AN WHO?! OF THE REATIONSHTPS AMONG AWESSNE BEHAVIOR, READTNG, MD 00mm IN DELMQUENT BOYS WORN EDUCATION mzcmm STATE unwmw DENNIS LARRY HGGENSO-N I968 This is to certify that the thesis entitled AN INVESTIGATION CF THE RELATIONSHIPS AMONG AGGRESSIVE BEHAVIOR, READING, AND DmMATISM IN DELINQUENT BOYS presented by Dennis Larry Hogenson has been accepted towards fulfillment of the requirements for Ed.D Special Education degree in J'- 4 I 2 s d d t " LIBRARY ‘ Michigan 97"" l Ungw‘f’ii- “s ~‘. -.. U' ’ Major professor Datef/CJ 2C”//.3:. /([‘fj 0-169 ABSTRACT AN INVESTIGATION OF THE RELATIONSHIP AMONG AGGRESSIVE BEHAVIOR, READING, AND DOGMATISM IN DELINQUENT BOYS by Dennis Larry Hogenson Michigan State University Purpose of the investigation. The relation- ship of school linked frustration to aggressive behavior is not well understood at present. Similarly, little is known about the relationship of delinquent aggressive behavior to dogmatism, intelligence, and school attitudes. A frustration-aggression hypothesis such as that formulated by Dollard and Miller, and as amended by later workers, might predict that boys confined in state training schools who had failed to achieve in reading would exhibit more aggressive patterns of delinquent behavior than confined boys who had succeeded in reading. The present investigation sought to determine correlational relationships among aggressive behavior, reading, Rokeache's construct dogmatism, school attitudes, and intelligence in two papulatibns of delinquent boys. Demographic factors inherent in the subjectsihome environments were also related to aggressive behavior. ugtho . Two groups of 48 subjects each were randomly selected from the pepulations of the Lansing Boys Training School (Michigan) and the Red Wing Boys Training School (Minnesota). Subjects were tested using the following instruments: Wechsler intelligence scales, the reading section of the inter- mediate level Stggford flievwnt Test, the Dogmtism §£§l§: and the Minngsgta Student Atttttde anegtgtz. Three measures of aggressive behavior utilizing data from.court reports, training school staff reports, and self reports of aggressive behavior, were employed. Data from.caseworker reports provided the demographic information for analysis. Ftndings. 1. While significant agreement was not observed between the three measures of aggression, the court report measure was found to correlate beyond .05 with reading in both groups. 2. Dogmatism was not observed to correlate in either group with any of the three measures of aggression. 3. Aggressive behavior was observed to correlate beyond the .025 level (staff report measure) with school attitudes in the Red Wing group, and beyond the .05 level (self report measure) in the Lansing subjects. 4. While aggression did not correlate significantly with IQ in the Red Wing subjects, this variable was observed to correlate beyond .05 for two measures of aggression in the Lansing sample. 5. Reading was found to correlate significant- ly with both aggression and IQ in both groups. Read- ing also correlated significantly with school atti- tudes in the Red Wing subjects. 6. Dogmatism.was not significantly correlated with either school attitudes or intelligence in the Red Wing subjects. It was correlated beyond .05 with school attitudes in the Lansing group. 7. School attitudes were correlated with IQ at a highly significant .001 level in the Lansing subjects, but failed to reach significance in the Red Wing Sample. 8. In the chi-square analysis of demo- graphic factors, no relationship between aggres- sive behavior and the demographic factors was observed. AN INVESTIGATION or ms RELATIONSHIPS AMONG AGGRESSIVE BEHAVIOR, READING, AND DOGMATISM IN DELINQUENT BOYS BY Dennis Larry Hogenson A DISSERTATION Submitted to the Faculty of the Graduate School Michigan State University In Partial FU1fillment of the Requirements for the Degree DOCTOR.OF EDUCATION June, 1968 ACKNOWLEDGEMENTS In the course of a research undertaking which has extended over more than two years, it is likely that the cooperation and encouragement of a number of people will have been necessary for a successful conclusion. Such has been the case in the present study. The writer wishes to extend particular gratitude to Dr. John L. Johnson and Dr. Donald Burke. It was through the imagination and trust of Dr. Johnson that the investigation was originally conceived and encouraged. It was through the counsel, toil, and considerable patience of Dr. Burke that the study was completed. The writer is also grateful to Dr.'s William.Durr, Charles Henley, and Arthur Seagull for agreeing to serve on the writers committee, for reading the manuscript, and for participating in itfs oral defense. Thankful appreciation must be extended to Dr. Paul Spata, Superintendent of the Lansing Boys Training School, and to Mr. Joseph K. Morgan, -11- chief psychologist at Red Wing. Both men extended their personal cooperation, and opened their schools to the writer in a most sterling fashion. Lastly, the writer would like to thank his patient and never-tiring secretary, Mrs. Margaret Bassett, for her help in the preparation of the manuscript. Minneapolis June, 1968 -iii- TABLE OF CONTENTS CHAPTER II. INTRODUCTION.............................. A. Background............................ B. The Problem........................... C. Formulation of Hypotheses............. D. Definitions........................... Aggressive behavior................. Boys training school................ Case history reports................ Delinquent juvenile behavior........ Dogmatism.and Open-closed mindedness........................ Frustration-aggression hypothesis... Reading underaChievenlerltooeoeoooeooe SChOO]. attitudes.................... Self report of aggressive behaVioreooooooooooooooooeooooeooe E. Assumptions............................ REVIEW OF THE LITERATUREooooooooooooeoooooo A. The Literature on Reading.............. Achieving and nonachieving readers... - iv- PAGE ox I» r- Id 10 10 10 10 ll 11 11 12 12 13 13 15 15 15 CHAPTER B. C. D. E. Causes of reading underachieve- ment.I.OOOOOOOOOOOOOOOOOOOOOO0.0.. Personality factors in underachieving readers............ The need for remedial measures...... The Literature on School Attitudes.... School attitudes.................... Teacher-student attitudes........... The Literature on Aggressive Behavior. Frustration-aggression theory....... Classroom responses to frustration.. The Literature on Delinquency......... Some international considerations... The origins of delinquent personalities....................o School achievement and delinquency.. Social factors in delinquency....... Perception and intellectual faCtors in delinunnCYOOOOOOOOOOOO The Literature on Dogmatism. o o o c c o o e o o Dogmatism and personality. 0 o c o c o o o o o Dogmatism and adolescence........... Dogmatism and Iearning.............. -v- PAGE 18 21 24 27 27 29 30 30 31 33 33 34 36 38 4O 42 43 45 47 CHAPTER PAGE 111. METHODOLOGY............................... 4’9 A. Restatm: Of Purposes............... 49 Be SUbjeCtSooeocccococooooeoooooecceceoec 50 Advantages of sampling two populations....................... 50 Criteria for selection.............. 53 Selection procedures................ 54 C. Measures.............................. 55 The Wechsler intelligence scales.... 57 The Stanford Achievement Test....... 58 The Dogmetism.Scale................. 59 The Minnesota Student Attitude InverltorYQcocooooocococococcoocooc 6o Aggressive behavior measures . . . . . . . . 61 D. Procedures............................ 63 Group procedures.................... 63 Individual procedures............... 66 Method of analysis.................. 67 IV. RESULTS AND DISCUSSION.................... 70 .A. Results............................... 74 Summary Of results.................. 96 Be Dian-1881011000000...ooooecoococccccoccc 98 CHAPTER PAGE Dissimilarity of the two SUbJeCt gmupscoococccocccoooooec 98 The interrelationship among variables........................ 100 Limitations affecting the inVGStigationeoceooooooeeoeoooooo 108 v. SWRY MD mNaUSIONSCCCOCCOCCOOCCCCCCC 112 A. smwoooooooooooooooooooooooooooooo 112 Be CODCIUSionSroeeoooooooeooooooooocooc 116 C. Recommendations and Need for Further smdyOOOOOOOOOOOOOOOOOOOOOOOO0.0... 119 Recommendations related to further researChcccocoocooeoeooooooocoooo 120 Recommendations related to theory.. 122 Recommendations related to educa- tion............................. 125 APPENDIXOOOOOOOO0.0...OOOOOOOOOOOOOOOOOO. 127 REFERENCES...0.00.0.0...OCOOOOOOOOOOOOOO. 162 -vii- ‘0’ TABLE I. II. III. IV. V. VI. VII. VIII. IX. X. XI. XII. XIII. XIV. LIST OF TABLES PAGE Summary of Sample Pepulation Data......... 56 Correlations Between Reading and AggreSSive Behavior.................... 76 Correlations Between Dogmatism.and AgthSSive Behavior.................... 78 Correlations Between School Attitudes & Aggressive Behavior.................. 80 Correlations Between Intelligence and. AggreSSive Behavior.................... 83 Correlations Between Reading, Dogmatism, Intelligence, and School Behavior...... 84 correlations Between Dogmatism, School Attitudes, and Intelligence............ 87 Aggressive Behavior and Parents Income Source................................. 89 Aggressive Behavior and Number of Parents in the Home.................... 90 Aggressive Behavior and Place of Residence.............................. 91 Aggressive Behavior and Birth Order....... 91 Aggressive Behavior and Race.............. 92 Aggressive Behavior and Age at First Court Contact.......................... 94 Aggressive Behavior and Highest school Grade Completed........................ 95 Dissimilarity of Subject Groups...........lOl -viii- LIST OF FIGURES FIGURE PAGE I. Population Data, Lansing and Red Wing Training SChOOlSoeocoooceoococooccoco... 52 II. Voluntary Statement of Aggressive BehaVioreoooooecccocccccooooccooooooocoo 65 .1);- CHAPTER I INTRODUCTION A. Background The second half of the present century has already witnessed more interest and research into the outcomes of public school education than any other comparable period in history. Much of this interest has focused on attempts to answer the question, to what extent does school eXperience influence later social behavior? While teachers, school administrators, and education associations have typically emphasized the positive social value of education, little effort has been expended in searching the curriculum for negative social influences. As McNally (1965) has indicated, such examinations have too often been considered fringe topics by educators. Juvenile delinquency is a significant social problem that would seem to merit considerable attention from professional educators. However, as Kirk (1963) has indicated, "Controlled research into the educational aSpects of delinquency is practically nonexistent.“ (p. 348). Retardation in the basic academic skills of Juvenile delinquents has been demonstrated in the context of numerous studies, however, questions dealing with the relationship of failure in school tasks and sub- sequent patterns of delinquent behavior are not re- solved in the literature.. Although it would appear that early school failure might be accompanied by considerable frustration for the learner, little has been done to eXplore the social consequences of school frustration. As Long (1965) has said, "The school is more than a community. It is the child's image of society. To fail in this first society is to start life with a mind set toward failure." (p. 295). Perhaps the most significant task confronting the child in elementary school is that of learning to read. In discussing student behavior Combs (1959) has said: Behavior as seen by the behavior is not due to chance, it is a caused and pertinent aSpect of the world as he eXperiences it (p. 13). Combs further suggests that the child who fails in reading views this eXperience as a failure of his first and crucial test in school. Such failures could have devastating repercussions for the personality as a whole. Similarly, in discussing children with histories of failure in reading, Rosewell (1964) has suggested that such chil- dren have suffered years of deepair, discouragement, and frustration. The wider the discrepancy between achieve- ment and ability, the more serious the consequences. -3- Dollard and Miller (1939) have suggested that overtly aggressive responses to school related frus- tration are not possible for many children. Parents and the school combine efforts to present a united front against such behavior. Moreover, during the child's early years in school, it is probable that only the parents' solidarity with the schools, and specific legal codes requiring school attendance, are sufficient to keep many children in school. B. The Problem The relationship of school linked frustration which might occur early in the child's life as a possible result of underachievement in reading, and later attitudes about school, evidence of a dogmatic cognitive outlook, and aggressive behavior in boys, are not well understood at present. Many writers (Kvaraceus, 1945; Bills, 1950; Gerstern 1951; Peck, 1955; Harris, 1961) have observed reading retardation as a consistent factor in the delinquent populations they were studying. Greenblat (1955) cites studies which show the startling frequency of reading retardation among aggressive and hostile delinquents. Sorenson (1950), in a longitudinal study, observed well adjusted nursery school children deteriorate in their adjustment as a function of not learning to read effectively in the primary grades. Bettelheim (1956) has said that among children coming to the Orthogenic School, academic failure was the most common presenting symptom, and reading retardation was the most common problem in this category. Similarly, Rabinovitch (1962) has observed that even severe emotional disturbances tend to be found tolerable in the classroom if the child is making adequate academic progress. Gates (1936) has found the incidence of emotional disturbance in groups of deficient readers to be as high as 75 percent. Furthermore, it would appear that school achievement is a matter of considerable sociological importance. Grann gt. El. (1956) grouped children in 12 South Carolina schools into socially most accepted and socially most rejected categories using sociometric and social distance scales. School achievement was the determining factor at all grade levels. In describing the problems of the dyslexic child, Saunders (1965) believes that had specific diagnostic and remedial measures been available, there is evidence to suggest that emotional problems and anti- social behavior might have been prevented. In spite of the above findings, many questions remain unanswered. Why have some delinquent boys achieved relatively well in school? Why might two boys from the same family exhibit opposite patterns of socialization, -5- and are such differences in behavior related to school success? Why are many delinquent boys not physically aggressive in their behavior? More importantly, on what dimensions do non-aggressive delinquent boys differ from their aggressive peers? To what extent is reading under- achievement a factor in the above? As Margolin (1955. p. 26) has said, "There is a good deal of overlap between delinquency and underachievement in reading, with no clear evidence as to which is cart and which is horse.“ It was the purpose of this study: (1) to extend the findings of earlier well known research with delinquent children (Glueck and Glueck, 1957; Bandura and Walters, 1959) with respect to the possible relationships between academic frustration stemming from retardation in the acquisition of reading skills, and later patterns of aggressive behavior; (2) to determine the degree of relationship which might exist between reading retardation and Specific attitudes about school, as expressed by a pOpulation of delinquent boys confined in state training schools in Mishigan and Minnesota; (3) to test a frus- tration-aggression hypothesis (Dollard and Miller, 1939; Miller and Dollard, 1950; Dinwiddie, 1955; Maier, 1956; Gottfried, 1959) which might predict that delinquent boys who had failed to achieve success in a school setting because of retarded reading skills, would exhibit more anti-social types of aggressive behavior than other delinquent boys showing less reading underachievement; (4) to determine if the cognitive construct dogmatism (Rokeach, 1960) might be associated with reading under- achievement, aggressive behavior, school attitudes, intelligence, and other deveIOpmental factors in a pOpulation of delinquent boys; (5) to determine the degree of relationship which might exist between aggressive behavior and intelligence, birth order, family, social, and other demographic data in the popu- lations described above. C. Formulation of Hypotheses The existence of an inverse relationship between criminality and educational achievement has been cited. In our society considerable importance is placed on successful school achievement. School failure is an eXperience which might be eXpected to produce frustration for the individual who has failed to perform adequately. However, socially acceptable outlets for aggressive responses which might accompany frustration are often lacking. Wickman (1928) has suggested that teachers have long viewed aggressive displays by their pupils as a behavioral threat to learning, and have sought to extinguish such behavior. Furthermore, parents have not typically permitted the expression of aggressive -7- behavior in the home, and society has viewed such behavior as a threat to social order. Dollard and Miller (1939) have discussed the delinquent offender and societies reaction to delinquency by suggesting that social disapproval is the most severe form of punishment for aggressive behavior. These writers have also called attention to the traumatic nature of adolescence by observing that adolescence is known in our society as a period of increased aggressiveness and irritability on the part of youth. There is general agreement that delinquent boys most frequently come to the courts attention during the ages of 13 and 14 years. If it could be assumed that aggressive response tendencies to frustration were cumulative, and if the school, home, and society combine to limit such reaponses in young children, then the occurrence of aggressive forms of delinquent acts under the additional stress of adolescence would seem reasonable indeed. This study has taken the theoretical position that children who fail to achieve in school, in the absence of adequate home and community controls on aggressive behavior, will be observed to become delinquent. The position was also taken that reading under-achievement presents one of the most obvious indications of school failure to the learner. In the absence of adequate ~8- controls on aggressive behavior, children who fail to achieve in reading may exhibit their aggressive behavior in a variety of delinquent forms. It has further been theorized that for students who do not achieve in reading, attitudes about past school experiences will be more negative than for reading achievers. It was also viewed as possible that a more structured, rigid, and guarded cognitive outlook toward new ideas and innovations might result from prolonged frustration occuring early in life. The construct dogmatism which has been described by Rokeach (1960) defines such behavior. The theoretical position for this study has rested on the observations by numerous writers that delinquent acts of an aggressive type, together with various forms of peer group alienation, have been observed to occur in adolescents whose school experiences have been essentially frustrating. Much existing evidence drawn from the work of others has tended to suggest that failure to develop adequate reading skills may contribute significantly to a generalized feeling of failure and inadequacy in the school setting. Because of additional evidence to suggest that most physical, intellectual, and emotional handicaps do not cause insurmountable educational problems when some tangible evidence of school achievement is forth- coming, the writer has become increasingly of the opinion that acts of aggressive delinquent behavior, directed outward toward society, may be the product of long years of frustration, which has been either caused or seriously aggrevated by school failure. In order to examine the theoretical position described above, the following research hypotheses were formulated: 1. There would exist a significant correlation between aggressive behavior and reading. 2. There would exist a significant correlation between aggressive behavior and dogmatism. 3. There would exist a significant correlation between aggressive behavior and school attitudes. 4. There would exist a significant correlation between aggressive behavior and intelligence. 5. There would exist a significant correlation between reading and either dogmatism, school attitudes, or intelligence. 6. There would exist a significant correlation between dogmatism and either school attitudes or intelligence. 7. There would exist a significant correlation between school attitudes and intelligence. 8. Aggressive behavior would be related to one or more of the following demographic factors at a significant level: parental income, number or sex of parents present in the home, -10- place of residence, birth order, race, age at time of first court contact, or highest school grade completed. D. Definitions Aggressive behavior. Throughout this investigation the term aggressive behavior was used to refer to acts of verbal or physical hostility directed at the person or property of others by boys confined in the training school populations. Aggressive behavior was differentiated from other forms of delinquent behavior such as stealing, truancy, and the illegal use of tobacco, alcohol, drugs, etc. as examples. It was further limited to acts directed against others, and not to those forms of aggression turned inward upon the self. Boys training_schools. Residential training facilities of an academic and/or industrial nature maintained at public expense for the purpose of correcting the delinquent behavior patterns of juvenile offenders. In this investigation the Lansing Boys Training School (Michigan) and the Red Wing Boys Training School (Minnesota) were specifically implied. Case history reports. Individual reports or files maintained for the purpose of presenting accurate family, medical, educational, behavioral, and legal information -11- pertaining to the residents of the boys training schools described above. File rooms were made accessable to the writer and information contained in the subjects' case history reports became an integral segment of this study. Delinquentjuvenile behavior. In this investigation the term delinquent behavior, and juvenile delinquency, referred to illegal acts committed by the training school populations who served as subjects in the investigation. Dogmatism and open-closed mindedness. For this investigation dogmatism and open-closed mindedness were used in the sense that these terms have been described by ROkeach (1960). The Dogmatism §£glg has been demonstrated to measure the individual's ability to synthesize new beliefs into belief systems with high dogmatic individuals being less able to perform this task. Dogmatism is a system variable, a characteristic of the total system. The open-closed concept was developed to bring into relief differences in the individual's behavior with resPect to the formation of new systems on a continuum from low to high dogmatism. In a sense, dogmatism is a measure of rigidity or inflexability of thought patterns and beliefs. The Dogmatism §£glg was used to measure this variable in the population of this investigation. Frustration-aggression_hypothesis. In this investigation, the hypothesis developed by Dollard and -12.... Miller (1939) and amended by Miller and Dollard (1950), Dinwiddie (1955), Maier (1956), and Gottfried (1959). Behavior which is frustrated in its intent or purpose always tends to be followed by aggressive responses. The frustrated persons behavior is goal-less and task orientation disappears. The possibility that reading underachievement is an intensely frustrating experience, and that aggressive behavior may result, were explored in this investigation. Reading underachievement. In this investigation readingxuflerachievement was determined by a mental age method for each subject. This method was suggested to the writer by Dr. Guy L. Bond of the University of Minnesota. Grade level reading expectancy was estimated by subtracting 5 from the subjects chronological age and multiplying this difference by the subject's Wechsler intelligence quotient. The resulting product was then divided by the subject's reading score obtained by using the Intermediate Battery, Form K., of the Stanford Achievement Eggt. The resulting quotient then became the measure of degree of reading underachievement. School attitude. In this investigation school attitude referred to the opinions, beliefs, and feelings about past and current school experiences as elicited by a 62 question instrument, The Minnesota School Attitude -13- Inventory, developed by Flanders (1960). Positive school attitudes were expressed by the magnitude of the total school attitude score for each subject. Self report of aggressive behaviqg. For this investigation it was deemed necessary to obtain three measures of aggressive behavior for each subject. One measure, the self report, was obtained by individual interviews with each subject responding to structured questioning based on a listing of 13 areas of aggressive behavior, ranging from threats directed toward others, to assault with a deadly weapon. The self reports of aggressive behavior thus obtained were used to assign subjects to groups for statistical analysis. E. Assumptions This study has assumed that school experiences of underachievement in reading may be deeply frustrating. It has been assumed that anti-social forms of aggressive behavior may be one form of reSponse to frustration, and that boys confined in state training schools provide an accessable population for the study of anti-social aggressive behavior. It has also been assumed that while the case history reports maintained for each boy at his state training schools may not have represented an inclusive summary of all available data about each .1;- boy, the data there contained were essentially accurate. It has further been assumed that currently existing instruments for the appraisal of reading achievement levels, intelligence, dogmatism, and school attitudes are sufficiently objective, valid, and reliable as to have provided usable information for this research investigation. Lastly, it is assumed that the test data obtained in this study have represented the best efforts for each subject, and that the data have been accurately scored and processed. CHABTER II REVIEW OF THE IITERATURE Much has been written about the incidence and causes of delinquence in the general school aged popu- lation. Similarly, much has also been written about personality deviations which often appear in association with academic underachievement, especially in reading. A growing but still relatively small body of knowledge dealing with dogmatism has appeared in the literature. Very little information regarding school attitudes and their relationship to academic achievement has been published. This chapter will attempt to present a relevant abstract of the above questions and to relate this literature to the current investigation. A. THE IITEP-ATUPE OI? READIIIG Achieving and non-achieving readers. Berkowitz and Rothman (1961) believe that a reading disability can cause complete failure in school, and the child who fails academically must suffer emotionally. These writers also believe that regardless of the problems which disturbed children present, the majority of such children suffer from some degree of retardation in reading. -15- Roswell and Natchez (1964) define a reading disability as a discrepancy between reading achievement and intelligence. The wider the discrepancy between achievement and ability, the more serious will be the effect. The psychotherapist Haworth (1964) has estimated that more than 10 percent of the children in public schools do not learn to read adequately in the regular classroom situation. She also believes that we have probably tended to underestimate the importance of this problem as a factor in delinquency. Natchez (1959), in a comparison of the behavior of readers and non-readers in fifth and sixth grades of three southern New York schools, noted that the retarded readers reaponded to school situations with significantly more frustration, aggression, dependency, and withdrawal. Ort (1962) stressed the fringe existence of non-readers, both during childhood, and even into later adolescence and adulthood. He has observed that non- reading appears to result in social alienation and self ex- trangement in our culture. Porterfield (1961) found a significant difference between peer prestige status and reading achievement. This relationship held true irrespective of a) the type of peer prestige, or b) the socio-economic level of the school community. -17- In a similar investigation, Tabarlet (1958) studied a group of retarded readers and a group of controls from the BatOn Rouge, Louisiana Schools. On the basis of a Mental Health Analysis, published by the California Test Bureau, the two groups differed at the .05 level on "behavioral immaturity, interpersonal skills, social participation, satisfying work and recreation, and adequate outlook and goals." (p. 524). McMurray (1963), using an N. of 742 Canadian subjects, found non-achievers in reading to diaplay more irresponsibility, shorter attention spans, and to receive generally poorer acceptance by peers. Carter (1964), using an N. of 900 California seventh and eighth graders, found students reading below grade level to be less successful in school, generally less happy, and to display more negative values toward schools. The high overall correlations between success in reading and success in most other academic subjects, at both the elementary and secondary school levels, have been described by several writers (Robinson, 1946; Hinkelman, 1956; Aaron, 1960; and Henderson et. al., 1965). It would seem apparent that a reading disability may be related to school failure, delinquency, and aggressive behavior. The degree of influence which such disability exerts on behavior may be associated with the discrepancy between intellectual potential -18- for learning, and actual level of achievement. The frustrating effects of a reading disability may also result in loss of peer prestige, and thus involve social as well as academic consequences. Reading under- achievers have been observed to diaplay immature be- havior, and to voice many negative attitudes about school. From the studies reviewed here, it would seem appropriate to narrow a study of reading disability to a limited number of possible behavioral correlaries of reading under-achievement, one of which might be aggressive behavior. Causes of reading under-achievement. The determination of why students under-achieve in reading is a complex undertaking. Durrell (1958), in a study of 2000 first graders, cited ineffective reading skills and inadequate word analysis techniques as important factors. He also found no relationship between chronological age and success in reading, and only a low relationship between mental age and reading success. Curray and Hughes (1961) found that "based on IQ pre- dictions, students eXpected reading achievement and actual reading achievement do not always correlate highly" (p. 91). Ames and Walker (1964) attempted to combine the Rorschach and Wechsler Intelligence Scale for Children to predict reading success. Their r.'s were -19- .55 Rorschach, .57 fll§g, and .73 for the combined method. Betts (1956) has strongly refuted the theory that we can predict reading success on the basis of intelligence alone. Shimota (1964) analyzed his data on 360 children ages 15-16 in the Western State HOSpital (Washington). Of these disturbed children, 51 percent were disabled readers. The disabled readers did not differ from their peers on the basis of intelligence, E.E.G. reports, neurological reports, type of emotional problem, mixed dominance, or any other physical handi- caps. Chiland (1964, p. 26) was able to trace reading and spelling errors to "emotional and family factors." Dentler (1964), however, stated that the causes and methods of preventing dropouts are largely traceable to what happens in the child's learning eXperiences in the primary grades. Some writers have attempted to ascribe the causes of reading under-achievement in the elementary school to sociological factors. Chandler (1966) feels there appears to be a rather direct link between reading achievement in the elementary school and l) father's occupation, 2) number of books in the home, and 3) annual income of parents. Roman, Margolin, and Harrari (1955) see reading under-achievement and delinquency to be related to the clash between some childrens value systems, -20- and the essentially middle—class value system of the schools. Granzow (1954) found parents of under- achievers to have lower socio-economic status, poorer homes, more indifference to schools and learning, poorer educations, more broken homes, and less peer acceptance. Crescimbeni (1964), investigating the variable of broken homes, found this factor to be more significant than I.Q., age, or grade level in pre- dicting reading under-achievement. Parental relationship factors have also been cited by some writers as possible reasons for reading under-achievement. Shaw and White (1965), using an adjective check list, found inadequate identification patterns in a pepulation of 114 school under-achievers. Nurberger (1955) cited a rather detailed case history which seems to show how reading under-achievement led to delinquent behavior as a result of the father's reactions to school failures. Coleman, Barnston, and Fox (1958) studied the parental trait complexes of male retarded readers referred to a reading clinic at the University of Southern California. A pattern of "domineering mother who exerted pressure for educational achievement, coupled with an inadequate father figure predominated." (p. 51). In conclusion, it can be said that current know- ledge suggests a multiple causation for reading under- -21- achievement. Poorly learned basic reading skills, or the absence of such skills, have been shown to be important factors. The use of I.Q. tests has; not always proven to betzdequate method. of predicting reading under-achievement in delinquent populations. Emotional stability, together with an adequate identi- fication figure, would seem to facilitate success in reading. Attitudes of parents and other significant adult figures about the seriousness of reading under- achievement may influence later classroom efforts. It seems clear that considerably more research will be needed concerning the causes of reading failures. It also seems likely that no single theory will adequately explain why all children fail to develop successful reading skills. However, this does not limit the feasibility of investigations into the behavioral consequences of reading failures. gersonality factors in under-achieving readers. The complex relationship between personality development and under-achievement in reading has received considerable attention in the literature, (Keshian, 1962; Healey, 1965; Knafle, 1965). Unfortunately the findings do not always agree, and occassionally distinct contradictions seem to appear. This should perhaps be viewed as a symptom of the complexity of the problem, rather than as a reason to ignore the findings, or to abandon the subject. -22.. As previously cited, the work of Sorenson (1950) would seem to suggest a very intimate link between success in reading, and satisfactory personality growth. Wattenberg and Clifford (1964), using an N. of 185 children in a Detroit study of self-concept, noted that feelings of competence and personal worth were significantly correlated with success in reading. Bruck (1962) reported an r. of .60 between self-concept (based on Machover drawings) and school achievement, using an N. of 60. In his doctoral thesis, Copple (1961) found positive self-concept and success in reading to be significantly related at the .01 level. Sophis (1966) stated that he was able to isolate the variable self-concept as a reader which he said related directly to reading progress. Unlike writers cited earlier, Berks (1957) was able to associate abnormal E.E.G.'s with percep- tual learning difficulties. Tuller and Eames (1966) reported similar success with respect to reading failures and E.E.G. evaluations. The suggest the problem is a lesion located in the parietal post- temporal region of the cortex. While abnormal E.E.G. findings do not always signify a personality or behavioral association, and while many subjects with normal E.E.G.'s behave abnormally, the occurance of abnormal findings has proven to be a valuable research avenue in numerous psychological studies. -23- The use of projective techniques has been re- ported to have been successful in differentiating non- achieving from achieving readers. Abrams (1956) noted more Rorschach C.F. and K. reaponses in his non-readers, indicating high anxiety levels and less maturity and impulse control. Knoblock (1965), however, did not find high reliability for the Rorschach in differentiating between verbal intelligence and reading, in his pOpu- lation of second graders. Solomon (1953) has reported being able to make such a distinction between achieving and non-achieving readers. Spache (1957, p. 467) used the Rosenzweig P-F instrument to isolate "conflict with authority figures and a tendency toward withdrawal behavior" in his sample of 125 retarded readers. Krippner (1966) successfully measured reading improvements and personality change using the Holtzman inkblot technique. Miscellaneous factors in the personality of under-achieving readers are mentioned frequently in the literature. Hummell and Sprinthall (1965) cited immaturity of outlook and unwillingness to postpone goal rewards. Brunkan and Shen (1966) found patterns of self-depreciation to relate to both rate and quality of reading; while Edwards (1958) found desire for social acceptance by parents, peers, and teachers to be a crucial variable. -24- It is difficult to summarize the literature dealing with personality factors common to reading under-achievers. Well controlled longitudinal research (Sorenson, 1950) would seem to point to a rather clear relationship between failure in reading and personality deterioration. Numerous studies have also been cited which correlate negative self-concept with unsuccessful reading skills development. However, E.E.G. studies and projective techniques have not detected specific or predictable neurological or behavior profiles in such individuals. It is probable, again, that the absence of dramatic findings is merely an indication of the complexity and dynamic nature of human personality. The need for remedial measures. Many writers have stressed the need for remedial reading, and for remediation in combination with therapy, for both delinquent and disturbed children. Rabinovitch (1962) points to the successes of speech remediation, and asks if more cannot be done with reading problems. Balow (1965) has published encouraging data on the long term effects of remedial instruction. A problem in the assessment of successful remedial programs center on the multiplicity of programs attempted. Staats and Butterfield (1965) have success- fully employed reinforcement principles with delinquents. -25- Bills (1950), Peck _e__t_. 31. (1955), Roman (1955). and Margolin (1955), have all demonstrated highly signifi- cant outcomes using remedial reading and psychotherapy in combination, as opposed to either remediation or therapy alone. It is very interesting to note that their significance levels have held for improved adjustment as well as for improved reading skills. One might seriously ask if additional study should not be given to the question of adding more remedial reading facilities to our psychological clinics when therapy with children is involved? Most remedial reading approaches have been of the typical variety, and some have been of a rather exotic or imaginative type. Robbins (1966) has reported finding no validity for Delacato's neurological retrain- ing exercises which were earlier reported to influence reading skills. Lay (1965) has reported the success- ful use of role playing techniques with a delinquent population. This procedure was reported to improve communications skills and self-concept, thus making other therapies more successful. Illovsky (1963) successfully used post-hypnotic suggestion with de- linquent boys to produce a remedial reading gain two and one-half times that of his controls. The suggestion "they need not fear reading," and that "they would be successful" (p. 65) was used. -26- Krippner (1964), working in the reading clinic at Kent University, noted improvements significant at the .01 level on the Mental Health Analysis Inventory, and significant at the .05 level on the Wechsler Intelligence Sgglgifgg Children, Verbal S2312, for his population of 30 subjects, following remedial reading instruction. Similarly, reduced anxiety levels and improved self confidence has been reported by Raygor (1959) for a post high school N. of 88, after seven weeks of "intensive" remedial reading. 4 One may conclude that a number of writers have cited the need for remedial measures with students who have not achieved in reading. While the number of specific remedial reading programs, some of them highly imaginative, are too numerous to analyze in detail, the rather encouraging results reported in most studies suggests that unsuccessful reading skills can be im- proved. This improvement has often been reported as successful when carried out in a tightly controlled environment, such as that of a psychotherapy group. It is possible that one reason for the encouraging results of remediation in a structured environment might be due to the increased opportunity for administering appropriate support and timely reinforcement. -27- B. THE LITERATURE ON SCHOOL ATTITUES School attitudes. Statistical analysis of the attitudes of delinquents toward the school are not well represented in the literature. Ball (1955), in his doctoral dissertation, has presented one factorial analysis of delinquent attitudes. He notes statistically "significant eXpressions of negative attitudes toward schools and educational values" (p. 2329). Flanders (1960) utilized a Health, Education, and Welfare Project Grant to develop and standardize a 62 item instrument, Th2 Minnesota School Attitude Inventory, which satisfactorily measures school attitudes. This instrument was employed by the writer in the present investigation. In their study of delinquent boys, Pearson, Barton, and Hey (1956) found the School Motivatigg Analysis Test to correlate highly with school reports for their academically achieving delinquent boys. These findings tend to suggest an association between academic success and positive school attitudes in popu— lations of delinquents. In a survey of the attitudes of 1154 public school adolescents, Clard and Wenningen (1964) found a close correlation between negative attitudes toward school and acts of admitted illegal behavior. Roebeck -28— (1964) studied two groups of public school retarded readers; one group at the primary level, and one group of adolescents. While low self esteem measures were found only in the older group, both grOUps exhibited negative attitudes toward the school. Wilson and Morrow (1962) found their public school under—achievers eXpressing significantly more negative school attitudes, and also "seeing the school as less supporting of emotional health and stability" (p. 702). Jackson and Getzels (1959) found sex difference in the attitudes of dissatisfied adolescents, with girls reacting to school dissatisfaction with feelings of personal inadequacy, and boys centered on projective attitudes of conflict with school authorities. In summary it can be said that few studies of school attitudes have been done with delinquents. Th3 Minnesota School Attitude Inventory promises to success- fully measure such attitudes in delinquent boys. The present study, therefore, should add materially to existing knowledge regarding the association of school attitudes and reading achievement; one indicator of school success. -29- Teacher-student attitudes. It would appear that student attitudes about school are related to school achievement, and to the past social behavior of the learner. There is some evidence to suggest that teachers' attitudes about their students, and about the value of education, may also be significant variables. Davidson and Lang (1960), with an N. of 203 subjects from fourth, fifth, and sixth grades in New York City, demonstrated that "Children's percep- tions of their teachers feelings toward them were correlated significantly with school behavior" (p. 116). Baracheni (1962) observed that "Teacher attitudes about academic achievement and student potential influenced social interactions and achievement motivation in the classroom." Weaver (1959) found the semantic difference in attitudes between students and teachers to be as closely correlated with student learning as intelligence test scores. While the present investigation will not directly examine the relationship of teacher attitudes toward delinquent boys, the knowledge that such attitudes exist and may be related to achievement must be emphasized. -30- C. THE IITERATURE ON AGGRESSIVE BEHAVIOR Frustration-aggression theogy. According to the theory deve10ped by Dollard and Miller (1939), behavior which becomes frustrated or blocked in its goal direction is always followed by some form of aggressive reaponse. Numerous examples from the behavior of individuals and groups are cited to support their theory. Their theory was amended by Miller and Dollard (1950) when responses to frustration were seen as occurring under the influences of cultural determinants. This helped to explain obvious differences in responses to frustration as functions of specific cultural values governing the behavior of its members. Miller (1941) had previously ascribed the cultural con- trols of behavior to a ’COpying of identification models' within the culture. Dinwiddie (1955) eXpanded these theories further, and demonstrated experimentally that: 1. The degree of similarity between the frustrated response and the aggressive ex- pression of this response will vary inversely with the strength of the drive inhibition for that response (i.e. social training). 2. The strength of drive motivating aggressive responses will correlate directly with degree of frustration in the absence of inhibiting drives toward overt aggression (p. 27). Because a population of confined delinquent boys may probably be assumed to lack aggressive drive in- hibitations (inasmuch as adequate drive controls would have functioned to limit the delinquent behavior which resulted in confinement), such boys provide an excellent group of subjects in which to test the above theories. Using a population of delinquent boys, Gottfried (1959) employed Miller's theory to demonstrate (significant at the .05 level) that his sample differed from the non- delinquent controls on the variables goal attainment and aggression. Maier (1956) had previously shown that "the frustrated person's behavior is goalless and task orientation disappears" (p. 2890). He also stated that "prolonged frustration, cepecially in important personal and social areas, may have a very serious effect." Frustration-aggression theory might easily be adopted as one possible explanation for aggressive be- havior in non-reading adolescents, if it can be demon- strated that failure to learn adequate reading skills is frustrating for the learner. It is also probable that many delinquent boys lack the social training and suitable behavior models which might inhieit overtly aggressive reaponsee to others. Classroom reaponses to frustration: As previously cited, numerous studies have shown the relationship of -32- school experiences of frustration with associated goal- less behavior, and acts of aggressive anti-social be- havior by the non-achievers. Quay and Blumen (1963) analyzed the court records of 191 white male delinquents for the presence of 13 delinquent factors. After rota- tion, four factors emerged; truancy, impulsivity, inter- personal aggression, and impersonal aggression. In a study of the behavioral manifestations of learning and non-learning, Harris (1961) found the non-learners to exhibit greater extremes of behavior, ranging from extreme submissiveness, to aggressive attacks upon other students. Their source of frustration was associated with under-achievement in reading. Magee (1964), and Shaw and Grubb (1958), have separately related school under-achievement to acts of aggressive classroom behavior. In an interesting study, Wagenheim (1960, p. 192) used "speed of recall" to evaluate memory for mobility and aggression. Subjects verbalizing the greatest numbers of such memories were also the ones showing the greatest spread between reading achievement and expected achievement. Pine (1965) evaluated the vocational aspirations of delinquents and found an inverse relation- ship between aspiration levels and histories of anti- social aggression. This might relate directly to the goal-less behavior pattern already mentioned in connection with frustration. -33- The studies cited above show that frustration might result from academic under-achievement. This frustration might be exhibited in behavior which might include truancy, impulsivity, and aggression. The behavior of frustrated students has been characterized as extreme. Delinquent boys have been shown to diaplay anti-social forms of aggression and low vocational aSpiration. These findings tend to reinforce the need for additional investigations into the dynamics of aggressive responses to school related frustration. D. THE IITERATUHE OBI DELINQUEI‘ECY Some international considerations. The problems of delinquency and its association with school under- achievement are by no means confined to the United States. Critchley (1964), in his book on dyslexia, cites the very high incidence of delinquent behavior in non-achieving readers in England and Scotland. Gregory (1965) used the British Social Adjustment Guide instru- ment with a population of “village" children in W. Berkshire. He found a significant connection between reading failures and "school restlessness, anxiety, and an inflated need for peer approval at the eXpense of social conformity and adult restraints" (p. 67). -34- In a highly relevant study in India, Tutto (1957) studied a sample of 100 maladjusted and delinquent students. He found poor reading achievement to be associated with delinquency, negative attitudes toward the school, and "significant resistance to new opinions, attitudes, and change" (p. 109). Both Kajimura (1958), and workers at the National Institute for Education Research (1959), in separate controlled research in Japan, found non-achieving students to diaplay significantly more delinquent and anti-social behavior than achieving students. Achieve- ment level was found to correlate with greater significance than the I.Q. for predicting delinquency and conforming behavior. In summary, it seems evident that delinquent behavior associated with school under-achievement has been observed internationally. It would seem reasonable on an a-priori basis alone to assume that the causes and effects of school related frustration might well transcend geographic and cultural boundaries. The origins of delinquentgpersonalities. Numerous writers have attempted to treat the question of origin of the delinquent personality. Most workers have approached the problem from the perspective of multiple causation (Mannheim and Wilkins, 1955; Block and Flynn, 1956; Roueck, 1958; Balogh, 1958). However, theories -35- of rather Specific causation have also been presented. Bandura and Walters (1958) deve10ped their classic investigation using the theory that aggressive behavior in boys is the result of forced close dependent involve- ments, because of unfavorable early socialization, in the absence of a suitable father-identification figure. These boys were controlled by fear of punishment rather than by internal controls. Bernabeau (1958) believes that delinquency results from the inability to give Up infantile fantasies of omnipotence. Nye (1958) sees delinquency as caused by a lack of family and social controls, while Mukherjeek and Kundu (1961) have found birth order to be significant. The Glucks, with a sample of 100 delinquents, associated delinquency with body type. Wirt and Briggs (1959), using 2,000 delinquents and 2,000 controls, found a lack of success in school, lower school grades, greater dislike for school, and more aggressive behavior patterns, in their delinquents. In an earlier paper, Backwin (1955) has asked if a major cause of delinquency might not center around the practice of prolonging school eXperiences for adolescents who have not deve10ped effective reading and other school-oriented study skills. -36- Although a multiple causation theory for the develOpment of the delinquent personality seems to be the most promising at the moment, it is perhaps only appropriate, in summary, to comment that no consistent theory has yet been developed which includes school related factors in a frustration-aggression type behavioral model. School achievement and delinquency. The in- ability to c0pe with normal school experiences has been described by Brownell (1954) and Dexter (1964) in their discussions of delinquent boys. The Harlem Project Report (1945) stated that in three Harlem schools containing the greatest percentage of delinquent boys in New Eork City, although the mean I.Q. for these schools was 99, only one-tenth of the boys were reading at grade level. Kvaraceus (1945) reported that his sample of 750 delinquents differed significantly from non-delinquents on grade repetition, with almost every delinquent having repeated at least one grade. He observed unhappiness and frustration to characterize this group. There is some evidence to suggest that potential delinquents can be rehabilitated through effective remedial teaching and other classroom techniques. Bowman (1959) reported a study in which potential -37- delinquents were transferred to classrooms described as warm and accepting, where it was possible to provide intensive remedial reading and prevocational work skills eXperience.. This group showed a 33 per- cent lower rate of subsequent delinquency than a similar group not so programmed. Arbuckle and Litwack (1960) examined the recidivism rates of 500 releasees from a Massachusetts correction school. Among the positive factors in success of parole was school grade completed. Gersten (1951) compared academic gains for two groups of delinquent boys over a twenty week period at the New York Training School for Boys. One grOUp received 20 sessions of psychotherapy (one hour per week), and the other group acted as controls. Pre and post testing revealed a "20 month gain for the eXperi- mental group as compared to a three month gain for the controls" (p. 317). Most studies of delinquent boys have cited school maladjustment as a definitive characteristic of these boys. Such observations are eSpecially common in urban pOpulations. It has also been shown that delinquent boys who were relatively successful in school were less likely to return to the training school a second time. It is likely that future investigations will provide additional data about the relationship of academic achievement and delinquency. -38- Social Factors in delinquenpy. It is a common observation that in urban communities the socio-economic status of parents determines to a large extent the location and type of one's residence, the schools available, social and peer group associations, and numerous other related factors. Such social consider- ations have been examined in detail by a number of workers, including Bandura and Walters (1959), in relation to influences on adolescent delinquent be- havior. The significance of economic minority group membership on personality and behavior appears to be dramatic. Dimitz, Kay, and Reckless (1958), using an N. of 717 sixth graders in Columbus, Ohio, employed three delinquency measures to examine the variables of sex, race, intelligence, neighborhood, school achievement in reading and arithmetic, and teacher nomination. Findings were in the expected direction, with smart white girls from good neighborhoods doing best in school, and showing the smallest incidence of delinquency, and dull Negro boys from poor neighborhoods achieving least well, and showing the highest rate of delinquency. Interestingly enough, the relationship between reading under-achieve- ment and delinquency was not significant in this study. -39- The high incidence of delinquency in a depressed Negro environment was reported by Segal (1966). He found "71 percent of Negro boys had committed offenses against persons or property, and 63 percent had committed related offenses in violation of social norms" (p. 29). It is very difficult to isolate specific variables in a causal relationship to delinquency. Havinghurst (1959) has traced adolescent patterns of delinquency, in part at least, to the school's failure to meet early readiness and reading problems. Glueck and Glueck (1957) see working mothers as a significant variable. lively, Dimitz, and Reckless (1962) believe that direction of socialization and inadequate self-concept have been the best predictors of delinquency. Bentley (1961), however, while finding delinquent acts to correlate with age, sex, birth order, and quality of parent-child relationships, did not find either socio-economic status or self- concept to be significant. In summary, it is probably accurate to say that the influence of socio-economic factors on delinquency and other behavior is extremely significant. Such factors are especially pervasive in urban areas. Minority groups have been shown to be especially vulnerable to economic and social pressures. Additional investigations relating family income, place of residence, race, and other factors to aggressive behavior are badly needed. -40.. Perception and intellectual factors in delinquents. Perceptual abnormalities in delin- quents have been reported by two writers. Zolik (1958) administered the Bender Gestalt instrument to 43 adolescent delinquents using Pascal and Suttell's scoring procedure. He found ”signifi- cant differences on all control group comparisons" (p. 26). Petrie, Asenath, and McCullock (1962) found a typical size comparisons in delinquents. The factor of perceived sense of social responsibility was related to successful school achievement by Narayana (1964). Johnson and Stanley (1955) compared delin- quent and non-delinquent boys, ages 10-12, on their perceived relationship to authority figures. The hypothesis that delinquent boys would exhibit significantly more hostile attitudes was not supported. However, both groups expressed significantly more hostility toward female figures. This might be eXpressed in the student-teacher relationships. Intellectual limitations have consistently been associated with delinquent populations. How- ever, Wheway (1958) states that a review of existing studies indicates sheer lack of intelligence -41- is seldom, if ever, a major factor in causing delinquency. Richardson and Surko (1956) found a mean IQ (WISC) of 88.4 in a population of 105 New Jersey delinquents. Quay (1965) believes that early studies seriously underestimated the intelligence of the delinquent. He believes that the earlier findings of 15-20 points below normal, and retardation rates of five times that of normal, are significant understatements of current findings. Quay has indicated that recent studies find means of minus eight points with consistent patterns of vocabulary lower than performance. Object assembly and picture arrangement are commonly observed to be higher than block design and picture completion. Abrams (1956) has reported similar findings. Coplan (1961) made the interesting observation that while low IQ delinquents show academic under-achievement and aggressive behavior patterns, high IQ delinquents tend to over-achieve in school, and to aggress through sociopathic channels. Shaw and McCuen (1960), and Stone and Rowley (1964), have also reported studies dealing with the relationship of intelligence and abnormal behavior. -42.. mile the perceptual skills of. delinquents have hem found to be atypical by some writers, intelli- gence has not been related directly to delinquent behavior. It is probable that earlier estimates of the intelligence of. delinquent boys have been too conservative. It remains for future research to . examine correlations between intelligence, aggres- sive behavior, and related cognative factors in papulations of. delinquent boys; Studies dealing with Rokeach's construct dog- matism (1960), and with its behaviorism correlation open and closed mindedness, while not numerous in the literature, show a definitely increasing inter- est in this subject by current workers. Since. the W m is a relatively new instrument, some of the current findings should probably be regarded as tentative, pending a more substantial body of literature. Unfortunately most of the existing research has been done with college-aged and adult powlations.‘ Little is known about patterns of dogmatic behavior in delinquent boys; -43- Dogmatism and personality factors. Rokeach and Fruchter (1956, p. 360), in their factorial study, found dogmatism to be "discriminable from authoritarianism, ethnocentrism, and rigidity." They also found that “dogmatism, paranoia, and self rejection are factorially similar." They further demonstrated, with an N. of 207, that dogmatism, paranoia, and self rejection would emerge with anxiety as a single factor. It should be noted that it was the self rejection link to dogmatism in Rokeach's findings that suggested the possibility of employing the Dogmatism Scale in a study of delinquent behavior. Alson (1959) attempted to correlate dogmatism with "Zajonc's measure of cognative structure" using a population at the Veterans Administration Center in Bath, New York. He noted correlations of practically zero. Zagona and Zurcher (1965), with an N. of 517, found that it was possible to demonstrate (beyond .001) an inverse relationship between dogmatism and verbal ability. Roberts and Herrmann (1960), and wrenn (1962), studied questions relating to time perspective, feelings of anomie, and the effect of preparatory involvement on goal valuation in high and low dogmatics. Their results indicate that high dogmatics tend to have -44- imbalanced rather than future orientatee time per- Spectives, and that this results in greater feelings of anomie, which exercises a disturbing influence on both present and future behavior. Moore (1962) studied verbal operant con- ditioning with high anxiety, dogmatic, college students. It was hypothesized that these sub- jects would condition easier than low anxious, low dogmatic subjects. The opposite findings occurred. With a similar population from the Counseling Center at Michigan State University, Kemp (1961) found that high dogmatic students had significantly greater numbers of personal problems, and that these prob- lems were not resolved in counseling. The tendency of high dogmatic subjects to deny defensive behavior was observed by Byrne, Blaylock, and Goldberg (1966). In a study of college students, Harvey (1963) noted greater tendencies toward conforming behavior in his high dogmatic subjects. In his study of social factors and dogmatism Sticht (1966) noted subjects experiencing high rates of geographic mobility to be significantly more anxious and dogmatic than subjects eXperiencing low geographic mobility. -45- The literature on dogmatism can be said, in summary, to indicate a need for additional studies before establishing possible relationships between this variable and others. It would appear that dogma- tism is related to self-depreciation and inversely related to verbal skills develOpment. It would also appear that high dogmatic subjects tend to have more personal problems, and that they lack a future time orientation. Because dogmatism has not been correlated with agressive behavior in delinquent boys, and because the relationship of dogmatism, reading, and school attitudes is not clear, the need for additional investigations in these areas would seem warranted. Qggmatism and adolescence. Anderson (1962), with an N. of 290 junior high school students in Edmonton, Alberta, observed that "there is a signifi- cant decline in dogmatism during adolescent years" (p. 135). He also reported a significant interaction effect between dogmatism, intelligence, and sex. Intel- ligent females tended to be more dogmatic than intelligent males. He further observed that child rearing practices were the basic determinants of dogmatism, and socio—economic status and dogmatism were inversely related. Bolmeier (1966) compared dogmatism in parents, to the adjustment of their -46.. high school aged children. He found that "In general, parents who were more Open in their thinking, accord- ing to the Dogmatism Scale, had children who were apt to score favorable on certain measures of adjustment“ (p. 5572). Meier (1960) was unable to establish levels of intensity of opinion, or resistance to accep- tance of teen age norms, by using the gpgmatism S3312 with a nation wide sample of high school students. He suggests that the whole concept of dogmatism is too loosely defined. He believes additional work should be done to strengthen its nomological net. Miller (1965) established that in his population of adolescence, "high dogmatic subjects under maximum conditions of involvement were most resistant to change" (p. 130). Paues (1963), using an N. of 675 students in Connecticut, studied dogmatism, self-image, and intelligence. She found "the higher the IQ, the more open the mind, and the less favorable the self-image; the more favorable the self-image, the more closed the mind; and, the higher the school achievement level, the more open the mind, and the less favorable the self-image" (p. 114). -47- Dogmatism and Learning. Virtually all of the available studies of the relationship between dogmatism and learning have involved college pOpu- lations. Ehrlich (1961), at Ohio State, confirmed the hypothesis that dogmatism implies "closed cog- nitive structure and so affects the capacity to learn in a negative direction." Rebhan (1966), in a study of three undergraduate college groups, demonstrated significant levels of test anxiety in high dogmatic subjects. However, Christensen (1963), using 166 students in an introductory psychology class, found "no support for the theory that the Dogmatism Scale predicts classroom learning" (p. 76). These findings are modified, however, by his state- ment that "aptitude and dogmatism are independent." Adams and Vidulich (1962), using an N. of 36 undergraduates at Louisiana State University, found high dogmatic subjects inferior to low dog- matic subjects in a paired association task. Kaplin and Singer (1963) found a significance level of .05 for the inverse relationship between dogmatism and sensory discrimination tasks. Oldfild (1964) established in her sample of college students that dogmatism does not influence choice of social -48- preferences, however, "high dogmatic subjects (significant at .005) could most readily be induced by the instructor to change their opinions about least accepted group members" (p. 2979). Lefcourt (1962) used an N. of 272 drug addicts, divided into neurotic, psychotic, and character disorder groups, to investigate the relationship of dogmatism and readiness for therapy through potential for change. Dogmatism and potential for change in therapy varied inversely. Because the gogmatism Scale is clearly being employed in a variety of new settings with divergent types of subjects, there exists a need to conduct investigations which might help to integrate and relate the construct dogmatism to a broader and more usable body of knowledge. CHAPTER III ME‘DIONLOGY W The present investigation was conceived as an experimental attempt to apply the frustration- aggression theory of Dollard and Miller (1939) to the behavior of two specific groups of delinquent boys confined in state training schools in Michigan and Minnesota. The investigation was designed with three rather specific purposes in mind. First, an attqnpt was made to examine the possible relation- ship of frustration arising from negative school experiences , as evidenced by reading under-achieve- ment, to patterns of aggressive behavior. Secondly, the relationships among reading, intelligence, school attitudes, and dogmatism were studied. Lastly, background data concerning enviromnental and behaviorism factors in the most aggressive members of each experimental group were compared to similar data for the population groups as a whole. The above design was adopted with the aware- ness that the behavior of delinquent boys is very -49.. -50- complex. No single theory has yet been advanced which satisfactorily accounts for all behavioral variables in delinquency. However, there is con- siderable evidence to indicate that school related variables, especially those associated with frustra- tion for the learner, have been significantly correlated with aggressive forms of delinquency in boys, B. Subjects Because delinquent boys who are confined in state training schools provide accessible popula- tions for the investigation of aggressive behavior together with the correlaries of such behavior, and because considerable recorded dataare'typically available concerning such boys, they present a logical source of subjects for research. Advantages of sampling two pepulations. The decision to sample two populations of delinquent boys was made for a number of reasons. It has never been established that two separate populations of delinquent boys confind in unrelated state training schools would exhibit similar characteristics. Furthermore, the behavior patterns of individual -51- boys within a specific training school are charac- terized by extreme heterogeneity. The Lansing (Michigan) and Red Wing (Minnesota) Boys Training Schools are approximately 700 miles apart, and so represent considerable geographic separation. Michigan is a heavily industrial and urban state, while Minnesota is essentially agricultural. The two boys training schools are of approximately the same size and maintain similar administrative‘educational practices. Nrthermre, the advantage of comparing two population:neasures of aggressive behavior, reading, school attitudes, dogmatism, and related danographic data seemed considerable. Figure one stmarizes population data for the two institutions, It should be noted that Lansing's population exceeded that of Red filing by only 10‘. The white population of Red Wing was considerably greater than that of Lansing and also contained a much.higher percentage of Indians. Negroes compromised 49 percent of Lansing's popula- tion and only eight percent of Red wing' s. The average length of confinanent at Lansing was approximately one-third longer than at Red Wing. -52- FIGURE I TOTAL POPULATION DATA LANSING AND RED WING BOYS TRAINING SCHOOLS Measure Lansing Red Wing Total Population 350 340 Percentage white 50 81 Percentage Negro 49 8 Percentage Other 1* 11* Mean Length 0 Confinement 8.4 months 6.2 months Mean Chronological * 15.7 (82-1.21) 16.4 (82-141) Mean Intelligence Quotient ** 95 ($2.03) 100 (32-144) *For population breakdown by ethnic groups, the one percent figure at Lansing is largely Mexican, and 331%“ percent figure at Red ‘Wing is largely ‘ “he mean difference bemen populations for C.A. is significant beyond .001, and the mean difference between populations for IQ is significant beyond .025. -53- The chronological age difference of the older Red Wing population was significant beyond the .001 level. The Red wing sample was also slightly more intelligent, with mean 1Q dif- ferences between groups significant beyond .025. We Th. two sample populations used in this investigation were selected at random from the total populations of the Lansing and Red Wing Boys Training Schools. Forty-eight boys were selected from each institu- tion, making a combined N. of 96. This figure represented approximately 15 percent of the total populations of the two training schools and was sufficiently large to permit an assumption of randomness in the distribution of data for statisti- cal analysis. As has been previously stated, certain subject variables were beyond the scope of this paper. line subjects from the Michigan sample, and six from the Minnesota sample, were excluded for one or more of the following reasons: 1) mental retardation, because individuals whose Vechsler lQ's were below 80 might be subject to the complex influences of retardation on personality, attitude formation, school achievement, and ability to -54- comprehend the instruments employed: 2) absence of both parents in the home, because the influences of foster home placement are not well understood, and show considerable individual variation in their relationship to personality development, and to overt behavior in schools and other social settings; 3) organic brain damage, because neurological influences on behavior vary from subtle to profound extremes, and may be negatively influenced in a testing situation; 4) psychosis, because valid responses to the testing instruments could not be assumed in the presence of inadequate reality testing, or unusual delusional systems. Exclusions for the above reasons were made only after a careful examination of documents contained in each subjects case history file was completed. W. The random selection of subjects for this study was accomplished in the following manner: with eyes closed a table of random numbers (Randreorporation, 1955) was entered separately for each of the two institutions. A coin was then flipped to determine whether rows or columns should be followed. The first 75 three digit numbers between one and the institutions total N. of boys were recorded. Each training school -55- registry (files in record room) was then entered, and from an alphabetical listing of all boys in the institution, the names whose location in the listing matched the first 48 random numbers were recorded and thus became the experimental group. he re- maining 27 names (from the original 75 random numbers) were used to replace boys who were truant from the school, too ill to be tested, on parole, or who were excluded for reasons described above. Table one smarizes data pertinent to the composition of the two experimental groups, It should be noted that the Red Wing sample was slightly older’and had come to the court's attention at a somewhat later date than had the Lansing subjects. Neither the mean lQ's nor their standard deviations differed dramatically between groups. The distribution of Negro and white subjects within the Lansing sample was approximately equal. Only six Negroes appeared in the Red Wing sample. Indian and Mexican subjects represented only six of the combined N. of 96 subjects. W The selection of instruments for this study reflected the need to accurately obtain measures of SUMMARY OF SAMPLE POPULATION -55- TABIE I DATA Heasure Lansing N-48 Re 6. Wi ng N348 Distribution by Race White Negro Indian Mexican Mean Chronological Age Mean Intelligence Quotient Mean Age at First Conviction 24 22 O 2 1508 (SD-lelg Ran5e=12.lO-l7.7) 95 ($938.54 Range-BO-lll) 12e8 (SD=2e1, Rangem7.3-l7.l) 38 6 4 O 1602 (SD-1.0, Range-14.0-18.0) lOO (30:12.0 Range-80-l23) 13.7 (SD=1.6, Range=9e 1'17s6) -57- intelligence', reading, dogmatism, school attitudes, and amount of aggressive behavior. For reasons of reliability and validity it was most feasible to adopt well known instruments that have been success- fully employed by other workers in related research. With the exception of aggressive behavior for which, no suitable objective measuring technique is cur- rently available, it was possible to satisfy the above criterion. be aggressive behavior variable was evaluated by three separate measures which will be described below. W6 The mwmam perhaps the best known and most widely used instruments for the determination of human intelligence at the present time. Wechsler chose to abandon the mental age concept, as such, in favor of a tabular method for determining intelligence based on standard deviations from the normal curve. His validity and reliability data are fully described (Wechsler, 1958). Verbal, performance, and full scale intelli- gence quotients are derived from these scales. be m (Wechsler Intelligence Scale for mildren) is employed with chronological ages up to 15 years 11 months. The M (Nechsler Adult Intelligence Scale) is used with subjects whose chronological ages are 16 years and older. It was necessary to unploy both scales in the present study because of the overlapping chronological ages represented in the samples. The reliability and validity of the above instruunts presupposed their use by a person with special training, usually a psychometrist or psychologist. This requirement, as will be later described, was met in the current study. For this study the reading section of the Intermediate Battery, Form K., of the W W 1955 (Kelley, Madden, Gardner, Terman, and. Ruch, 1953) was administered to all subjects. Detained validity and reliability data for this instrument are provided with each packet of tests, and is also available from the publisher, World Book Company. his reading section of this instrument consists of two sub-tests, one a 48 item test of paragraph meaning, and the other a 48 item test of word meaning. The two sub-tests each contain reading grade level nouns to which raw scores are converted. A third score, that of average reading, is obtained by finding the arithmetic mean of the -59- combined grade level equivalents for the paragraph and word meaning sub-tests. Directions proceed each test and time is not a factor with this instru- ment. Because this instrument reliably measures reading skills from the primary through the secondary grade levels it was found to be readily usable with a population of delinquent boys. W. This instrument, which has gone through five editions, and for which validity and reliability data are available for a number of groups (Rokeach, 1960), consists of Forms D. and E. Form D. containing all of the final 66 items of the complete scale was adopted for this study. The W m and instruc- tions to the subject are found in Appendix B. It should be noted that the subject is required to respond to each of the 66 statment itmns in one of the following ways: +1: I AGREE A LITTLE; +2 I AGREE ON THE WHOLE; +3: I AGREE VERY MUCH; I DISAGREE A LITTLE; -2: I DISAGREE ON THE WHOLE; -3: I DISAGREE VERY MUCH. Scoring the U P so Dogmatism Scale consists of finding the algebraic sum for all ituns. Because this is the only instrument which has been demonstrated to measure the dogmatic factor of open-closed mindedness, and because this -60.. instrument has not been used with delinquent boys, its adoption is believed to be justified in the present study. Numerous instruments for the measurement of opinions, beliefs, and attitudes, have been described in the psychological and sociological literature. However, the selection of a valid and reliably school attitude instrument that would be usable with a population of delinquent boys showing a considerable span in reading and intellectual skills was a difficult task. Such an instrument has been deve10ped by Flanders, while working at the University of Minnesota. Its standardization has been described (Flanders, 1960) in a report of a Cooperative Project Grant from the Department of Health, Education, and warm- me mm mm mm W consists of 62 statements which are each to be answered in one of the following five ways: SD: STRONGLY DISAGREE; D: DISAGREE; U: UNDECIDED; A: AGREE; SA: STRONGLY AGREE. Scoring is according to a formula provided in the key, and magnitude of the total score reflects the degree of positive school attitude measured. 1m: scale's statements refer to -61- a specific classroom or school situation. Since the subjects in the current study were assigned to a variety of academic and vocational situations it was necessary to conclude the instrument's printed directions with the following statement: What I want you to do when marking each statement is to think of the teacher and classroom or shop you have here at training school that is most like your idea of the usual teachers and classes you have had in the past. Appendix 0. consists of the W m W and instructions to the student. satisfactory instrument meeting acceptable standards of validity and reliability has yet been shown to measure aggressive behavior in delinquent boys, three separate measurements were employed in this study. Each measurement received a separate statistical treatment. Measurement one consisted of a careful analysis of the individual case history reports maintained within the training school at which each boy was confined. Such histories included court reports of previous illegal acts, social case work reports, police statements made by the subjects about their past behavior, and reports from school -52- and comunity sources. The reports also contained information about behavior while under confinement at the training school. Through access to the training school file room it was possible to tabulate behaviorism information about each subject for later statistical analysis. Measurement two was derived by providing each training school senior psychologist, director of social services, director of home life (housing etc.) services, and the director of educational ' services, with an alphabetical listing of all subjects selected from his school. These pro- fessional staff members who were all in close con- tact with the schools total population were then asked to list in rank order as many boys as possible, using the following instructions: Your cooperation is requested for the successful completion of an experi- mental research project; being gflcud at ya tra schooI. It is m important that the experimentor know which of the boys listed below are regarded by you as the most mm and W toward either 592].“ or the other QEE. This hostility may be expressed e er ysically (fighting) or verbally threats). The experimentor is interested in non-aggressive del quent behavior such as unauthorized smoking, truancy, stealing, homosexual acts, etc. -63.. From these rankings it was possible to tabulate behavioral information for the second statistical analysis of aggression. Measurement three consisted of a self report obtained in individual interviews with each subject. Following the establishment of rapport, which did not prove to be an unmanageable task, responses to the W W m illustrated in Figure II were obtained. By tabulating this inter- view data it was possible to do a third statistical analysis of aggressive behavior. Iii—W Cooperation with the writer was excellent on the part of the superintendents and staffs of both training schools. No serious problems were en- countered in either the securing or housing of subjects for testing. Access to record files and other necessary data was readily extended. W6 1110 reading. school attitudes, and dogmatism instruments were ad- ministered in small group settings consisting of five subjects per group. Meetings in the group provided the first contact by the subjects with -64.. FIGURE II VOLUNTARX STATEMENT OF AGGRESSIVE BEHAVIOR BEHAVIOR” Threats directed toward peers: Fighting with peers: Threats directed toward adults: Minor malicious preperty destruction: School or home incorrigibility: Aggressive gang or group behavior: Physical assault on parent or other adult figure: Attempted rape (not statutory): Sadistic forms of injury to others: Arson or major malicious prOperty destruction (purposeful): Felonious assault with the intent to do serious bodily harm: Rape (not statutory): Assault with a deadly weapon: *Note: Illustrations of each type of behavior were provided by the examiner. Subjects were asked only if they had participated in the Specific behavior described, and if so th item was checked. -55- the examiner. All groups were seated at tables located in a quiet room of the main administration building at each school. Precautions in seating were taken to Space subjects beyond visual access to answer sheets of other subjects. When the grOUps had assembled, the examiner introduced himself, and said that he would like to read the following statement: I have asked you here today to help me with an eXperiment I am doing for Michigan State University. I want to know more about the attitudes and Opinions of boys like yourself to a number of statements about your school eXperiences, and the world you live in. I have no connection with this school, and no one here at the school will ever know what you have written down or told me. There is no way that this information can in- fluence your personal life in any way. Are there any questions you wish to ask me? Following questions about the study, its use, the writer's purpose in doing the study, why and how subjects were selected etc., the dogmatism, school attitude, and reading instruments were administered, in that order. Because reading ability is a significant factor in testing, each statement on the dogmatism and school attitude instruments was read aloud twice by the examiner while the subjects were -66.. reading the same statemmts at their tables .' No new statement was read until the previous statement had been answered by all subjects. Naturally no assistance beyond directions and clarification was given during the reading testing. W. Intelligence testing and the self-reports of aggressive behavior were done on an individual basis. Subjects for these procedures were called to the psychologist's office where adequate interview controls were easily maintained. All needed intelligence testing was done by the writer who is certified by the State Board of Examiners of Psychologists in Minnesota as both a Practicing and School Psychologist; All sub- Jects who did not have a W IQ recorded in their school files were tested with that instrument. Subjects whose M were more than 36 months old were retested. It was necessary to do 19 intelligence tests with the Michigan population, and 31 tests with the Minnesota population. Before the self-report of aggressive behavior interview began) each subject was carefully reassured that none of the information given would ever be used in any way to his disadvantage. It -57- ‘was further explained that most boys, including the interviewer when he was a boy, have done many of the things being discussed. It was also made clear that the interviewer did not want to know the names, dates, or specific details of the events being asked. It was only necessary to know if the subject had or’had not engaged in the specific behavior being described. ‘All interviews ‘were conducted by the writer and in no case was it not possible to complete the interview form. ‘ugghgg_gg_gnllzlig, All demographic data were collected by the writer following a careful reading of the case files. Appropriate information for the analysis of demographic factors was imediately transferred to record fauna and stored in the possession of the writer’until data ‘were available for all subjects. Similarly, all testing was done by'the writer. The writer twice repeated each step in the tabulation of demographic data, and in the scoring of test instruments, as a precaution against error factors. In consultation with educational research specialists at both Michigan State and the University of Minnesota, the appropriate statistics for'the analysis of data were determined. It was -53- decided that the most usable statistic for computing large sample relationships of the subject variables under investigation in this paper would be the W 53131; 231g; ngfigieng 91 correlation, corrected for ties (Edwards, 1960). This statistic which has the following form: = .. 4C2: DL”) R l [ INCNN) provides a convenient means of ranking subjects who display behavior on non-discrete population variables, such as aggressive behavior, and whidi are not readily available for analysis by other means of either central tendency or of variability. Further- more, Hays (1965), has described a t-test for signi- ficance level determination with two ranked variables. This test has the following form: .t: mm with N minus two degrees'o-f Yf-reaedom, and is "satisfactory for N's larger than 10." The chi-square test for two independent samples (Siegel, 1956) is a well known non- parametric technique for analyzing data for -59- two independent samples of unequal size. This statistic, which has the following form: was found to be readily usable for comparing demographic factors present in the most aggressive 1/3 of each training school sample, with similar factors present in the remaining 2/3'3 of each sample. The decision to make comparisons on a 1/3 — 2/3 ratio was arrived at because it was feared that the middle one—third of subjects might dilute the results, and because the scores on aggression seemed to break naturally into approximately a 1/3 - 2/3 ratio. Comparisons of demographic factors were made for aggression and the following variables: 1. income source; 2. parents present in the home; 3. place of residence; 4. birth order; 5. race; 6. age at first court contact; and 7. highest school grade completed. A one tailed test at the five percent level (.05) was used for all determinations of signifi- 081108. CHAPTER IV RESULTS AND DISCUSSION The present investigation has sought to ex- amine the relationships among aggressive behavior, reading, and dogmatism in two specific groups of delinquent boys. It has further sought to examine certain attitudes about school, and the relation- ship of certain specific demographic factors which might be associated with aggressive behavior in the above groups. This investigation has taken the theoretical position that early school experiences of failure, especially in reading, might predispose a child to significant levels of frustration. Such frustra- tion might, in the absence of aggression-inhibiting social controls, be expressed in various forms of anti-social aggressive behavior. It has further been theorized that feelings of failure and frustra- tion might influence attitudes about school, as well as certain personality characteristics described by Rokeach (1960) as dogmatism. Finally, it has been viewed as possible that aggressive behavior might -70- also relate to certain demographic factors present in the individual's environment. Delinquent boys confined in two state training schools were selected as subjects to examine the above theoretical positions. The decision to use such boys was based on the probable assumption that they have demonstrated lower than normal levels of inhibitions toward aggressive behavior. The two groups of 48 boys each were randomly selected from.state training schools in Michigan and Minnesota. In order to examine the theoretical position described above, the following research hypotheses were formulated: 1. There would exist a significant correlation between aggressive be- havior and reading. 2. There would exist a signficant correlation between aggressive behavior and dogmatism. 3. There would exist a significant correlation between aggressive behavior and school attitudes. 4. There would exist a significant cor- relation between aggressive behavior and intelligence. 5. 7. -72- There would exist a significant cor- relation between reading and either dogmatism, school attitudes, or intelligence. There would exist a significant cor- relation between dogmatism and either school attitudes or intelligence. There would exist a significant cor- relation between school attitudes and intelligence. Aggressive behavior would be related to one or more of the following demographic factors at a significant level: parental income, number or sex of parents present in the home, place of residence, birth order, race, age at time of first court contact, or highest school grade com- pleted. Hypotheses one through seven were tested using the Spearman Rank order Correlation (Hays, 1965). Hypothesis eight was tested using the Chi-square test for two independent samples (Siegel, 1956). All significance levels were set at .05 (one tailed). -73- Three measures of reading achievement were derived by grade level equivalence. The first measure indicated paragraph meaning or reading comprehension. The second measure indicated word meaning or vocabulary development. The third measure was that of average reading, and was de- rived by obtaining the arithmetic mean of the first two measures. The third or average reading score was used for statistical treatments. . The results of individual intelligence testing were derived in the usual Wechsler style of verbal, performance, and full scale IQ' s. School attitude scores were derived by magnitude of total score, with high scores indicat- ing positive or favorable attitudes about school experiences- m W W mm: W has been included as Appendix C. Reading achievement ratios were derived by the mental age method previously explained. Since actual grade reading level was divided by expected grade reading level, high ratios indicate greater reading W- In summary, it should be noted that a rank of one was assigned for the most reported aggressive -74- behavior, greatest reading under-achievanent, most dogmatism, least favorable school attitudes, and lowest full scale IQ. All subsequent rank order correlation findings were derived from these rank- ings. Ties in ranking positions were corrected using a method described by Hayes (1965), in which the arithmetic mean of the tied rankings were divided by the number of tied scores, with the quotient so derived becoming the rank order position of all tied scores. The test results and rank order of subjects can be found in Appendix D. A. RESULTS As was previously indicated, the. absence of a suitable objective instrument for measuring aggressive behavior necessitated the employment of three separate measures of this variable in the present investigation. Table two and Appendix E sunmarize the Spearman rank-order correlations between reading and the three measures of agree- sive behavior. -75- Based upon the statistical analysis, it was found that the court report measure of aggressive behavior correlated significantly with reading in both sample populations. It was thus possible to accept hypothesis number one for the correlations between reading and accounts of aggressive behavior reported in the case folder of each subject. It was not possible to accept hypothesis number one fdr'the correlations between reading and the self report measure of aggressive behavior. These correlations did not reach significance in either sample papulation. The correlations between reading and the staff report measure of aggressive behavior reached significance only in the Lansing population. Thus it was not possible to accept hypothesis number one as it applied to staff reported aggressive behavior. In summary, it was found that significant correlations existed between reading underachieve- ment and acts of aggressive behavior performed by the delinquent boys who constituted the two sample populations. However, these correlations were ob- served only for the court report measure of aggressive behavior. -75- TABIE II CORREIATIONS BETWEEN READING AND AGGRESSIVE BEHAVIOR BOTH GROUPS* Measure r t—value p Lansing court report data .35 2.58 .025 Lansing self report data .18 1.24 NS Lansing staff report data .35 1.89 .005 Red Wing court report data .40 2.94 .005 Red Wing self report data .13 .92 NS Red Wing staff report data .02 .08 NS *A rank order of one was assigned for the greatest reading underachievement and the most aggressive behavior observed. -77- Table three and Appendix E summarize the.Spear- man rank-order correlations between dogmatism and the three measures of aggressive behavior. Based upon the statistical analysis, it was found that none of the correlations in either group reached sig- nificance. It was not possible, therefore, to accept hypothesis number two. It was observed, however, that the correla- tion of .25 between dogmagism.and the staff reported measure of aggressive behavior in the Lansing sub- jects did approach significance. No relationship whatever was found in that group between.dogmatism and court reported aggressive behavior, and only a very slight relationship was observed for the self report measure. In the Red Wing subjects the correlation between dogmatism.and the staff reported measure of aggressive behavior was also the highest of the three measures. In that group, however, none of the correlations approached signifi- cance. It might have been possible to suggest that direct observations of the subjects by their train- ing school supervisors was the better method of -78- TABIE III CORREIATIONS BETWEEN DOGKATISH AKD AGGTVSSIVE BEHAVIOR BOTH GROUPS“ Measure r t-value p Lansing court report data .00 .00 KS lensing self report dara .02 .14 KS Lansing staff report data .25 1.31 NS Red Wing court report data .04 .27 NS Red Wing self report data .11 .76 KS Red Wing staff report data .19 .77 NS *A rank order of one was assigned for the highest dogmatism score and the most aggressive behavior observed. -79- detecting those aspects of the dogmatic personality that correlate with aggressive behavior. This possibility, however, could not have been confirmed with the current data. Hypothesis 3, There would exist a significant rre t on between ressi e be via d ch 'gtgitudgg. The Spearman rank-order correlations between school attitudes and the three measures of aggressive behavior have been summarized in Table four and Appendix E. Following the statistical analysis of data, it was observed that no one specific measure of aggressive behavior had reached significance for both populations. It was not possible, therefore, to accept hypothesis number three. Unlike the findings for the first two hypotheses, no directional effect was observed with respect to type of aggressive behavior measured across popula- tions. While one of the three measures of aggressive behavior did reaeh significance in each sample popula- tion, the two significant correlations were not for- the same specific measure. Furthermore, in the Lansing sample, court reported aggressive behavior approached the level of significance, while in the Red Wing -80... TABIE IV CORRELATIONS BETWEEN NEGATIVE SCHOOI.ATTITUDES AND AGGRESSIVE BEHAVIOR BOTH GROUPS“ Measure r t-value p Lansing court report data .18 1.24 NS Lansing self report data .28 1.97 .05 Lansing staff report data .10 .69 NS Red Wing court report data .07 .48 NS Red Wing self report data .14 .99 38 Red Wing staff report data .28 1.97 .05 *A rank order of one was assigned for the most negative school attitudes and the most aggressive behavior observed. -31- population this measure produced the lowest of the three correlations. It might have been possible to suggest that the heterogeneity of agreement between papulations was due to the difficulties encountered in attempting to quantify two variables as intangible as aggressive behavior and school attitudes. It was also very interesting to observe that court reports of aggressive behavior did not correlate significantly with school attitudes for either group, while the correlations for the other two somewhat more subjective measures of aggressive behavior produced mixed findings. The possibility of relating the above findings to factors inherent in the geographic separation of the two schools should also be considered. Wess- The findings for the statistical analysis of data between the correlations for intelligence and the three measures of aggressive behavior~have been summarized in Table five and Appendix E. Because intelligence was not observed to correlate at a -82- TABLE V CORPS IAT I ON S BET WEE N I NTE III GEN CE AN D AGGRE SSIVE BEHAVIOR BOTH GROUPS* Measure r t—value p Lansing court report data .28 1.97 .05 Lansing self report data .05 .34 NS Lansing staff report data .35 2.89 .005 Red Wing court report data .12 .84 NS Red Wing self report data .10 .69 NS Red Wing staff report data .11 .76 NS *A rank order of one was full scale intelligence score behavior observed. assigned for the lowest and the most aggressive —83- significant level with any of the three measures of aggressive behavior in the Red Wing population, it was not possible to accept hypothesis number four. The inconsistencies in findings across pOpula- tions further suggests that the two subject groups were dissimilar. In the Lansing subjects low intelligence would seem to be closely associated with objective accounts of aggressive behavior. However, it was not possible to note this observation in the Red Wing group, for whom none of the correlations even approached significance. Hypothesis 5. There would exist a significant correlation between reading and either dogmatism, school attitudes, or intelligence. Table six and Appendix E summarize the Spearman rank-order correlations between reading and the separate variables of dogmatism, school attitudes, and intelligence. Based upon the statistical analysis, it was observed that the correlations be- tween reading and intelligence were highly signifi- cant, making possible the acceptance of hypothesis number five for both groups. No other correlations involving reading and the above variables were found to be significant across populations. -84- TABLE VI CORREIATIONS BETWEEN READING AND DOGMATISM, READING AND NEGATIVE SCHOOL ATTITUDES, AND READING AND INTELLIGENCE BOTH GROUPS“ Measure r t-value P Lansing reading and dogmatism data Lansing reading and school attitudes data lensing reading and intelligence data Red Wing reading and dogmatism Red Wing reading and school attitude data Red Wing reading and intelligence data .12 .31 .71 1.67 2.21 4.95 NS NS .005 NS .025 .001 *A rank order of one was assigned for the greatest reading underachievement, highest dogmatism score, most negative school attitudes, and lowest full scale IQ observed. -35- The dissimilarity in correlations across sample populations was again observed. Except for the expected relationship between reading and intelligence, no other correlations were signifi- cant in the Lansing subjects. However, in the Red Wing subjects both school attitudes and intelligence correlated significantly with reading. In addition, the correlation betwaen dogmatism and reading approached significance. The Spearman rank-order correlations between dogmatism, school attitudes, and intelligence, have been summarized in Table seven and in Appendix E. Based upon the statistical analysis for the combined populations, it was not possible to accept hypothesis number six. No correlations among these variables reached significance in the Red Wing subjects. However, the correlation between dognatism and intelligence approached significance. In the Lansing subjects, the correlation between dogmatism and school attitudes was significant beyond .05. The observation of -86.. dissimilarity between sample groups was again supported. w ld n i d 8 2221115222. The findings for the statistical analysis of data between school attitudes and intelligence have also been summarized in Table seven and in Appendix E. Because school attitudes were not observed to correlate with intelligence at a significant level across populations, it was not possible to accept hypothesis number seven. In the Red Wing subjects, the correlation of.lS between school attitudes and intelligence did not approach significance. However, in the Lansing subjects the correlation of .81 for the same variables was the highest noted among any of the variables investigated. -87- TABLE VII CORREIATIONS BETWEEN DOGMATISM, NEGATIVE SCHOOL ATTITUDES, AHD INTEIIIGEECE BOTH GROUPS“ Measure r t-value p Lansing dogmatism and school attitudes .26 1.89 .05 Lansing dogmatism and intelli- gence .09 .62 NS Lansing school attitudes and intelligence .81 8.32 .001 Red Wing dogmatism and school attitudes .08 .54 NS Red Wing dogmatism and intelli- gence .19 1.36 NS Red Wing school attitudes and intelligence .15 1.15 KS *A rank order of one was assigned for the highest dogmatism score, most negative school attitudes, and lowest full scale intelligence score observed. -88- birth ogdeg, ggce, age at time of first court h t s r d com 1 ted. The relationship of the above factors to aggressive behavior was tested by the chi-square test for two independent samples (Siegel, 1956). By using the court report measure of aggressive behavior, which was found to be the most stable of the three measures, it was possible to compare the most aggressive one-third of each sample population to the remaining sample for the above demographic factors. The 1/3-2/3 split for comparisons of aggression and the demographic factors was adopted for two reasons. It was feared that the middle one-third of subjects might dilute the results of the comparisons. It was also ob- served that the ranking of subjects for aggressive behavior seemed to break naturally into approximately a l/3-2/3 ratio. Tables eight through fourteen present the findings from.the analysis of demo- graphic factors. It should be noted that none of the chi-squares reached significance, although non- significant relationships differed somewhat between variables. -89- TABLE VIII Aggressive Behavior and Parents Income Source ____Lms.ma Realises Most Others Most Others €I73) (2/3) Total €I§3> (2/3) Total Employed 8 17 25 8 15 23 Others 8: l5 23 _§_, L7 25 Total 16 32 48 16 32 48 Analysis of data yielded a chi-square of approximately .001 for the Lansing sample and approximately .190 for the Red Wing sample. Obviously both values were substantially below the 3.84 required for significance (.05 level). The employed parents were found to be working primarily at unskilled and semiskilled occupations. The unemployed parents were receiving various forms of public welfare assistance with‘A.D.C. funds predominating. -90- TABLE IX Aggressive Behavior and Number of Parents in the Home miss Miss Most Others Most Others €§§3> (2/3) Total €123) (2/3) Total Both Parents 7 12 19 8 15 23 Others 9 20 29 8 17 25 Total 16 32 48 16 32 48 Analysis of data for the relationship between aggression and number of parents in the home also revealed chi-squares which fell far short of significance. The value .011 was observed for both the Lansing and Red Wing samples. It should of course be noted that while number of parents present in the home was not significantly related to aggression, the incidence of broken homes in both the more and less aggressive subjects was very high (see Appendix.A-l). -91.. TABLE X Aggressive Behavior and Place of Residence lasting fi__£E&—Red w Most Others Most Others €73) (2/3) Total €23) <2/3) Total Detroit or T.C. 10 12 22 12 21 33 Others 6 20 26 4 11 15 Total 16 32 48 16 32 48 While neither chi-square for aggressive behavior and place of residence reached significance, the 1.09 for the Lansing subjects was considerably larger than the $109 value for the Red Wing population. It should also be noted that 55 of the total N cf 96 subjects lived in either Detroit or the Twin Cities. This pr0portion of urban residence is much larger than that observed in the general pOpulations of the two states. -92... TABLE XI Aggressive Behavior and Birth Order Lans Red Wing Most Others Most Others ?I73) (2/3) Total ?I§3) (2/3) Total First or Last 3 ll 14 10 14 24 Others 13 21 34 6 18 24 Total 16 32 48 16 32 48 The chi-square of 1.742, although not significant, was much larger for the relation- ship between aggression and birth order in the Lansing subjects, than was the .844 value observed in the Red Wing population. This finding contributed to the numerous observations of dissbmilarity between the two sample papula- tions. -93- TABLE XII Aggressive Behavior and Race Lg£§1§g__ Red Wing Most Others Most Others ?§§3) (2/3) Total ?§§3) (2/3) Total White 6 18 24 12 27 39 Others .19. .14. as. .9. .2. .2. Total 16 32 48 16 32 48 Appendix A91 summarizes the distribution of both sample populations by race. The percentagewof Negro subjects was much higher in the Lansing subjects. However, neither that groups chi-square of .805 or the Red Wing chi-square of .158 even approached significance. TABLE XIII Aggressive Behavior and Age at First court Contact ngsing Red Wing Most Others Most Others €I§3) (2/3) Total €I§3) (2/3) Tbtal Under Age 12 5 10 15 l 6 7 Others 11 22 33‘ 1: zg 41 Total 16 32 48 16 32 48 The-chi-square for aggression and age at first court contact was extremely small in the Lansing subjects (.002). The chi-square of .522 observed for this relationship in the Red Wing subjectaswas larger, but obviously far from significant. The data of first court contact can be found for all subjects in Appendix.Arl. -95- TAELE XIV Aggressive Behavior and Highest School Grade completed mm W6 Most Others Most Others, €I73) (2/3) Total €I§3) (2/3) Total Through 8th 12 19 31 4 ll 15 Others .5. .13. .12. 42. _2_l_ .22. Total 16 32 48 16 32 48 Thirty-one of the 48 Lansing subjects had formal educations which terminated in the eighth grade or earlier. Fifteen of the Red Wing subjects had reached only this level. However, the chi-square values for the relationship between aggression andischool grade completed failed to reach significance. While the chi-square of 2.865 for the Red Wing group approached significance, the .555 figure for the Lansing group indicated a minimal relationship. On these variables also, the groups were dissimilar. -95- W- The test performance of the Lansing and Red Wing subjects was found to follow a highly dissimilar or heterogeneous pattern in distribution of correlations among the variables under investigation. This observation also held for the chi-square analysis of demographic factors. In summary, the following results were obtained; 1. 2. 3. While significant agreement between all three measures of aggressive behavior was not obtained for either population, the court report measure of aggressive behavior was found to correlate with reading beyond the .05 level in.bg§h,groups. Dogmatism.was not observed to correlate significantly with any of the three measures of aggressive behavior for either of the two sample populations. The staff report measure of aggressive behavior was observed to correlate beyond the .025 level with school attitudes in the Red Wing group, and beyond the .05 level (self report measure) in the Lansing subjects. 4. 5. 6. 7. -97.. While aggressive behavior did not correlate significantly with IQ in the Red Wing subjects, this variable was observed to correlate beyond the .05 level for both the self and staff report measures in the Lansing sample population. Reading was found to correlate signifi- cantly with both aggressive behavior and IQ in m groups. Reading also correlated significantly with school attitudes in the Red Wing subjects. Dogmatism was not significantly correlated with either school atti- tudes or intelligence in the Red Wing subjects. It was correlated beyond .05 with school attitudes in the Lansing group. School attitudes were correlated with IQ at a highly significant level (.001) in the Lansing subjects, but failed to reach significance in the Red Wing sample population. -98- 8. In the chi-square analysis of demo- graphic factors, none of the factors examined reached significance. How- ever, aggressive boys in larger than expected numbers tended to come from urban centers, and in one or the other groups.to have had unemployed parents, to have been nonwhite, and to have had meager formal educations. B. DISCUSSION Acts of aggressive behavior involving delinquent adolescents are commonly agreed to constitute a significant social problem. Little research into the possible school related factors associated with this behavior has been done. Al- though many studies have noted academic under- achievement, especially in reading, to be a fre- quent finding among delinquent youth, most writers have viewed this as merely an additional indication of maladjustment and inadequate coping behavior. The present study has found court report information about aggressive delinquent acts to -99- correlate significantly with under-achievement in reading. This finding was observed in two sample populations of delinquent boys, widely separated geographically. Although two other measures of aggressive behavior did not produce significant correlations with reading under- achievement across populations, such correlations were observed to reach significance within sample groups. It might therefore be said that students who are experiencing frustration in school because of ineffective reading, especially those from high delinquency urban areas, should be provided with intensive individualized remedial measures while still in the elementary school. It might also be said, based on the literature previously cited in this investigation, that state training schools might profitably devote more attention to remedial reading as an adjunct to their existing programs. D ss r t the sub e t rou . In addition to the significant relationship between reading under-achievement and aggressive behavior, perhaps the most important finding derived from this investigation was that two populations of -100- delinquent boys separated geographically exhibited very heterogeneous characteristics. Table 15 has compared the test performance of the Lansing and Red Wing subjects for reading, dogmatism, school attitudes, and IQ. The heterogeneous nature of the two sample populations has been dramatically illustrated with t-test differences in each of the four test means significant beyond the .05 level. The Red Wing subjects were better readers, exhibited less dogmatism.(.001), and had higher full scale IQ's (.025). These significant differences in test per- formances constitute an important variable which was not hypothesized in the design of the present investigation, but which had been suggested to the writer in 1965 by Dr. John L. Johnson, now of Syracuse University. It is difficult to conceptual- ize how cultural, racial, and socio-economic factors present in the environments of the two sample populations might have been sufficient to produce the heterogeneity observed. One needs to ask why the two sample populations should differ three grade levels on reading, 49 points on dogmatism, 18 points on school attitudes, and five points on IQ. Put LP —.._--. -Ma -101- .n-OH oases Sam mo saucy 5 one nouoom oowowdzou-Hm one .33 awn-mag Scum estrus—u mead; ouaaoubd no make» a.“ one mooaauue Hoe-Lou amalgam-m on“. Sou.»- ootfiuou amnesty condone-w no mason 5" one nouooa Seduce—woo rage." ooonw no canon s..- ouo mouooa washroom mmo. mm.~ o.~a cos an.” o.mo ooaowaaaouta so. mn.~ oo.m~ mm.oas N~.N~ mm.mma aoosuauu< Hootum boo. oo.e mn.om oo.as oe.oe ma.oa sonatawoa Boo. ma.e me.~ mm.w mo.a mn.n magenta not: a osao>lu mm was: com um coo: menace; ousaoqz E mmDQmU HUMhmDm ho WHHM0m ensue n03 3 «.3 Ken n o o." xuoz .Couosh H05: nun asausom no 3 . 33: 3 . m." :m n n a 3 9330: £8 a»: 3.53 83 z 3.3 Some N o 0.3303 w0suoz nun 38909 S 3 z .3 . n." s.— Ann n o goose: ugh-oz run seasons: «o m.— 3 o .3 an on n o o xuoz beacon noon vouch—95 uaouuon Home 3 3 .3 n mom n a s 255?: «0&3: sue.— uaouu0n no a.— 3 n .3 on R n e o Monsoon. u0nush nun 333m on an 3 n .3 .3 mm n N n 83:03 H050: one 3890.— 33 z effi— shown o n 0.3303 not»: nun usausom no a." 3 o .3 n .— con N n 005.353 noon 90.5.5.5 9.“ chin no a." z .— . h.— 3 con m 0 use: .huouosh neon sue uaowu0n no a." z 3 .3 3mm n e a nonsense". 58 vacuums: 38qu 33 3 «.3 363 m, .— o aaoauuozu noon sum 2360.8. .uz no a.— 3 h . n.— on R n u a .— o . “5. sodusum soon too.— uaouu0n no a.— z o . 3 e on n n . n n 0.3303 90590: rue 32.909 .3 m.— z o . 3 n a an n n e easy 303 “.0932 snowman nachos no a.— z n .3 au R. n N m sadness: spun fie .38 use 83 3 3.3 33m a o 0.3303 soon «insane: udouusn no a." z n .3 «an n agenda ooh—sow 06oz ua0a0usam 00903 s0¢ 2300..— 001m 02 M0352 ueouo no 08003 5 ocean no .33 assoc .3 ocean If nub—«ml woos: season 350m used 9.0.“ Hus” wade—"sum A098» 652 mwa "up—Hag 4H3 20H Ramon Hag AI< Nun—Em: -130- WEIF-II& n 03 030030: anon nun 0000a030u «003 z n.03 nonnn n N 0003303 moans: 0000300: 0300003 3003 3 5.03 on30n 3 n 303003 anon nuo3 003unom 0003 a n.03 3nnnn m n 030030: 00:00: 303 nouns anon 0n03 z 0.03 mnenn 3 n 34 000.32 .3050: fin .326 033033 303 z 0.3 33.3 n 33 3302 3300003 anon nun 0033n m003 z n3n3 33oo0 3 3 0000000004 0000 nuo3 0030003 n003 z n.03 nannn n n 03003:: anon nun 0000aa303 nn03 z n.n3 annnn n a 030030: 0003 :00 0300034. n003 3 3.33 3300n n n uoxusua anon nun 30333. 3003 s 3.03 003nn 0 n 3003 muououn 30000: 000 0033m n003 z o3.03 noooe n n 03003:: 0003 nun 3003000 0003 2 03.03 n00nn 0 n 000330: 00:00: 300 0000a0303 0003 a c3.n3 n3o0n n 3 .3000 .>03u 30:00: 000 03ouuon «003 z n.n3 03nnn n 0 0003303 00:00: :00 0303303 nn03 z 3.03 3nnnn 0 0 30030 30300: 0000000: 030300n 0003 2 m.n3 msonn n n 03000303 00:00: 000 03onuon m003 z o3.~3 nnmnn n 0 030030: 00300: 303 00330030: n0o3 z p.03 0330n 3 n 3003 nuouuum anon 3033 osmon 00>3m 0003 z 0.53 ncnnn a n 030030: 30:00: :30 a0>0m 00000 nmo3 z 0.03 30o0n n 0 0003303 00000: 300 0030003 n003 a 0.n3 nnncn n 0 300: 5000003 upon :00 00300: 0003 a n.03 03o0n 0 n 0303303 00000: 0000000: 3300000 0003 3 0.m3 N3nnn 3 o sue: 3000003 0003 :00 u3oguoa 0003 z 33.n3 onnnn n n 000030: 0303 005 03:0»33 n003 z n.n3 neonn 0 0 0303303 nonugx 030 3300003 n003 z n.03 nonnn .HO—vho ugflndm 830m 080m ”guidm Oogvdnflm 380m 001% 02 MODE—.32 5.330 30 00.803 03 000.6 30 .3030 0.38 383300 909332 nugflm uuflflhflm an: UOOdHa magnum... 2.3» .. $33.3 L. -131- o o.” 09.30: Hanna: 5a «0.— uhon: 3o.— 3 5.3 duunu." a .— 55 23m .3502 5m hogan... 3 a.— 3 s .3 4003." .n N 2353: "050: 5.....— ufiomuoqfiz $3 3 h. D 403.3 m e 93.8930 58 Ava agenda; .33 3 oéa dome: m h 3.88 uofium flog «3.58:5: 32 z .33 3:3 a m 88303 uofiox £3 «388.3: 32 3 .23 .333 .— N 23:03 go: duo." «.3 09353: 33 z .n 6." <83.— e o 33083: mafia: nun Ramadan com.— 33 ~93 <33." a H 2.833 uofium fig .883» 82 z m.D 332 e n “.353qu 58 duo.— huao amuse 83 3 0. ma 4303 n n 835:5 gonna: 5m 0030 So." 3 3.3 4.33.. n m 045300: 58 duo." «3093532 Ra.“ a 3. a.— 4:3." a N 30568 38a anon sue.— adaonuog $0.— 3 m6." 303.." m e 9.3303 audio: £00 #933 $3 65” n. n.” (mam: e m non—0nd.” anon duo." «333:5: mom.— 2 o. o.— <39: H e House—do anon duo nouuonoom .33 3 s. n.” 4.33.— m. N 83:03 .350: 5m agenda; ova.— z «.3 4933 .— m 3.3.8 53 duo." agenda; moo.— 3 o. n.— <33." m 5 gas 53 5a 32.3 $3 3 3.3 433." n e 903039 fiom 5w «$38.; moo." z n. 3 4&9: H m 8.303 uofio: 5m «3.53:5: 33 a e. D <29: Hoouo «Edna» ooh—now 060m ugouHm 00903on vacuum 003“ and Mona—.5 nun—«m no 3805 ovuuo no nude £300 doonum uonfisz uuaouum «vacuum :31! 903mm wade-mm Aswan 3554mm. mwom 02H: OHM <93 zanagmom an; «:4 “35“qu - 132- N 3 88¢... £8 fi: 8:333 $3 3 3.: <33: 3 3 385.3 58 53 333cm 33.. 3 N.3 3N3.— .. m .3030 3.33: 38 £3.33 93 .33 $3 3 N.3 2 .1._ use: .2. Reading 15274.25 .18 1.24 NS Dogmatism. 18181.75 .02 .14 NS School Attitude 13318.25 .28 1.97 .05 IQ 17665.50 .05 .34 NS Red Wing Subjects N819 2 . 4AEE:!3§A¥§-filhfl!12£LflEfi5 .—JE— .Jfih. 5:!II2§H_R— Reading 16203.50 .13 .92 NS SchOOI Att1EUGe 15985.50 .14 .99 NS IQ ' 16587.00 .10 .69 NS -157- APPENDIX E (continued) commons OF AGGRESSIVE BEHAVIOR STAFF REPORT DATA BOTH GROUPS 1| Lansing Subjects N828 2 . WW: .2. .2... 21:19.: .9. Reading 3355.25 .09 .46 NS Dogmatism 4602.00 - . 25 1. 31 NS School Attitude 4033.00 -.10 .50 NS IQ 28030.50 .35 1.89 .05 Red Wing Subjects N819 . 2 new: .9. .1. 21.12. _2.. Reading 1165.50 -.02 .08 NS Dogmatism 926.00 . 19 . 77 NS - 158- APPENDIX.E (continued) READING CORRELATIONS BOTH GROUPS .- 1 Lansing Subjects N: 2 . W: __L .s. 5:51.912: .2. School Attitude 19543.00 - . 06 . 41 NS IQ 11527.50 .38 2.77 .005 Red Wing SubJects N248 2 . W: .2. .2. 2131s: .2. Dogmatism 14040. 75 .24 1. 67 NS School Attitude 12880.00 .31 2.21 . 025 IQ 5451.00 .71 4.95 .001 -159- APPENDIX E (continued) IQ CORRELATIONS BOTH GROUPS Lansing Subjects N348 19.29.: .23.. .1... 21.1.9. .2. Reading 11527.50 .38 2.77 .005 Dogmatism 20170.50 -.09 . 62 NS School Attitude 21191. 50 . 81 8. 32 . 001 Red Wing Subjects N348 152.929.: .22.. .1... -v u .2. Reading 5451.00 .71 4.95 .001 Dogmatism 14880. 25 . 19 1. 36 NS School Attitude 15705.00 . 15 1. 15 NS - 160- APPENDIX E (continued) DOGMATISMIAND SCHOOL ATTITUDE CORRELATIONS BOTH GROUPS Lansing Subjects N348 2 . .Essastiaa.asg: ..£L._. .1... 3:23122H_2.. School.Attitude 13707.75 .25 1.89 .05 10 20170.50 -.09 .62 us fishesl.fi££issss.ss§; 10 21191.50 .81 8.32 .001 Red Wing Subjects N348 2 . .Eanstisa.assfi ..£L... ..s.. .5:29129.;E_. School.Attitude 17044.50 .08 .54 ns 10 14880.25 .19 1.36 88 S e de : IQ 15705.00 .15 1.15 NS -161- APPENDIX E (continued) commons MDNG THE THREE MEASURES OF AGGRESSIVE BEHAVIOR BOTH GROUPS Lansing Subjects 2 Essays: J: .2. .s. 21:12: 2 Court to Staff Reports 28 1278.75 .75 4.75 .001 Self to Staff Reports 28 3605.00 .02 .10 NS Red Wing Subjects .1: _.93 312131932 Court to Self Reports 48 20529.00 -.11 .76 NS Court to Staff Reports 19 1080.50 .06 .25 NS Se1f to Staff Reports 19 842.50 .26 1.83 .05 mm: REFERENCES Aaron, 1. E. Comparisons of good and poor readers in the fourth and e th grades ,1. mg, am, 1960, 54, 34.3. ’ Abrams, J. C. A study of certain personality characteristics of non-readers and achieving readers- W. 1956‘. 16. 377-378. Adams’, H. E. Dogmatism and belief congruence in paired-associate learning. W, 1962,10:(1), 91-94. Adorno‘, J. W.‘, e um .‘ New York: ey an ns c., 1 . - Alson, E. Cognative structure and dogmatism. W: 19595 203 365'366‘ American Psychological Association, W W. Washington. a ;; e A. e . knee, L. B. 6: Walker, R. N. Prediction of later reading ability from kindergarten rorschach and 1.Q. scores. ,55: (6), 1964, 309- 313. Anderson, C. C. A developmental study of dogmatism during adolescence with reference to sex differences. We.» Arbuckle‘, D. 6: Litwack, L. A study of recidivism amng juvenile delinquents. Egg“; 2223235128.. 1960. 24: (4). 45- Bakwin’ H. Causes of juvenile delinquence. “r. ’W 1955. 89. 368-373. Ball, J. C. A scale and factorial analysis of delinquent attitudes. W. 1955, 15, 2328-2329. -162- - 163- Balogh', J. K. Juvenile delinquency proneness: a study of the predictive factors involved in delinquent phenomena. W W-. 1958‘. 48. 61 - 1 . Balow‘, B. Long term effects of remedial reading instruction. W, 1965,18, 581-586. Bandura, A. 5: Walters, R. Dependency conflicts in aggressive delinquents. W 1958, 14: (3), 52-66. ’ Bandura, A. & Walters, R. escen r as : d of -, _:uence of 1d n t ' d .H ' ‘_ :37‘_;_ ‘en: .':0 New York: Ronald Press, 1 . Baraheni, M. N. An inquiry into attitudinal concomitants of success and failure at school. M, 5:(1), 1962, 63-67. Berkowitz, P. 6. Rothman, E. W. New York: N. Y. U. Press, 1 1. Berks, H. 1". Brain pathology. W, 1957, 24, 169- 174. Bernabeu, E. P. Underlying ego mechanisms in delin ency. PW 1953 27 383- 338. ’ ’ ’ Bettelheim, B. W. New York: he Free Press, 1 . Betta, E. A. Are retarded readers "dumb?" W, 1956,76, 569-575. Bills, R. E. Nondirective play therapy with retarded readers- 11.922511154519921... 1959, 14, 140-149. " - . -om Housel ‘- a l a o O ' O a a e " C I C I I s s O C a . o e e r 9 , e I a a q ' O O o e C ' C e ) a e D C a o I a ' . . n O O O a 8 I I s -164- Bolmeier, G. A. The relationship of dogmatism in parents to various aspects of adjustment among high school students. W M, 1966, 26:(9), 5571-55 . Bowman, P. H. Effects of a revised school program on potential delinquents. W W. 1959. 322. - - Bruck, M. &Bodwin, F. The relationship between self-concept and the presence and the absence of scholastic under-achievement. J, .113, Pszgh2;., 1982, 18:(2), 181-182. Brunkan, R. & Shen, F. Personality characteristics of ineffective, effective, and efficient readers. W, 1966, 44, 837-843. Byrne, D., Blaylock, B. 6: Goldberg, J. Dogmatism and defense mechanisms. W” 1966, 18:(3), 739-742. Caplan, M. S. A.comparative analysis of average and superior I. Q. delinquents. Unpublished doctoral dissertation, western Reserve University, 1961. Carroll, J. B. The nature of data, or how to choose a correlational coefficient. W. 1961, 26, 347-372. Carter, H. D.’ Over and under achievers in California. 311;, J, gas, 123,, 1984, 15:(4), 175-183. Chandler, T. A. Reading disability and socio- economic status. W, 1966, 10, 5-21. Chiland C. Eches scolaires au cours preparatoire (academic difficulties in the preparatory course>- WW. 1964. 95(1), 15- O Christensen, C. M. A note on dogmatism and learning. W. 1963. 66:(1). 75-76- Clark, J. 5: Winningen, E. The attitudes of Juveniles toward the legal institution. W W. 1 . 1 . -165- Cochran W. G. 6: Cox, G. Experigtggtgé design (3g, 2. New York: iley, l . Cbhen,.A. K. Sociological research in juvenile delinquency. 9ggr, g, Qgthgpszghgag. 1957, 27, 81- . 001m, Jo Ce, Barnston, Fe Le & FOX, Jo Parental attitudes as related to reading disability in children. M, 1958, 4, 47-51. (Zomba. A. W- 6: Snyss. D6 W6 New York: Harper 5: c., l . Copple, L. B. Mater deve10pment and self-concept as correlates of reading achievement. W“. 1961. 22:(4). 1241. Crescimbeni, J. Broken homes affect academic achievement. W. 1964, 84:(7), 433-441. Critchley. Macdonald . We Springfield, 11 .: rles mas and Sons, 1964. Curry, R. 5. Hughes, H. Relationship between measured and anticipated achievement in reading. W, 1961, 32, 91-96. Davidson, H. 5. Lang, G. Children's perceptions of their teachers' feelings toward than related to self-perception, school achievement, and behavior, M, 1960, 29, 10 118. Dentler, R. 8. Monroe, L. Social correlates of early adolescent theft. W, 1961, 26, 733-743. Dentler, R. A. Dropouts, automation, and the 2751283' Teach. col. Rec., 1964, 65:(6), DeVos, G. 6: Mizushima, K. The school and delinquency: perspective from Japan. W. 1962, 63:(8), 828-838. - 166- Dexter, L..A. 4.. 2 111' o; :u u°_ -;: -1 -¢- .° -1 - -3; -ro-,:u 0' s - - t~. ‘ew ‘or : Baa c :o-ks, l‘- . Dimitz, S. & Kay, B. Group gradients in delinquency poetential and achievement scores of sixth radars- W. 1958. 28. 98-605. Dinwiddie, F. W. .n :9. 1'2 cn of s_e -r - - -f res--ns~ ;- era- .st on ,- he d s 1 :- W- Washington. D. C.: thol c University ress, 1955. Dollard. 3- 5‘ Miller. No W- New Haven: Yale Univers ty Press, 1 . Darrell, D. D. Success in first grade reading. J. Educ., 1958,58,1-48. Edwards ’ A. Edwards, D. L. Reading from the child's point of view. §l5m1_§pg., 1958, 35, 239-241. Edwards, W., Lindman, H. & Savage, L. Bayesian statistical inferencem psycholo ical research. W, 70, 196 , 193-242. hrlich H. J. Dogmatism and learning. W 522.122.1511.. 1961. 62. 1.3-149. Fink, M. B. Se1f-concept as it relates to academic underachievement- 91W. 1962,13:(2), 57-62. Flanders, N. A. 5ress, 1960. Gates,.A. I. Eailure in reading and social malad «mm W. 1936. 25, 2 5-206. - 167- Gersten, C. An experimental evaluation of group therapy with juvenile delinquents. W. 1951. 1. 311- Gibbs', D." N.‘ Student failure and social maladjust- ggtgssl’erson. G: guid. J., 1965, 43(:6), smack; & smock. War... New York: Harper ros., . Glueck, 6: Glueck, Working mothers and delinquency. H , 1957, 41, 327-362. Gnagey, W. J. Do our schools prevent of promote (Zieéinquencfl W, 1956,50, 215- l Gottfried, N. W. Psychological needs and verbally expressed. aggression in adolescent delinquent 33%;. Dissertation Abstr'.‘, 1959, 19, 3352- Granzow, K. R. A comparative study of under achievers , and over achievers in reading. Dissertation Abstr.', 1954, 14, 631-632. Greenblatt, H. J. I hate reading. W Mn... 1955.1. 8-14- Gregory, R. E. Unsettledness, maladjustment, and reading failure: a vill e study. fig}. ,1" W,1965,353%1), 63-68. Harlem Project. W ' ‘1' Q5: _-1.'. ” 5. -_d 118 'LlL-H 1d d: ”I c s e" ’ : s - mm s ew or ‘ty: -e New York Foundation and the N.Y. C. Board of Education, 1945. Harris. 1. D. W 1.. York: The Free ress, . Harris, 1'. L. Sumnery and review of investigations relating to reading July '64-June '65. W, 1966, 59, 256-258. v 1 Q I I O e 0 0 ID 0 e e . ~ O I a s O s O I O . e o D 0 e . ' ‘ e o a o 0 Q t ' A i D 4' O Q C 5 0‘ . I O I a U . v a u . t 4 O O . - 168- Harvey, J. B. Type of influence, magnitude of discrepancy, and degree of dogmatism as determinants of conformity behavior. W. 1963‘. 24: (2). Havinghurst, R. J. Poor reading and delinquency may go hand in hand. W, 1959, 64, 55-58. Haworth, M. R. (1%.), Wm. New York: “31C BOOkS has“, 0 Hays. W. L3 W6 Holt. Rinehart, ston, . Healy, A. K. Effects of changing children's attitudes toward reading. W, 1965, 42, 269-272. Henderson, E. H. Sol-social constructs of achieving and non-achievin readers, W, 1965, 19,114-11. Hinkelman, E. A. Relationship of reading ability to elementary school achievement. M m, 1956,42, 65-67. Hunmel, R. 6: Sprinthall, N. Underachievement related to interests, attitudes, and values. 23259.9... w, 1965, 44, 388-395. Illovsky, J. An experience with group hypnosis in reading disability in primary behavior disorders. WW, 1963,102: (1)’ 61-67. Jackson, P. 5: Getzels, J. Psychological health and classroom functioning: a study of dissatis- faction with school amo adolescents. m. m, 1959, 50, 295- 00. Johnson, G. O. 8: Stanley, J. C. Attitudes toward authority of delinquents and non-delin uent boys- W: 195.51. 712- 716. . e I O I n o O I o O s Q o O O s . . e e u a I I I 0 l . o . I O 0 e e O , C e s o . Q n D u I O O u 0 O C v e . I O n C Q U D Q I O O O n O V O O I - 169- Kajimura', Y. Agpsychiateric study of the Juvenile delinquents. W6 1958, 7, 38-63. Kaplin, M. F. & Singer, E. Dogmatism and sensory alination: an empirical investigation. W. 1963. 27=<6). 486- Kelley, 'l'., Madden, R., Gardner, E., Tom, L., and KICh’ G; 0 New York: r1 Boo c., . Kmp’, C. G. Influence of dogmatism on counseling. W. 1961‘. 39'. 662-665. Kendall. M4 6. W11 London: r fin, . Keshian', J. G. Is there a personality pattern common Eggsgggessful readers, W., 1962, 39, Kirk. 8. 5: Welner. B. W W3 ashlnston. D- c.: -‘l‘he Council for ceptional Children, 1963. Knafle', J. D.‘ Personality characteristics, social adjustment, and reading effectiveness in low achieving college freshman in a reading program. W, 1965, 59, 149-153. Knobloch, P. Rorschach investigations of the reading recess. W 1965, 33, 277- 82. Krippner, 8. Relationship between reading improve- ment and ten selected variables. Ger-seam; W. 1964. 19:(1)'. 15-2 . Krippner, 8. Reading improvement and scores on the Holtzman inkblot technique. m W. 1966, 19, 512-522. mi: Yo ers-on- son, ew or: World ok, 1945. 1 n I ' I o e s . A a . a s 1 Q C ; o a e o q 1 e - I o G i v C l - e c 1 s s o e a a e I O o .. . . . O a p e a O a I‘ O s C 9 s O 0 u a ‘ -170- Lay, '1'. Language facilitation among delin uents: a pilot study- W. 1 65.15. 21-225. Lefcourt‘, H. M. Clinical correlates of do tism. W. 1962‘. l8: (3). 32 -328- Lindquist, E. £1 ,1 Lively, E., Dimitz, S. 5: Reckless, W. Golf concept as a predictor of delin uency. W. 1962. 3 :(l). 153-155 Lloyd, R. G. The relationship between academic achievement of pupils and the social structure of the class room. W, 1956, 21, 179-180. Long, N. J., Morse, W. & Newman, R. W the classroom. Belmont, Ca1if.: rth Publishing, 1965 . McMurray, J. G. Some correlates of reading diffi- culty in satisfactory and disabled readers: a preliminary study in grade three. am, 1963, 5: 2), 149-157: McNally J. Delinquency and the schools Educ, R35, 1965,77:(3), 212-214. ’ - Magee, R. D. Correlates of aggressive defiant behavior in elementary school bo s. Dissertation Abstr., 1964, 25:(2, 1340. Meier, M. H. Dogmatism related to attitudes toward adolescence as an institution. W m, 1960, 20, 2890. Meier, N. Frustration theory: restatement and extension. W, 1956,63, 370-388. London: :_ HerMajesty s Stationary Office, 1955. a e O a I c g - e a C O a a r q a I s s a . ‘ . a I D ‘ C C Q . O O . O O O D C O I . I 9 r 9 O C O C U Q C l I I .‘ . a 0 . ' I 1 a . -171- Margolin, J., Roman, M3 & Harari, C. Reading dis- ability in the delinquent child: a microcosm of psychological patholo y. J r s , 1955,25: 1), Miller. Na Es MW- New Haven: Yale versity ress, l 1. Miller. M & Dollard. J- W IEDE2Efl2EflflHaLLJELJEEHDEEGLJELJLJEEHfifiE learning. New York: McGraw-Hill, 1 . ‘Miller, N. Involvement and dogmatism.as inhibitors of attitude thense- W... 1965,1:(2), 121-132. Moore, D. N. Dogmatism and anxiety in the conditioning of verbal behavior. W 22: <7). 1962. 2468. Mukherjeek, . & Kundu, S. .A study of birth order and family'positions of the criminal. Indian J, pszggg;,,1961, 36: (3), 127-132. Narayana, R. S. A.study of the sense of responsi- bility and it's relationship to academic achievement. Psychol. Studies, 1964, 9: (2), 109-118. Natchez, G. Personali .atterns and oral readin;: : <1 iIIIIEIEESYENEFIB1IFEZIQZTT ; stioa {I :ve s :e t_-ns - d:- ;.d.u aggression, angw Ehdgfwg; in chiidggg, New York: New Yor Un versity ress, 1959. National Institute for Educational Research (Japan). Comparative study of delinquent normals of the secondary school pupils at Keihin area. WW. 1959. 13. Newman, H. Reading habits, attitudes, and achieve- ments of high school drOpouts. §§5Q_figig§§, 1966, 48, 36-48. Nurburger, Howard M. .A case of reading disability. Nt b e s ', 1955, 1, 15-19. s I a a I s e s o O C o O o , a . a a a . q I a O a a e t 0 a l a - . . 0 g a a . .1 C O V o g u r a a C I I ' A -172- Nye. F. 14 W W. New York: John Wiley, 1 . Oldfield, E. J. The relative influences of dogmatism and situational praise on variability in social reference. W. 1964, 24: < ), 2978-2979. Ort, L. L. Reading difficulties, a contributing factor to underachievement and failure in school. W, 1962, 28, 489-492. Pannes, E. D. The relationship between self- acceptance and dogmatism in junior-senior high school students. W, 36:(9), 1963, 419-426. Peck, H. B. A new pattern of mental health services in children's court: round table 1954. W. 1955‘. 251(1). 1-50. Petrie, A., McCulloch, R. & Kazdin, P. The perceptual characteristics of juvenile delinquents. W. 1962.5 .415-421. Person, G.‘, Barton, V. 6: Hey, G. S.M.A.T. motivation factors as predictors of academic achievement of delinquent boys. PM, 1956,63, 370-388. Pine, G. J. Occupational and educational asperations and delinquent behavior. W, 1964-65, 13: (2), 107-111. Porterfield, O. 6: Schlichting, H. Peer status and reading achievement. W, 1961, 54, 291-297. Quay, H. 6: Blumen, L. Dimensions of delin ent 3713337717.; J. soc. Psychol., 1963, 1:(21), 1 . ~173- Rabinovitch, R. 8: Ingram, W. Neuropsychiateric considerations in reading retardation. W., 1962, 15, 433-438. Raygor', A. L. College reading improvement and personality chanso- W8. 1959, 6, 211-217. Rebhun, M. Dogmatism and test anxiety. ML... 1966, 62: (1), 39-40. Richardson, H. 5: Surko, E. W.1. S. C. scores and status in reading and arithmetic of delinquent children. WM, 1956, 251-262. Robbins, M. P. Study of the validity of Delacato's theory of neurological organization. m. m, 1,966 32, 517-523. Robeck, M. C. Effects of prolonged reading dis- ability: aper liminarys mm;- W motor Skills., 1964, 19: (l), 12. Roberts, A. H. 8: Herrmann, R. S. Dogmatism, time perspective and enemies W. 1960, 16, 67-72. Robinson, H. M. W, micago: University 0 cago ress, l . Rokeach, M. 5: Fruchter, B. A factorial study of dogmatism and related concepts. W. 2.229221.” 1956', 53, 356-360. Rokeach', M. W. New York: Basic Boo s, . Roman, M.’ Tutorial group therapy: a study of the integration of remedial reading and group therapy in the treatment of delin ints. W. 1955. 15. 17 13 Roman, M. Margolin, J. 8: Harrari, C. Reading retardation and delin ency. W W. 19 5. l. l- 1 O O O O O t a a e 1 x v g . . _ a e , s a . 9 . . o - C i t I . O c . v 0 O f I F I . C 9 U C l r I I I D a Q o A O . V S I e I O O I d ’ o - a i . Q Ir - 174- Roswell, F. 8: Natchez, G. W, siasnsaistgaaaggfsasssst- New Yo = Basic Boo , . Roucek, J. 8. (Ed.) W, New York: Philosophical L rary, 1 . Saunders, R. E. Dyslexia, more than reading retardation. ‘, 1965, 11: (3), 137-14 . Segal, B. E. Racial perspective and attitudes among Negro and white delinquent boys: an gmpggical examination. _m, 1966, 27: (1), Shaw, M. C. & White, D. L. The relationship between child-parent identification and academic underachievement. W, 1965, 21: (1), 10-13. Shaw, M. C. & Grubb, J. Hostility and able high school underachieverh W. 1958, 5, 263-266. Shaw, M. C. 5: McCuen, J. The onset of academic underachievement in bright children. 9, edug, Psychol., 1960,51, 103- 109. Shimota, H. E. Reading skills in emotionally dis- turbed, institutionalized, adolescents. W. 9 1964,58,106-111. Siegel, S. 9"3...-’.LL3 a. 111.11": '- . ‘- b :.' --;_. W. New ork: Graw-‘ 1, ' -. Solomon, R. H. Personality adjustment to reading 31199983 and failure. We 9 1953, 77, 64-82. Sopis, J. The relationship of self image as a reader to reading achievement. aheadsmi_£hs£322_flsazt- 1965-66, 1: (2), 94-101. - 175- Sorenson, H. A longitudinal study of the relation- ship between various child behavior ratings and success in primary reading. Unpublished doctoral dissertation, University of Minnesota, 1950. Spache, George, Personality patterns of retarded readers, W, 1957, 50, 461-469. Staats, A. W. 5: Butterfield, W. H. Treatment of non-reading in a culturally deprived juvenile delinquent: an application to reinforcement principles. W, 1965, 36, 925-942. Stevens, 8. S. New York: Sticht, '1‘. G. 6: Fox, W. Geographical mobility, dogmatism anxiety y. and a8e1 mm. 1966, 68: (1), 171-174. Stone, F. 6: Rowley, V. Educational disabilities in emotionally disturbed children, M, 1964, 30, 423-426. Tabarlet, B. E. Poor readers and mental health. M, 1958', 35, 522-525. The Rand Corporation, . u n di; _ s w .1 H '0' 1:11- .—_,_. New York: eFree ress, 1° . Tuller, D. & Eames, T. H. Electroencephalograms of children who fail in reading. m. 9311g,, 1966,32, 637. Tutto, D. N. Maladjustment among adolescents, 1 "' 110 Wagenheim, L. First memories of accidents and read- 1118 difficulties. WM: 1960, 30, 191-195. - 176- Wattenberg‘, W. & Clifford, C. Relationship of self-concept to beginning achievement in reading. W., 1964,35, 461- Weaver, C. H. Semantic distance between students and teachers and it's effect upon learning. WELL. 1959. 26. 273-2811 Wechsler, David 1 ,1 11-.:: 31.3% Late “gage. ‘ rBlt 11- re _ : 1 - - f am and lkens Company, 1958. Wheway, J. P. Intelligence and delinquency. W119”. 2. 208-2141 d . New York: The Conunonwealth” Md, 1 8. Wilson, R. C. 5: Morrow, W. R. School and career adjustment of bright high-achievers and under-achievers in high school boys MW. 1962.101. 91-1531 Winer, B. J. Wirt, R. D. 6: Briggs, P. 1". Personality and environmental factors in the development of delinqueney1 We... 1959. 73, (whole # 485) Wrenn, R. L. The effects of preparatory involvement on goal valuation in open and close belief systems‘1 W11962.23=(2)1 735. Zegona, S. 5: Zurchor, L. The relationship of verbal ability and other cognative variables to the open-closed cognative dimension. Zimmerman, I. L. 5: Allebrand, J. Personality characteristics and attitudes toward achievement of good and poor readers. 11.99.151.393... 1965, 59, 28-30. -177- Zolik, E. S. A comparison of the Bender gestalt reproductions of delinguents and non-delinquents. J. Clin.gpsychol., 195 , 14, 24-26.