers in School H who parti- cipated in the study were eXperienced, tenured teachers. School A is located in a wealthy suburb on the peri- meter of Washington, D.C. Its administrators boast that it is one of the "best schools" in the country. All but one of the teachers who participated in the study had been teaching fewer than four years. These teachers came from - several parts of the United States, but none had grown up in the community. Although all three schools are public elementary schools with Caucasian rural or suburban pOpulations, there seem to be differences. Schools D and H appeared to be traditional schools with relatively rigid principal-teacher-pupil hier- archies. The classrooms are bright and cheerful, the teach- ers benignly authoritarian and well trained. The teampteach- ing attempt in the third grade at school D was the only "experimental" class. 27 At School A more attempts to have student-chosen and student-organized activities were in evidence than at schools D and H. Students ran a visitor's patrol which guided their frequent visitors around the school and directed strangers such as the experimenter. Bus patrols, library and lunch patrols similarly guided students and teachers through these activities. The teachers seemed enthusiastic about various kinds of experimental classes. Some classes had seminars for interested students. Many classes exchanged pupils for various functions, including an older class which helped teach a younger class. Although students and teachers at School A are not "free" in the sense that "free" is used in "free school", they seem to have more Opportunities for eXperimenting with education than is traditionally allowed. Group IQ;tests had been given by the schools to the children in kindergarten and second grade. At these grade levels there were no Itkdifferences between the children in Schools H and D and the children in School A. Instruments Both teachers and children were interviewed about their punishment beliefs. In addition all teachers were given the pupil-teacher interaction scale, and children in all grades were given a sociogram. Punishment Judgment Interview: The punishment judgment 28 interview consisted of six stories, each describing the com- mission of a misdemeanor. The subject was required to say what should happen to the child who committed the misdemeanor and to explain his reply if he could. The misdemeanor in each story was lying or stealing. The person lied to or stolen from was a mother, a teacher, or a child-friend of the thief or liar. Therefore there were two stories, one with a misdemeanor of stealing and one of lying, involving each victim, mother, teacher, or peer. The stories are comparable in all other aspects. Therefore the type of judgment could be related to the situation, lying or stealing, or the relationship, child-mother, child- teacher, or child-peer. The stories were derived from Piaget's (1932) original stories rather than from stories used more recently in the moral judgment literature (Johnson, 1962; Kohlberg, 1969). The original stories were designed for young children rather than for adolescents (frequently the subjects in Kohlberg's studies are adolescents), and are simpler. Because the stories used by Kohlberg and Johnson contain many elements (disparity in ages of characters, relationship between characters, severity of crime, number of alternatives Open to the criminal, etc.) differences in the type of response given to each story cannot be attributed to any one of these elements, but each disparity perhaps cancels the effect of 29 the others. The interview with the teachers differed from that with the children only in that the six stories of interest were embedded in six other filler stories originally used by Johnson. An earlier study (Haviland, 1968) found that some adolescents tended to give the same reply to each story, perhaps in an effort to be consistent, since the stories are very similar. It was hoped in the current study that interference from other stories would emphasize their dif- ferences. Three of Johnson's stories were used as filler for the fifth graders in pre-testing, but they proved to be too confusing and were not used in the study. It seemed as if a partial situational explanation of the problem inspired the children to ask for more explanation. A simple story (containing few variables) seemed to encourage generalization. The interview stories and the replies to them are con- tained in Appendix A. The appendix also indicates how the replies were classified. The replies were grouped according to the procedure outlined in the following results section. Pupil-Interaction Scale: A COpy of the pupil-inter- action scale is in Appendix B. It contains the written instructions which accompanied the scale. This scale was developed from a similar scale used by Gary Stollak. The scale is used to discriminate differences in the way that 30 a person perceives himself to communicate. For the purpose of this study only interactions between teacher and student(s) were counted. If the teacher said that she would solicit an Opinion, statement, or comment from a student she received a point for that situation. If the reply to the situation involved no interchange, no points were given. The scale describes three school situations and requires the teacher to complete the script for each. The teacher is required to describe her reaction as if she were writing a play. No reasons, thoughts, or feelings are required, only a description of the action and a record of the verbal interaction. The verbal interactions then were counted and used as a measure of teacher-pupil interaction. For this study the situations described (a) a pupil's favorable reaction to an account of a robbery, (b) a class's questioning of the teacher's beliefs about cheating, and (c) the teacher's reaction to a child looking in the teacher's desk drawer. A panel of three judges rated the quality of the inter- actions on a five point scale. (The lowest point on the scale was "one-way interaction with teacher instruction only” (1) and the highest was "teacher solicits student Opinion and.uses it to form her instruction" (5).) There were no differences in ranking teachers based on counting interactions and quality of interactions. All analyses 31 are based on rankings formed from the number of interactions between the student and teacher. Sociogram: After administering the punishment judgment interview in June, the members of each class were asked to write the name of his or her best friend in that class on the back of a slip of paper. The experimenter said that she wanted to find out whether children who were best friends had the same ideas about the stories she had just read. She promised to be very careful picking up the slips of paper so that no one would know what each had written. Occasion- ally someone asked if the experimenter meant boy or girl friends and the experimenter replied that it did not matter, just so long as it was the person he liked best in the class. In three of the fifth grades a girl asked whether she could name her teacher and the experimenter replied that she wished to compare only the choices of other children. The children understood the instructions in all the classes and all the children replied very quickly although it was obviously a matter of great merriment and secrecy for many of them. "Best friend" nominations were used as an indicator of peer interaction on the basis of an experiment mentioned by Kohlberg (1969, p. hOO). Kohlberg claims that "children highly chosen by their classmates" are faster in moral dev- elOpment than their peers. Since no other reference to KOhlberg's study could be found, it was hOped that "best 32 friend" would correspond to "highly chosen" in relation to the develOpment of judgments about punishment. Perhaps, though, Kohlberg means that the children were most Often chosen for certain activities, not necessarily involving friendship, but he does not indicate this specifically. It was hOped that a relationship between the views of pOpu- lar children and their friends could be shown, especially if the views of the pOpular children were more advanced than those of their peers. A child was called "pOpular" if he or she received three or more "best friend" nominations. Using this method (there were at least three children in each class who were called "pOpular". Procedure All September testing was carried out during the second and third weeks of school. First graders were interviewed individually by the experimenter and two assistants, one a middle-aged woman, one a young:man.* There were no inter- viewer effects. The children were selected by the teacher to meet the interviewers in the hall. The interviewer greeted each *I am deeply indebted to Gary B. Johnson, graduate student at Michigan State University, and Helen A. Jones, High School teacher. 33 child by name, having listened to the teacher, or asked the child his name. The interviewer then engaged the child in a casual conversation during which he or she explained to the child the purpose of the study. Usually the instruc- tions were as follows: I am studying children's ideas about different things that happen to them when they do something wrong. There are no right or wrong things to say about the stories I will tell you, you should just say what you think would be the best thing to happen, what would be the best thing for the child to teach him not to make the same mistake again, what would be the fair- est thing to happen for everybody. Quite often there would be a discussion of "ideas" and "opinions". The eXperimenter illustrated how different peOple could have different Opinions. In this case the child would also be asked to give an example of different Opinions about some subject, such as favorite flavors of ice cream. The presentation order of the interview stories varied randomly from child to child. Each child was read a story and then prompted to give his Opinion about what would happen to the child who "took something", or "lied to his friend", etc. (A OOpy of the stories is in Appendix A). An effort was made to help each child understand the story situation even if the stories had to be repeated or stated differently; therefore, the stories were not always verbatim OOpies of the printed outlines. Many of the first graders were shy 31+ and hesitated to speak to the interviewers, fearing that this was their first "test" in school. Every effort was made to dispel this feeling in the children. As often as not, introductions with the child took as long as the inter- view itself. Third and fifth graders were group tested in their 'classrooms. Each child had a blank piece of paper on which to record his Opinions. The instructions were the same as for the first graders. Third graders were told that they could guess on the spelling of words or raise their hands and the experimenter would come and help them. Again, the experimenter did not hesitate to explain the stories more fully or to answer questions. Since all three schools often entertain psychological researchers, all the older children were familiar with the testing situation. The teachers received a OOpy of the questions that were read to the children with the "filler" stories randomly interspersed among the six stories of interest. Between each story there was space to reply to the situation. The teachers received the same instructions as the children. The teachers recorded their Opinions in the teachers' lounge either during their breaks or while the experimenter read the stories to their classes. In the June testing, interviewing of all grades follow- ed the smme procedure as in September. The teachers received 35 the pupil-interaction scale in June, which they filled out in the teachers' lounge during their breaks or while the experimenter read the stories to their classes. The instruc- tions for the pupil-interaction scale were printed on the scale and space was allowed on the scale for the teachers to write their replies (See Appendix B). After the June interview, the children were asked to write the name of his or her best friend on a slip of paper and the experimenter collected the papers. As explained above in describing the sociogram, the children in all classes did not hesitate to follow this simple instruction. RESULTS The children's and teachers' responses to the punish- ment interview were classified in six categories (origin— ally recommended by Piaget, 1932, and used by Haviland, 1968) . 1) 2) 3) h) S) 6) These categories are: physical punishment: spanking, whipping, beating, fighting, slapping, hitting with objects, etc. restriction: loss of privileges, confined to room, corner, chair, closet, etc.; also includes lying or stealing with greater intensity than original culprit used, 1. e. punitive reciprocity. reference to authority: "tell on" culprit to mother, father, principal, teacher, big brother, etc. reciprocity: do the same thing -- lie or steal; or remove the problem e.g. child cheats, teacher removes paper and supplies another. loss of relationship: leave school, house, room, yard, etc. or not be friends. explanation: talking, explaining, yelling, shout- ing, etc. In order to categorize the replies, the responses from all students were written on cards with no reference to grade, sex, school, or teacher. All duplicate cards were 36 37 eliminated. Then the cards were placed in the six categor- ies shown above. The responses given above as examples are typical of the responses in that category. For a com- plete catalogue of the responses in each category, see Appendix C. Then the reference cards were used to place individual studentfis and teacher's responses in the assigned categories. The numbers one to six are not rankings. The results do not show evidence for a hierarchy within each division of punitive and restitutive judgments. Within the punitive judgment category, physical punishment, restriction, and reference to authority were used almost equally at each grade level. Reciprocity, loss of relationship, and expla- nation in the restitutive category are also used about equ- ally at each grade level. Although older children were more likely to suggest restitutive punishments and less likely to suggest punitive ones, the kind of punitive or restitutive punishment did not vary according to age. Kohlberg (1969) has suggested that there are at least six hierarchical divisions in his schema of moral judgment development. No one had tested the divisions used in pun- ishment judgment to see whether they were correspondingly related. The current results indicate that they are not. The two divisions -- punitive and restitutive -- seem to account for the develOpmental variation among first, third, 38 and fifth graders. In general, responses from all six stories fit quite well into the six categories. A few unresolved problems lay in the differences between punishments logically placed in the same category while referring to different stories. Fbr example, it is possible that suspension from school (loss of relationship) is quite different from staying out of a friend's yard (loss of relationship) mainly because suspension involves a referral to other authorities such as the principal and parents, who are likely to inflict other punishments, and consequently deprives the child of more than his relationship with his teacher. All the res- ponses in the "loss of relationship" category for the teach- er story (child lies to or steals from a teacher) were of this nature. There were no responses in this category for the parent story (child lies to or steals from his mother). A second problem lay in placing yelling, scolding, and other verbal reproaches in the explanation category. Among peers this verbal abuse may be reciprocal and not particularly punitive, but from a mother or teacher to a child it may fall occasionally into physical punishment -- that is, a loud abusive noise may be employed to frighten or humiliate the child, not to restore the relationship. On the other hand, some adults with little intention of punishing with words generally tend to explain things to 39 children in loud voices, believing that their words will "sink in" better if prOpelled more forcefully. Therefore some adults were given the benefit of the doubt in having their verbal corrections categorized as "explanation". The ambiguity here is exemplified in typical comments from children such as, "She would explain it to me why not". E: "How would she do that?" Child: "Well, she would yell..." Changes and variations in beliefs To determine whether types of responses changed from the first to the third to the fifth grades, combined pun- itive and restitutive categories were compared by grade using a multivariate chi-square (Sutcliffe, 1957). The results appear in Table l. The analysis reveals very significant differences among all grades (p <.OOl). A chi-square of differences between first and third, third and fifth, and first and fifth grades reveals that the difference is equally reliable between all grades (p (.001). A sex difference is apparent: girls were slightly more mature than boys, especially in grade one (p‘<.005). The grade by sex interaction was not signifi- cant («*<.lO). All children, regardless of school, teacher, or grade, gave more restitutive responses at the end of the year than ho Table l. Multivariate X2: Punishment judgment differences and interactions for grade, sex, story, and response change. Interaction df X2 Gr adea x Changec 6 70. 20M Sex x Change 3 18.37** Storyb x.Change 3 2h.56**e Grade x Change x Sex 6 11.06% Grade x Story x Change 6 18.86** Sex x Story x Change 3 16.kke* Sex x Story x Grade x Change 6 3.08 aGrades: bStories: cChange: 6%“ < .10 awn-«x < .005 WK ( e001 1: 3. 5 Teacher-child vs. peer-child relationship (1) (2) (3) (h) from punitive to restitutive punishment from restitutive to punitive punishment no change -- remained restitutive no change -- remained punitive A1 at the beginning (p< .005). The significant changes during the school year and between grades support the first hypo- thesis that, in general, changes in judgment change with age. Older children give more restitutive judgments and younger children give more punitive judgments. On this basis, restitutive responses are more mature than punitive responses. To determine whether there were variations in the res- ponses in different stories, the reaponses were grouped sev- eral ways. First the differences in responses to different stories were examined at each grade level. At all grade levels there were very significant differences between the responses given to stories on which the relationship was between an adult and a child and those in which the relation- ship was between two children with more restitutive responses when the relationship was between two children (¢‘<.001). At no grade level was there a difference between the responses given to stories in which the adult was a mother and those in which the adult was a teacher. This confirms, to some degree, that the difference is between adult and child re- lationships, not between family and authority and friends. In addition to the variation in responses because of the relation between the victim and the thief or lier, there were differences in responses to stories about lying and cheating. Stories about lying evoked more mature responses h2 at all grade levels than stories about stealing (d‘<.OOl). This also confirms the first hypothesis in showing that situational variations exert a strong influence on the ma- turity of response, even though there is a clear develOp- mental trend. Table 2 shows the per cent of response in each change category by grade. The difference between the responses to adult and peer relationship stories seems to be due to different factors at each grade level. In the first grade the difference occurs because 18% of the children change from restitutive punishment for peers to punitive punish- ment (most punitive responses fall in the "refer to auth- ority category"). Perhaps this reflects experiences during the school year in which the teacher intervened or encour- aged children to come for help. This change from resti- tutive to punitive responses is the greatest negative change at any grade level, to any story. It is negative in the sense that it contradicts the general trend to change from punitive to restitutive responses. But further analysis by individual classes will show that a short-term negative trend was observed under other circumstances. The differences in the third grade to stories involving adult and peer relationships seem to be due to the greater percentage of mature responses given to the Peer story as compared with the teacher story. This same difference is #3 Table 2. Percent of types of reaponse changes in each grade to teacher and peer punishment interview. % Response change Grades punitive to remained restitutive remained restitutive restitutive to punitive punitive Teacher Story First 25 12 6 S7 (N-96) Third 25 7 11 57 (N-118) Fifth 31 26 10 33 (N-th) Peer Story First 17 15 18 50 (N-96) Third 33 23 6 38 (N-118) Fifth 36 26 12 25 (N-1h5) ALL also true in the fifth grade but is not so pronounced. A second variation in the develOpmental change in judgment was predicted to be related to the beliefs of a class's teacher. As other studies had suggested (Bier- stadt, 1969; Measel, 1968; Guha, 1967), elementary school teachers tend to give homogeneous, relatively authoritarian responses to tests concerned with teaching methods. This study indicates that they also give homogeneous, relatively authoritarian responses to questions about punishment judg- ment. Of sixteen teachers, only one consistently chose a restitutive response, six chose the restitutive response half or more than half the time, the other nine chose the restitutive response less than half the time. For compar- ison, by the end of the fifth grade more than 65% of the children use restitutive responses half or more than half the time. Since comparable data on other adults are not available, there is no evidence to test whether teachers, especially these teachers, are typical adults in their beliefs about punishment: A post hoc hypothesis about teachers' beliefs was that the teachers would be influenced by the children they teach and thus offer less mature responses than expected. If this were the case, then one would expect that first grade teachers would have the least mature responses and fifth grade teachers the most mature. However, this 1L5 hypothesis was not supported. There was a significant difference between teachers' responses by grade, but the first grade teachers had the most mature reSponses, not the least (°‘<.05). There were significant differences between teachers who taught in schools D and H and those who taught in school A (“~<.05). The responses given by teachers in school A were more often restitutive than those given by teachers in schools D and H. Perhaps this difference is due to the fact that corporal punishment is allowed and practiced in schools D and H but not in school A. Perhaps the ubi- quity of corporal punishment makes it easier for these teachers to suggest and justifies it for them. Comparing the most severe punishments for each school -- corporal punishment for schools D and H, and suspension for school A -- there were still differences in the number of teachers in each school who suggested the most severe punishment as an effective response in situations involving children stealing or lying. This difference suggests that regardless of different practices in the two school systems, the teachers in schools D and H were more punitive. The data indicate that the differences in punitive beliefs between teachers in the different school systems affected the beliefs of the children. Table 3 shows the differences in responses to the teacher by grade for each h6 Table 3. Chi-squares of differences in responses to the teacher stories for school systems D and H vs A by grade. Grade (if x 2 First (fall) 11 19.07n°3' (spring) 11 27.16% Third (fall) 11 Sl.12** (spring) 11 50.87%* Fifth (fall) 11 29.26% (spring) 11 65.lh** *- °< < .005 aux-cc < .001 A? school system. For the first graders in each school system, there are no significant differences at the beginning of the school year. By the end of the school year there are dif- ferences in response to the teacher story only (v<<.005). Fbr the third and fifth graders there are significant dif- ferences in reaponses to both stories. The children in school A, where teachers are less punitive, tend to give more restitutive responses than their peers in schools H and D to both peer and teacher stories. An example of the effect of a teacher who gave con- sistently restitutive responses is shown in Table h. An analysis of the responses given by this first grade class shows that the re8ponses are not significantly different from those of the average fifth grade responses. The res- ponses to the peer interview story are most like those of the averaged third grade responses and they are not signi- ficantly different from the fifth grade responses. They are significantly different from the averaged first grade responses (“~<.05). This example further supports the hypothesis that, contrary to general develOpmental trends, situational variance can affect the responses of children significantly. Further analysis of the differences between teachers' 118 Table h. Comparison of a first grade taught by a teacher giving consistent restitutive responses with averaged fifth grade reaponses. averaged 1st grade 5th grade Response Change teacher peer teacher peer story story story story punitive to _ restitutive 29% 27% 30% 3u% remains restitutive 21% 21% 2u% 27% restitutive to punitive 18% 18% 10% 15% remains punitive 32% 3h% 26% 23% M9 responses supports the hypothesis that teacher differences are related to class differences. Table 5 shows that the response changes during the school year vary significantly between the class whose teacher gave the most restitutive reaponses and the class whose teacher gave the most puni- tive responses at each grade level. The difference at the first and third grade level for all responses (combined mother, teacher, and peer story reSponses) is significant (o<<.005). Most of this difference is contributed by the responses to the teacher story'(“<}OOl). The fifth grade teachers were either not sufficiently different or exerted less influence on their pupils' judgments. There is no significant difference in the fifth grade between classes with restitutive or punitive teachers. These results sup- port the second hypothesis by showing that for first and third graders the children's beliefs are more consonant with their teachers' beliefs at the end of the year than at the beginning of the year. First and third graders with punitive teachers were more likely to give punitive responses than were first and third graders with teachers who advocated restitutive responses. To test for specific teacher effects, that is, the teacher's effect on specific categories of responses (as Opposed to changes from punitive to restitutive responses), the teacher's response to a particular story was compared 50 Table 5. Comparison of response changes between classes of most and least mature teachers at each grade level. Grade Response Change lst grade 3rd grade 5th grade most least most least most least punitive to restitutive 29% 18% 26% 17% 82% 31% remains ' restitutive 21% 0% 21% 1% 21% 17% restitutive to punitive 18% 5% 1n% 9% 8% 12% remains Punitive 32% 77% 38% 78% 29% h0% x 2 12a21.uu x 2325.66 12:2.00 0‘ 0‘ ( e001 °‘ ( e001 11.8. Note. -- There is no significant difference on peer stories. 51 with her class's response. If the percentage of responses matching the teacher's category was higher than the average percentage of responses in that category for that grade, then that class was rated plus, showing a positive teacher effect. If the class had fewer than the average number of responses in the category chosen by the teacher, then that class was rated.minus, for no or negative teacher effect. For the first grade 60% of the class responses were rated.minus. For the third grade hh% Of the class responses were rated minus. For the fifth grade 33% of the class responses were rated minus. These results indi- cate that the teacher's particular mode of punishment (use Of authority, suspension, whipping, etc.) was not indica- tive of the:mode of punishment her students use, even though the maturity of her response affected the maturity of her students' responses. Therefore the second hypothesis was not supported entirely, since the children's beliefs did not become more consonant with.their teacher's beliefs in terms of specific categories of responses. The second hypothesis prOposed that specific teacher effects on the content of the replies would be more pro- nounced for first graders than for fifth graders. The results indicate that this is incorrect. If children tend to adOpt the teacher's responses, then such tendency is more likely in the fifth grade (77% positive teacher effects) 52 than in the first or third grade as indicated above. This (non-significant) increase in agreement from first to fifth grade may have been due to the fifth graders' increased ability to verbalize adult punishment beliefs, making it more likely that they would state a particular adult's belief. The first graders were apparently not able to state adult beliefs as readily as the fifth graders, although they were influenced by them significantly; therefore, they did not adOpt a particular adult belief. Teachers' beliefs and repgrted teachipg practices To determine whether teachers with restitutive punish- ment beliefs were more likely than were teachers with pun- itive beliefs to ask their classes for a discussion of pun- ishment issues, the rankings of the teachers on the pupil- interaction scale were compared with their rankings on the punishment judgment interview. Goodman and Kruskal's gamma for ordered data with ties predicts that the probability of finding a pair with the same ordering is .hS more than that of finding a pair with different orderings for this data. Therefore, teach- ers with.more:mature punishment judgments tended to inter- act verbally more with their students in situations where moral or value judgments were to be made (evaluation of a robber, a child cheating, and a child looking into the 53 teacher's desk) than were teachers with less mature judg- ments. This difference suggests that teachers with mature punishment beliefs tend to encourage their students to discuss their beliefs rather than to lecture them on cor- rect beliefs. The Opening phrase for interacting teachers was generally, "I would ask...", and for non-interacting teachers, "I would tell...". Peer pgpularity and punishment beliefs The fourth hypothesis was based on a claim by Kohl- berg (1969) that pOpularity is related to moral judgment. To test this, the scores of pOpular children (three or more "best friend" nominations) on the punishment judgment interview were compared with the scores of other boys or girls in their class. There was no significant difference in the categories of responses or in the maturity of res- ponses of pOpular boys or girls in any grade. Perhaps this is another difference between "moral judgment" and punish- ment judgment. Because there were no discernible differences between pOpular children and their peers, it was not possible to test for peer effects by noting changes in the beliefs of children associated with "pOpular" children. Therefore a second test of peer effects was attempted. Assuming that conformity to peer Opinion would be reflected in conformity 514 of response at the end of the school year within a class, the most often chosen categories for each class were exam- ined at the beginning and at the end of the school year. However, the spread of responses in the various categories remained about the same over a year's time. That is, if "loss of relationship" was most often chosen at the begin- ning of the school year it was unlikely that it would be even more often chosen at the end of the school year. Therefore, no peer effects were found in this study. CONCLUSIONS This study about children's judgments of punishments gives some indication of how changes in these judgments occur and why differences in judgments among children of the same age are common. Changes and variations in beliefs First, the hypothesis that judgments change from puni- tive to restitutive as children grow older was supported. From.the earliest study by Barnes (1902) to the current day, this hypothesis has been supported. But in all studies, although a significant change occurred, there were always some children in the youngest groups who seemed."mature", and some in the oldest groups who seemed "immature". The develOpmental change in punishment judgment was not so ex- tensive as in supposedly related areas of cognitive judgment such as conservation. Why should there exist a particular range of’mental ages in which.changes in conservation be- liefs occur, before which ages practically no child has mature beliefs, while there apparently is no such demar- cation in beliefs about "morality", or in this case, about punishment? One possibility was-that there was no cognitive aspect to this change at all, that the belief depended.upon simple SS 56 reinforcement. A two-year old taught restitutive beliefs should be equally as convinced of their effectiveness as a ten-year old. In this case the beliefs would be related only to the training given to the child, and a change to a change in treatment. Perhaps adults believe that one type of punishment is more effective for young children and another kind for older children, and because of this adult belief, a develOpmental change occurs in the child- ren's beliefs. This possibility was not supported by the data of this study about schools and teachers. The teachers of fifth graders were more, not less, convinced than first grade teachers that punitive punish- ments were effective. Nevertheless, fifth grade children were more mature than first grade children --I- almost more mature than their teabhers. On the other hand, a first grade class, whose teacher consistently prOposed restitu- tive punishments, gave responses not significantly differ- ent from fifth grade classes. These results taken together would seem to indicate that reinforcement can make a signi- ficant difference in children's responses, but that most adults do not consistently reinforce mature responses. Nevertheless first and fifth graders are different. Perhaps their parents are treating them differently. This aspect of the training has not been investigated. Or per- haps there are really developmental differences which are 577 not caused solely by direct environmental changes or by changes in adults' behavior. If develOpmental differences occurred independently of environmental changes, changes in the "mature" first grade class must be explained, the differences that occur- red between classes taught by more and less mature teach- ers must be explained, and the different kinds of responses given to interview stories that vary in terms of their interactions or the type of transgression.must be explained. An analogy from cross-cultural studies of conserva- tion.may be useful in illustrating how such discrepancies may occur in develOpmental phenomena. Studies of conser- vation among the Wolof, for example, have shown that in primitive cultures the rate of acquisition of conservation varies much.more than it does in technological cultures (Bruner, 1966). In primitive cultures some few children have acquired conservation concepts at about the same age as in technological cultures, but there are some adults in primitive societies who have never acquired these con- cepts. What would happen if one trained prhmitive children in conservation problems? Would it be more or less easy to train them than to train immature (in terms of conser- vation) children in technological cultures? Unfortunately, we do not know. In a prhmitive society everyday events do not convince develOping or mature persons that conservation 58 is a "true" view of reality. Therefore, perhaps, many pOOple who are capable (capable in that they would espouse conservation if reared in a technological culture) of com- prehending conservation do not. Perhaps in our society everyday events do not always convince develOping or mature persons that restitutive pun- ishments are more effective than are punitive punishments in dealing with crime and criminals. (In fact, experimental evidence has not managed to prove one more effective than the other in all circumstances.) This analogy, of course, leads one to speculate about a society in which all mature persons would deem restitutive punishment as more effective. In such a society, would young children still feel that punitive punishments were more effective? It would not be possible to examine this question in its most general form, but it could be studied under particular laboratory conditions: A provocative example was the first grade class whose teacher apparently advocated consistent resti- tutive punishment, nevertheless not all the children gave a restitutive response at the end of the year. The class as a whole was indistinguishable from a class of older children, however, in beliefs about how teachers should punish, and significantly more mature than other first grade classes. This indicates that first graders can verbalize the same beliefs as older children. Perhaps even pro-school 59 children could learn to use restitutive responses. There were other indications, however, that restitu- tive punishment beliefs, in general, mature almost inde- pendently of direct training. No class was less mature at the end of the year than at the beginning, although many teachers were very inconsistent in choosing restitu- tive punishments and several did not choose them for teach- er-child situations at all. This seemed to indicate that punitive and restitutive punishment beliefs form an invar- iant sequence, not for individuals in a short period of thme, but for groups over a long period of time. The kinds of variations in beliefs found in the cur- rent study can be summarized in terms of Kohlberg's "rules" for defining a cognitive stage. The first rule requires that each stage implies distinct or qualitative differences in solving the same problem. The data on punishment be- 1iefs show that there are two distinct types of responses among first, third and fifth graders. Although the res- ponses may be divided into any number of categories (this study used six) the only apparent develOpmental changes occur between the divisions called "punitive" punishment and "restitutive" punishment. Punitive and restitutive responses are distinctly and qualitatively different. Therefore the first "rule" for defining a cognitive stage is satisfied. 6O Kohlberg's second rule requires that the stages form an invariant sequence in individual develOpment. This requirement was not fulfilled by the results of this study. The stages were relatively invariant for groups of children, but not for individuals. The variation from the beginning of the school year to the end of the school year showed that many children who chose restitutive punishments at the beginning of the year chose punitive ones at the end of the year. The results from the punishment interview with the teachers, using them as examples of "mature" adults, do not indicate a clear preference for restitutive judgments either, so there may not be a clearly defined change from one kind of belief to another. There is, however, a clear tendency for older children to make the restitutive judg- ment more often than younger children. Therefore, an em- pirical test of this second rule has not justified its use in the strong form that Kohlberg prOposes. Perhaps an amendment stating that the sequence should be generally invariable as shown by group data would be sufficient. The third rule requires that each belief "represents an underlying thought-organization". As Piaget has sug- gested in his analysis of moral judgments, punitive pun- ishments are part of a world view in which rules are in- variant and rulers (adults) are infallible. From this point of view there is no restitution, actions may not be reversed 61 and peOple do not reciprocate, they rule or obey. Resti- tutive punishments, however, are part of a cooperative world, in which many rules are relative, actions nearly reversible and peOple reciprocally equal. The last rule requires that the more mature stage be logically more integrated and.more differentiated. If one considers punitive punishments in reference to a dichotomous world of authoritatively decreed rights and wrongs, such a world does seem less complex than a world in which dif- ferent coOperating groups of peeple may decide upon rights and wrongs and.may even change the rules if it seems appro- priate to them. If one advances from a concept of obed- ience to a concept of mutual responsibility, the possibil- ities for variations in behavior and thought should increase. Therefore, of the two possibilities (1) that punish- ment beliefs are purely situational (may be easily trained by direct reinforcement) or (2) that punishment beliefs are aspects of cognitive stages, the latter is more sup- ported by the data of this study. Almost in spite of teacher training, most students changed from punitive to restitutive beliefs. Those students whose teachers were more mature showed significantly more mature responses than those children with less mature teachers. However, even classes with ex- ceedingly immature teachers were not less mature at the end of the school year than they were at the beginning. This 62 shows that there is an invariant sequence in beliefs, that it is not as easy to train children of school age to accept punitive beliefs as it is to train them to accept resti- tutive beliefs and that this change may occur in Spite of a teacher's influence. There are differences in the maturity of the responses to the different stories. The responses given to stories involving two children were more mature for all children than the responses to stories involving adults, either mothers or teachers. This difference may be related to Piaget's hypothesis that peer cooperation (as illustrated by Piaget (1932) in games) is one eXperience that contri- butes to the acquisition of mature concepts. Presumably the equality of the actors and the relativity inherent in such situations leads to concepts of "morality beyond duty". This result leads one to propose other methods for studying the develOpment of punishment concepts. One would hypothesize that a training situation in which trainee and trainer were of equal status would engender the most resti- tutive corrections; on the other hand, the more different the status of trainee and trainer, the more punitive the corrections would be. One must then question the effective- ness of each training situation, of course, and try it under various conditions. There were also differences in the maturity of responses 63 to the different transgressions. Lying was seen as a lesser offense than stealing, apparently, and was accorded lesser punishment in two ways. The punitive punishments were less severe, and it was more likely that a restitutive response would be given. This sort of result is not really surpris- ing, since one sees evidence of this type of thinking per- vading our whole social system. Most peOple seem more will- ing to rehabilitate thieves and more willing to punish murderers severely than vice versa, although the principals of life and property are almost equal before the law. Although this difference in the severity of the crime and punishment responses seems obvious, previous studies have not noted its effect on responses. All moral judgment studies (Piaget, MacRae, Johnson, Kohlberg, et al.) have used several interview stories which involve various trans- gressions and various relationships between the victim and the criminal. Because the stories have not differed from each other in systematic ways, these effects have not been detected, though they certainly exist. For example, several of Kohlberg's interview stories (moral dilemmas) involve a choice between the life or happiness of a beloved spouse and a crime (stealing, murder). Kohlberg has not reported differences based on the type of crime or on the relation- ship between the person in the "dilemma" and his choices. Kohlberg claims that such differences are irrelevant to the 61+ reasoning behind the "justice" of the decision in the dil- emma. The results of the current study strongly suggest that both the decision and the reasoning may be influenced by the details of the "dilemma". Teacher's beliefs and reported teaching practices The results show that there are differences among teachers' responses to the punishment interview. These differences are related to similar differences on the scale of teacher-pupil interaction. Most importantly, these differences are related to changes in their students' res- ponses. The close relationship between rankings on the punishment interview and the teacher-pupil interaction scale indicates that teachers who responded maturely to the punishment interview tended to ask their students to offer Opinions about the morality and reSponses to steal- ing and cheating rather than simply to offer their own opinions. The latter was the typical response of the teacher who gave punitive responses. The change in the students of teachers who gave resti- tutive responses may be due partly to direct training, but it is more likely to be due to the teacher's having commun- icated to the children that their Opinions and here were equal. It also is possible that the major factor insti- gating change is the opportunity granted by the teacher 65 to consider moral problems and the ensuing confusion as many different Opinions were presented. One of the exper- iences that is hypothesized to instigate change in many types of beliefs is an experience of incongruency, the dis- covery of a situation in which an old solution does not fit. For the children of mature teachers there may have been situations which did not fit a perception of a world ordered into superiors and inferiors in which the good is to obey; these teachers expressed the view that the good is more than obedience and illustrated the variations in obedience. Three further aspects of the teachers' responses were 1) the general immaturity of the responses 2) the differ- ences between the teachers of different grades and 3) the differences between the teachers in different schools. Further study of adults' responses is needed to show whether these particular teachers' responses are typical or, for that matter, whether teachers are typical of adults in general. If teachers' responses are, in fact, less mature than those of a non-teaching adult, one would then need to investigate how this occurred, whether it involves a retrogression dependent upon the number of teaching years or whether the difference is already apparent in college students who intend to teach. One would also wonder whether or not teachers who are immature in this area are effective 66 teachers. The results indicate that first-grade teachers are more mature than teachers of third and fifth graders. Since the sample of teachers is relatively small, not too much should be made of this difference. If the difference is supported by further evidence, one might hypothesize that fifth grade teachers more than first grade teachers need to be more adamant about exerting authority, since fifth graders are beginning to doubt the veracity of auth- ority. The differences among teachers at different schools seem quite reasonable. One principal remarked that one should test the principal instead of his teachers, since he would quite naturally hire teachers compatible with him and his philosOphies. This alone would account for differences, even though other influences no doubt Operate. In any case, these results indicate the necessity for in- cluding different schools in any study Of teachers. Peer pOpularity and punishment beliefs The only peer effects found were the indirect ones showing more mature responses to the interview story in- volving peers rather than an adult and a child. Contrary to expectations, pOpular ("best friend") children were not more mature than their peers. This result is contrary 67 to a claim by Kohlberg (1969) that moral maturity is re- lated to pOpularity. Because there was no difference, changes due to the influence Of a pOpular friend could not be studied. In retrospect, one might predict that very pOpular children enjoy a status slightly more like that of an adult than a peer. Perhaps some children are pOpular because they are "fair" in sports and other activities and know both how to lead and how to cooperate. Some children, however, may be pOpular because they play an adult role well. If the latter is the case, it may explain the results insofar as the pOpular children would be expected to assume authori- tative roles with respect to correcting the mistakes of others and to advocate punitive punishments. On the other hand, another possible reason for this discrepancy is suggested by Bronfenbrenner (1970). In discussing differences between children in the U. S. and the U. S. S. R., he comments that peer groups have a bene- ficial influence in training children in the U. S. S. R., but not in the U. S. The peer group is used as a training influence by Soviet educators, while in the U. S. the peer group influence Often contradicts the major aims of edu- cators (p. 156). If it could be shown that educators, both parents and teachers, preferred restitutive punish- ment, then active children's leaders might prefer punitive 68 punishments. This possibility is not statistically sup— ported by the study, but a direct test of peer effects should account for it. Therefore, in future research, this method of assess- ing popularity should be abandoned, and other methods which ensure the equivalent status of individuals should be em- ployed to measure peer effects. If the study involves hypotheses about the interactions among peers, the frequency and type of interactions must be examined rather than assumed. REFERENCES REFERENCES Bandura, A., & MacDonald, F. The influences of social reinforcement and the behavior of models in shaping children's moral judgment. Journal of Abnormal and Social Psychology, 1963, 61, 27h-281. Barnes, E. Growth of social judgment. Studies in Education, 1902, 2, 203-217. Berenda, R. W. The influence of the group on the Jgdgments of children. New York: King's Crown, 5 . Bjerstedt, A. A simulator-test approach to the study Of teacher's interaction tendencies. Educational and Psychological Interactions, 1968, 22, 21 p. Blatt, M. The effects of classroom discussion upon children's level of moral judgment. Unpublished doctoral dissertation, University of Chicago, 1969. Cited by L. Kohlberg, Stage and Sequence: The cog- nitive and develOpmental approach to socialization. In D. Goslin (Ed. ), Handbook of socialization. new York: Rand McNally, 1969. Bronfenbrenner, U. The role of sex, class and culture in studies of moral develOpment. Religious Education, 1962’ 21! 3'17 0 Bronfenbrenner, U. Tonworlds of childhood; U.S. and U.S.S.R. New'York: Russell Sage Foundation, 1970. Bruner, S. J., Olver, R., Greenfield, P. M. Studies in Cognitive Growth. New York: Wiley, 1966. Church, R. M. The varied effects of punishment on behavior. Psychological Review, 1963, 19, 369-h02. Cowan, P. A., Conger, J., Heavenrich, J., & Nathanson, M. Social learning and Piaget's cognitive theory of moral develOpment. Journal of Personality and Social Psychology, 1969‘, _1_}_,"261-2"71I. 69 7O Dewey, J. Human nature and conduct. new York: Random House, 1957. Festinger, L. A theory of cognitive dissonance. T‘— Evanston, 111.: Row, Peterson, 1957. Graham, A. w. DO teachers who use democratic methods develOp democratic attitudes. E1ementary,School Journal, 19h6, September, 22-27. Hamm, N. G. & Hoving, K. L. Conformity of children in an ambiguous perceptual situation. Child Devel- Harrower, M. E. Social status and moral develOpment. British Journal of Educational Psychology, 193k, E: 73.93- Hartshorne, H. & May, M. A. Studies in the nature of character: Vol. I,_Studies In deceit° Vol. II Studies in self:control° VOIZ III Studies in the organization of character. New York: Macmillan, 92 -1 O. Haviland, J. M. Judgments of the justice of punish- ment by adolescent delinquents. Unpublished Master's thesis, Michigan State University, 1968. Hoffman, M. L. Power assertion by the parent and its impac: on the child. Child Development, 1960, ii, 129’]. 30 Hoffman, M. L. Childrearing practices and moral devel- Opment: Generalizations from empirical research. Child Development, 1963, 24, 573-588. Hoffman, M. L. & Salzstein, H. Parent discipline and the child's moral develOpment. Journal of Personality and Social Psychology, 1967, 5, E5-§7. Holstein, 0. Parental determinants of the develOpment Of moral judgment. Unpublished doctoral disserta- tion, University of California, Berkeley, 1968. JOhnson, R. A study of children's moral judgments. Child Develgpment, 1962, Q2, 327-35h. 71 Kohlberg, L. DeveIOpment of moral character. In M. L. Hoffman & L. W. Hoffman (Eds.), Review of child develOpment research. New York: Russell Sage Foundation, 196k. Kohlberg, L. Moral education in the schools: A deve- lOpmental view. School Review, 1966, IR, 1-30. Kbhlberg, L. Moral and religious education and the public schools: A developmental view. In T. Sizer (Ed.), Religion and Eublic_gducation. Boston: Houghton-Miffl n, . Kohlberg, L. Stage and sequence: The cognitive and develOpmental approach to socialization. In D. Goslin (Ed.), Handbook of socialization. New York: Rand McNally, I939. The concomitant develOpment of cognitive A test of selected de- Lee, L. C. and moral modes of thought: ductions from.Piaget's theory. Unpublished doctoral dissertation, Ohio State University, 1968. Cited in Dissertation Abstracts, 1968, g2, 1856-1857. LeFurgy, W. G. & Woloshin, G. W. Modification of children's moral judgments through experimentally induced social influence. Proceedin s of the 76th Annual Convention of the APA, 19 ,‘2, 3 7-3h8. MacRae, R. A test of Piaget's theories of moral devel- Opment. Journal of Abnormal and Social Psycholo , 1951+, 9,—Iu-180 The relationship between teacher influence M93391, We We and levels of thinking of second grade teachers and Unpublished doctoral dissertation, University pupils. Cited in Dissertation Abstracts, of Michigan, 1968. 1968. g. 1955. Minuchin, P., Biber, B., Shapiro, E., & Zimiles, H. The cholo ical impact of the school experience. fiew‘YorE: Basic Books, Inc., 1969. Mohandy, G. S. & Guha, T. N. A comparative study of authoritarianism among service officers and higher secondary school teachers. Ps cholo ical Researches, 1967, -2_, 9-12. 72 Parke, R. D. & Walters, R. H. Some factors influencing the efficacy of punishment training for inducing reaponse inhibition. Mono ra h of the Society_for Research in Child DeveIOpment, 1967.l2§, léfls. Peck, R. F. & Havighurst, R. J. The s cholo of character develOpment. New'YorE: Wiley, 1960. Piaget, J. The moral udgment of the child. Glencoe, 111.; Free Press, 1 82 (Originally published,‘ 1932 . Rosenthal, R. Experimenter effects in behavioral re- search. New York: Appleton-Century-Crofts, 1966. Rousseau, J. J. Emile. New York: L. P. Dutton & Co., 1911. Sears, R. R., Rau, L., & Alpert, R. Identification and child-rearing. Stanford, CalifTYE Stanford University Press, 1965. Sullivan, H. S. The interpersonal theo ofgpsychiatry. New'York: W. W. Norton & Co., Inc., 19 . Sutcliffe, J. P. A general method of analysis of frequency data for'multiple classification designs. Psychological Bulletin, 1957.72g. 13h—137. Turiel, E. An experimental test of sequentiality of develOpmental stages in the child's moral judgments. Journal of Personality and Social Psychology, 1966, 1, 611-618. Turiel, E. Developmental processes in the child's moral thinking. In P. Mussen, J. Langen, & M. Cov- ington (Eds.), New directions in develo ental 3 ch- ology. New York: Holt, Rinehart & Winston, 1959. Winer, B. J. Statistical_principles in ex erimental desigg. New York: MEGrawbHill BOOEECO., 1952. APPENDICES APPENDIX A Puni shment Judgment Intervi ew *1. *3. APPENDIX A Punishment Judgment Interview The stores downtown are always worried about shep- lifting. In the past few years the stores have been good to students by giving them part-time work, especially at Christmas, to provide those who don't have much money with some buying power. The police have been working on the shOplifting problem, too, arresting many more peOple. ShOplifting has de- creased recently. What do you think is reSponsible for most of the decrease? Why? A boy was alone in study hall and decided to look in the teacher's desk. There was some money in the tOp drawer and he took it. The teacher found out. What do you think she would do? Another boy did the same thing. He took some money from a teacher's desk. But the teacher talked to him and explained that he was taking something that was important to someone else. Later these two boys found a teacher's wallet in the school parking lot. One of them wanted to take the wallet to the teacher. Was it the first boy or the second boy? Why? Two boys got caught stealing a car. Their homes, their school records, etc., were exactly alike, and neither had led the other into taking the car, but for some reason one was sent to a Training School and the other was put on probation and was allowed to stay at home. The one who stayed at home had a good probation officer who worked with him a lot. The one who was sent to the Training School was given lots of discipline till he was released after 73 *Distractor stories. Responses to these stories were not evaluated. 71L about a year and came home. Now they were both in their neighborhood again. About six months later one of them stole another car. Which one do you think it was? Why? Some girls borrowed a car one night from the Older brother of one of them. They went for a ride in the country and, since they all had drivers' licenses, they took turns driving. They all drove carefully but one was driving a little fast and skidded on the ice. The car went into the ditch and got a banged up grille and right fender. Now was the one driving responsible for the accident or should they all be held responsible? Tom knew that his Mother kept money in the bedroom. One day he took some, but his Mother discovered it. What would his Mother do? Sam stole some money from.his Mother, too, but she did not punish him. She talked to him.so that he would understand what stealing was. A few days later Tom and Sam found somethings that their Mothers had left on the front step. One of the boys wanted to sell what they had found and keep the money. The other boy wanted to give them back. Which boy wanted to give the things back to his Mother? Why? A teacher caught a student cheating on a test. The student had all the test answers on a sheet of paper underneath the test. What should the teacher do to the student? Another teacher made a practice of never punishing students for cheating. This teacher would make a point of talking to students who were known cheaters, and explaining the meaning of and effects of cheating. Which teacher had the most students cheating, probably? *7 . 75 A boy was out with a group of friends one evening. They got to wrestling and one boy began wrestling ‘ each of the others, even when they didn't want to. He got to wrestling with a boy who was wearing glasses and the glasses fell off and broke. Now there were all sorts of things that the grownups could do to him: they could give him a spanking or they could have him work to pay for replacing the glasses or they could break something of his. What would be the worst punishment? What would be the fairest punishment? Which one do you think that they gave him? Peter was over at his friend's place one afternoon waiting for him to come back. He found some money in the room and took it. His friend found out and was really mad. What do you think his friend did? Another time John was at a friend's place and he took some money from his friend. His friend found out and explained to John what he was going to do with the money and what he thought about stealing. Peter and John found something their friend had left at school a few days later. One of the boys wanted to take it back to the friend. Who wanted to take it back? Why? A girl was supposed to have her room cleaned up every day before she left the house. However, she made a practice of stuffing clothes and things under the bed instead of cleaning. Her Mother found this out and decided to do something about it. That should she do? This girl had confided her "cleaning" method to a friend who also began to stuff things under the bed instead of cleaning up. When her Mother dis- covered it she had a talk with her daughter and explained to her why she wanted the room clean. Vhich girl will keep her room clean the longest? 76 If each of these girls had a younger sister who knew the above incident, which girl's sister would be most likely to keep her room clean? *10. A science class was on a trip in the spring and they stOpped to look over an apple orchard that was in bloom. While the leader wasn't around, one person in the group out off many limbs, to get apple blossoms. He did this when none of the others were looking, so that none of the group knew who did it. The farmer saw the cut branbhes in the cars and was very angry. He complained to the group leader but no one knew who did it so no one could tell the group leader who had done it. The farmer and the group leader decided to charge everyone in the group equally to pay the damages, since they couldn't find out who had done it. Was this right? 11. A boy told another kid in his class that the Friday quiz had been canceled. Because of this the kid failed the quiz. Vhat do you think the kid should do? The same thing happened to another boy. Another student in his class told him a quiz had been canceled and the boy failed the quiz. This boy went over to the lier after school and talked to him about it. Which boy would play the same trick again, probably, the first one or the one who was talked to? Why? *12. This group of boys went into a candy store every day after school. Then they started out, every- thing was all right, but soon they began to take candy and things. One day the owner caught them. They admitted that they'd been taking things for quite a while. The owner didn't know whether to call the juvenile squad of the police or to give 77 the boys a spanking or to have the boys do things like sweeping and mOpping the store till they'd worked to pay for what they had taken. Which punishment would be the fairest? What punishment would be the hardest? what did the storekeeper do? AP PENUIX B Pupil Interaction Scale APPENDIX B Pupil Interaction Scale Name School_fi Grade Instructions A series of situations will be found below. You are to pretend or imagine that you are the teacher of the child described. The child is to be considered the sane age as those you are now teaching. Your task is to write down exactly how you would respond to the child in each of the situations, in a word, sentence or short paragraph. write down your exact words and actions, but please do not explain why you said or did what you described. Again, write down your exact words or actions as if you were writing a script for a play or movie (e.g. do not write "I would reassure or comfort him", instead for example write "I would smile at him.and say "Don't worry, Billy, I'll explain it to your mother".) 1. In a discussion of current events, Jimmy describes a robbery with Obvious admiration for the robber and little regard for his victim. 2. Lee was left alone in the classroom for some reason. When the teacher and the rest of the students return they see him.standing at the teacher's desk, filling his pockets with small items from her drawer. 3. When.the current researcher (myself) leaves the classroom, one of the children asks you what you would answer to the story about the boy who cheats on a test. The rest of the class also seems very interested in your reply. 78 APPENDIX 0 Punishment Judgment Categories APPENDIX D Punishment Judgment Categories Explanation: "tell him not to do it" "speak to him" "she will explain to him" "talk to him about..." "tell him he won't have any friends" "tell him it hurts peOple's feelings" "forgive him" "make him apologize" "yell at him" "remind him.when he lies" "bawl him out" "scold him" Loss of relationship: "stOp being friends" "not play with hLm" "ignore him" "get mad at him" "never talk to him again" "8° home" "don't let her visit for a week" "make him go out" "suspend him" "put dunce cap on him" "throw him out" "get out of her room" "don't like him" "never trust him" "don't believe him anymore" "look mad at him" 79 80 Reciprocity: "tell him same thing" "do it over" "make him pay it back" "throw her paper away" "do something like that to him" "tell a lie to him" "make her clean the house" "take some of his allowance" "make her do some work" "make her clean her room" "throw away her paper" "rip up her paper" "take test over" "move her desk" "stuff her mess under the bed" "make him work" "make him give it back" "give her a zero" "take away a toy and give it back to the store" "erase the paper" "make student study more" "hide your money" "do the same thing" "make him earn money" "put a mouse trap under her bed" "unstitch his baseball" "give her another test" "let her room get messier and.messier until she cleans it up" Refer to authority: "tell his ma" "send him to office" "call the principal" "call the police" "tell teacher" "tell daddy" "write about it on report card" "have big brother settle it" 81 Physical restriction: "put him in corner" "make him write 100 times" "stay after school" "stay in house" "ground him" "restrict in yard" "take TV away" "go to bed at 6:00" "stay in from recess" "make him do a lot of (home) work" "no dessert" "stand on tip-toe in closet" "put him in back of room" "move his desk" "take away privileges" "put him in hall" "put him on probation" "do dishes for a week" "take his comic books away" "lock her in her room" "put his head down" "let her sleep in the yard" "tie him to a chair" "throw her in the garage" "take him back to the kindergarten (lst grader)" Physical punishment: "spank him" "beat his hands" "Whip" "thunk him" "whomp" "pound him" "tan his hide" "smash his mouth" "sock him up" "give him a black eye" "choke him" "use the belt" 82 Physical punishment: (continued) "get in a fight" "throw an acorn in his pants" "kick in his teeth" "kick his rear" "belt him" "give him a black eye" "paddle him" "shake him up" "kill him" "hit him in the eye" "hit him with a baseball bat" "spank him with the Board of Education" "slap his hands" "kick him" "trip him" "wash her mouth with soap" "beat his brains in" "tie an anchor around him and throw him in the middle of the ocean" "pick a fight" "rough him up" ”‘A@1me