A COMPARISON OF THE DIDACTIC INTERACTIONS OF MOTHERS AND FATHERS. WITH THEIR PRESCHOOL CHILDREN ‘ Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY IO LYNN CUNNINGHAM 1972 W“I”WI‘IIIIIWHWW L M m ‘3 m . h: I? I >ll 4 :‘ I .'. Ix'r . l‘frl I‘VE I t! . 1".l IIIIIIIIIIIII‘NI Ila. \Q ABSTRACT A COMPARISON OF THE DIDACTIC INTERACTIONS OF MOTHERS AND FATHERS WITH THEIR PRESCHOOL CHILDREN BY Jo Lynn Cunningham The basic objective of this study was to compare the interactions of mothers and those of fathers with their preschool children. An un- structured dyadic teaching situation was used in which each parent was paired individually with his preschool child. More specifically, this study sought to identify potential simi— larities and/or differences between mothers and fathers on the follow- ing aspects of their interactions with their preschool children: communication process, verbal communication, nonverbal communication, and teaching procedure. A secondary intent was to investigate these comparisons in relation to socioeconomic group, race, and sex of the child. A completely crossed and balanced 23 design was used, with inde— pendent variables of race (black and white), socioeconomic group (lower and middle), and sex of child (male and female). A stratified random sample of thirty-two families was selected from a larger representative sample in order to achieve an orthogonal design. The unit of analysis was thus the family, while the dependent variables of interest were Jo Lynn Cunningham the scores of the mother and the father as they interacted with the sampled child. Only subjects having complete data were selected, and only one child from a family was eligible for inclusion in the sample. In addi- tion, the following criteria were maintained: (a) Both parents were living in the home with the child at the time of assessment. However, either natural parents or step-parents were accepted for inclusion. (b) Both parents were of the same racial group as the sampled child. ’ (c) The age of the child was between 2 years 10 months and 3 years 9 months inclusive at the time of assessment with each parent. (d) The child had less than one term's preschool experience at the time of assessment with each parent. Data were collected using a situation which involved having each parent teach the child a simple two-dimensional sorting task. A toy sorting task was used with the mothers and a nine-block sorting task with the fathers. The sessions in which each parent attempted to teach the child the designated task were videotaped and then rated using a structured observational rating procedure. The specific variables of interest were identified as follows: (a) Communication process--initiation, response, response object, reinforcement, feedback, and reward; (b) Verbal communication--verbalization, verbal fantasy, voice tone, verbal specificity, time orientation, task orientation, and nature of interference; (c) Nonverbal communication--affective tone, anxiety, level of Jo Lynn Cunningham involvement, and physical behavior; (d) Teaching procedure--concept specification and teaching method. A three-way multivariate analysis of covariance was carried out for each of the four sets of variables. The dependent variables were the contrast scores obtained by subtracting the father's score from the mother's score for each of the variables of interest; covariates were length of the session and age of the child. Scores for the qualitative variables were obtained by computing logit transformations of relative prOportions for both dichotomous and polychotomous cases in order to stabilize the variances of extreme scores. Results indicated that there were both similarities and differ— ences between mothers and fathers in their interactions with their preschool children in the dyadic teaching situation. Differences were concentrated in the areas of nonverbal communication and teaching procedure, with marginal differences in communication process. Simi- larities occurred primarily in the area of verbal communication. The differences in teaching procedure were a function of an interaction between race and sex of the child, while the differences in nonverbal communication and the trends in communication process were a function of socioeconomic group. In addition, a number of individual variables reflected potential differences between parents. A general comparison between mothers and fathers indicated that, in contrast to fathers, mothers tended to be relatively more verbal. In addition, the nature of this verbalization differed in several ways. Mothers used greater specificity than did fathers in their verbaliza- tion, and they used relatively more verbalization dealing with tension management. Fathers, on the other hand, were more decision-oriented Jo Lynn Cunningham and more evaluation-oriented. In addition, their non—task verbaliza- tion dealt with matters less closely related to the immediate situation. There was a tendency for interaction sequences in the teaching situation to be somewhat shorter with mothers than was the case with fathers, although mothers tended to continue interaction by initiation of a new sequence more frequently than did fathers. The interactions of the mothers also had a relatively low frequency of cases in which there was no immediately preceding communication, which indicated that the child less frequently terminated the interaction sequence than with fathers. Teaching procedure used by parents differed both in kinds of concept specification used and also in teaching method. In general, mothers tended to introduce new concepts more often than did fathers, while fathers' teaching behavior was characterized by relatively more use of general orientation, use of new approaches, and use of demonstration as a teaching method. A COMPARISON OF THE DIDACTIC INTERACTIONS OF MOTHERS AND FATHERS WITH THEIR PRESCHOOL CHILDREN BY Jo Lynn Cunningham A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Family and Child Sciences 1972 © Copyright by JO LYNN CUNNINGHAM 1973 DEDICATION To a very spec1al pair of parents-- Helen Yarbrough, who for many years was both mother and father to me, and Dick Yarbrough, the father whom I finally had the unique privilege of helping to choose. ii ACKNOWLEDGMENTS "No man is an island, entire of itself . . ." --John Donne, For Whom the Bell Tolls. And no dissertation is completed without the assistance of untold numbers of friends, colleagues, and associates. Like many islands, however, the greatest mass is hidden from view, with the written report representing only the final peak of a much vaster graduate program. My entire graduate program has been rich both in experiences and in friendships. While not all can be mentioned individually, some of those individuals whose contributions to this project are particularly obvious deserve special recognition. CHEERS AND THANKS . . . First and foremost, To Robert P. Boger, guidance committee chairman for my entire graduate program and supervisor for the graduate assistantship under which I was able to collect the data for this project. His tolerance, encourage- ment, guidance, harrassment, good humor, and faithful support were effective catalysts for the many invaluable opportunities he provided for varied experiences, professional growth, intellectual challenges, and most importantly, the chance to be mg_throughout it all. To the other members of my doctoral guidance committee-— --Margaret Jacobson Bubolz, who was also department chairman, con- tributed to my enrichment through both theoretical content and a variety of opportunities for professional involvement beyond the classroom. --Don Melcer provided invaluable guidance through his calm, in- sightful, and rational perspective, turning many frustrating events into learning experiences that ultimately strengthened my professional development. iii --Ellen Strommen provided the stimulation and encouragement which helped me to develop both knowledge and confidence at a critical time in my program. Her continued support presented a model for the highest academic standards. --William H. Schmidt was officially added to my guidance committee near the end of my program, but his contributions were significant (and_ meaningful). His statistical expertise, coupled with a convincing be- havioral demonstration of patience (exemplified by periodic reminders in my moments of discouragement that trying new statistical techniques is indeed fun), proved helpful on both qualitative and quantitative dimensions. To a number of individuals who were not official members of my guidance committee but who played particularly important roles in my graduate program and the development of this research-- --Andrew Porter, by giving me the opportunity to work as his graduate assistant and exposing me to a broader view of research and varied methodological problems, had a substantial influence on my doctoral program and thus indirectly on the subsequent completion of this research. --Others for whom I had the privilege of working and learning as a graduate assistant included Cole Brembeck, Eileen Earhart, Hyman Rodman, and Irving Sigel. Through discussions and various other activi- ties, they provided many of the best experiences of my graduate program. --Beatrice Paolucci provided a stimulating challenge throughout my graduate program. She also read the first draft of this dissertation and offered helpful comments. --Jere Brophy also reacted to the first draft of this research, as well as to the development of the ideas for this study from their very initial stages. To staff members of the Institute for Family and Child Study at Michigan State University who assisted with the collection and processing of the data used for this study-- —-As an assistant and friend, Marilyn Olson helped in innumerable ways throughout this study. In particular, her capable and conscien- tious assistance with rating of the videotapes, supervision of the other raters, and quality checking of all the coding contributed significantly to the quality of the data--and consequently to my mental health. --Others who assisted with the tape rating were Kellery Dempsey, Jan Gillings, Gina Green (who won the endurance record for length of service), Kathy Hynes (who doubled as artist in residence while cinching the "Boob Tube Award" for greatest total amount of parent-child tape rating), and Deborah Pappas. Their perseverance in watching the local research reruns was matched only by the sensitivity with which they were able to implement an admittedly complex procedure. --Responsibility for the videotaping of the sessions was assumed by the media unit of the Institute. Special credit goes to coordinator Jim Respress for his ability to calm the parents, charm the children, create esprit de corps among the psychometric crew, and coax the creaky equipment into professional-level production. --Collection of the parent-child data was facilitated by the competence of psychometrist Sharon Evans. iv --Equa1 skill at the other end of the production line was shown by Jean Lundeen, who coded all the data with record speed and accuracy, and Gayle Swanbeck, who keypunched and verified the majority of the data cards. To other Institute staff members who helped in innumerable ways with the completion of this study-- --Verda Scheifley gave both pep talks and strategy assistance in my programming battles with the computer. --Pixie Vose performed a variety of tasks with characteristic cheerfulness and efficiency--and without ever issuing a reminder that running a wakeup service was never mentioned as a condition of her employment. --Cynthia Zinn shared her typewriter while I typed--and typed-- and typed. --The interest and support of many other staff members during this entire study provided encouragement that was both personally and pro- fessionally gratifying. We shared many happy and frustrating times, and from them I learned many things during my graduate program. To the Michigan State University Computer Laboratory staff, particu- larly Dean Franklin and his midnight shift. Their patient assistance, trust, and cooperation in allowing me to spend many nights running my own data made possible the accomplishment of an otherwise impossible task. To the many students and faculty members with whom I have consulted on their research and in spite of whom I finally finished my own. Working with them provided important evidence of the value of learning by doing, for as they did, I learned; the task of completing this dissertation was greatly facilitated by their experience. To the many friends who provided the moral support that is essential to the completion of any individual's graduate program-- --Alberta Dobry ran a friendly "listening ear" service during the entire project, assisted with proofreading this dissertation, and re- peatedly braved the discouraging task of trying to wake me from too- infrequent naps. --Fritz Briscoe also assisted with proofreading and fed me both nourishment and persistent assurance that there was hope! --Many other friends also offered their assistance, shaped my ideas by sharing theirs, subverted my leisure moments, tolerated my tirades, sympathized during my moments of crisis and quasi-crisis, and celebrated each small accomplishment with me. They share perhaps a bit of the blame for my sometimes-slow progress--but much of the credit for the completion of this project--and for my survival. To the Office of Economic Opportunity, which funded the larger study through which these data were collected. The financial assistance of these grants and the support of all those who contributed in any way to the larger study are gratefully acknowledged. Last but not least, To the parents and children who participated in this study. Their cooperation with what may well have appeared to be idiosyncratic de- mands of research was the primary factor which made this study possible. From them I learned far more than can be indicated by statistics and data summaries. jlc © vi Chapter 1. TABLE OF INTRODUCTION . . . . . . Statement of the Problem Objectives . . . . . . Conceptual Definitions . Assumptions. . . . . . . REVIEW OF LITERATURE . . Socialization Theory . . Interaction Theory . . Communication Process Overview . . . . . Initiation . . . . Response . . . . . Response Object. . Reinforcement. . . Feedback . . . . . Reward . . . . . . Verbal Communication. Overview . . . . . Verbalization. . . Fantasy Verbalization. Voice Tone . . . . Verbal Specificity Time Orientation . Task Orientation . CONTENTS Nature of Non-task Interference. Nonverbal Communication Overview . . . . . Affective Tone . . Anxiety. . . . . . Level of Involvement Physical Behavior. Teaching Procedure. . Overview . . . . . Concept Specification. vii 12 l6 l6 l6 18 18 l9 19 20 20 20 20 21 23 23 23 24 25 25 26 26 27 27 28 28 3O 3O 3O Chapter Teaching Method. . . . Methodological Considerations. Family Relationships . . . . . Parent-Child Relationships. Mother-Child Relationships. Father-Child Relationships. Comparison of Mother-Child and Father-Child Interaction. . . . . . . RESEARCH DESIGN AND PROCEDURES Design . . . . . . . . . . . . Sample Selection . . . . . . . Description of Sample. . . . . Data Collection Procedures . . Measurement. . . . . . . . . . Mother-Child and Father-Child Parent-Child Interaction Rating Hypotheses . . . . . . . . . . Operational Definitions. . . . Data Reduction and Analysis. . Data Processing . . . . . . Statistical Analysis. . . . RESULTS. . . . . . . . . . . . Communication Process. . . . . overView. O O O O O I O O I Test of Three-Way Interaction Tests of Two-Way Interactions Tests of Main Effects . . Test of Grand Mean. . . . . Contribution of Covariates. Verbal Communication . . . . . overview. O I O O O I O O 0 Test of Three-Way Interaction Tests of Two-Way Interactions Tests of Main Effects . . . Test of Grand Mean. . . . . Contribution of Covariates. Nonverbal Communication. . . . Overview. . . . . . . . . . Test of Three-Way Interaction Tests of Two-Way Interactions Tests of Main Effects . . . viii Interaction . Procedure . Page 31 32 35 35 4O 43 46 51 51 53 55 6O 65 65 66 68 69 72 72 74 88 89 89 89 91 94 103 107 107 107 110 110 112 116 122 122 122 125 125 127 Chapter Test of Grand Mean. . . . Contribution of Covariates. Teaching Procedure . . . . . Overview. Test of Three-Way Interaction Tests of Two-Way Interactions Tests of Main Effects . . Test of Grand Mean. . . . Contribution of Covariates. Summary. . Overall Comparisons . . . Comparisons on Individual Variables Communication Process. Verbal Communication . Nonverbal Communication. Teaching Procedure . . CONCLUSION Discussion Summary of Results. . . . Overall Comparison. . . . Socioeconomic Group Sex of Child Comparison . Race Comparison . . . . . Covariates. Limitations. Implications . Suggestions for Further Investigation. BIBLIOGRAPHY. . . APPENDI II. CES. . . . INSTRUMENTS. Mother-Child Interaction: Father-Child Interaction: Comparison. Toy Sorting Task. Nine Block Sorting Task Parent-Child Interaction Rating Procedure. CELL MEANS Cell Means Cell Means Cell Means Cell Means Cell Means OF of of of of of VARIABLES FOR ALL DEMOGRAPHIC GROUPS Communication Process Variables. Verbal Communication Variables Nonverbal Communication Variables. Teaching Procedure Variables . Covariates . . ix Page 134 134 139 139 139 141 149 151 159 159 159 161 161 166 167 168 171 172 172 172 173 176 178 179 180 182 184 187 202 202 202 207 212 243 243 253 265 271 276 LIST OF TABLES Design Matrix . . . . . . . . . . . . . . . . . . . Age of Parents at Time of Assessment. . . . . . . . Educational Level of Parents. . . . . . . . . . . . Occupation of Fathers . . . . . . . . . . . . . . . Age of Sampled Children at Time of Parent-Child Assessment . . . . . . . . . . . . . . . . . . . Family Size of Sample Families. . . . . . . . . . Ordinal Position of Sampled Child in Family . . . . Quantification of P-CIRP Variables Used in Analysis Summary of Multivariate Tests of Differences between Mothers and Fathers on Communication Process Dimensions of Parent-Child Interaction . Analyses of Covariance for Race x Sex x SES Interaction: Communication Process. . . . . . . Analyses of Covariance for Sex x SES Interaction: Communication Process. . . . . . . . . . . . . . Analyses of Covariance for Race x SES Interaction: Communication Process. . . . . . . . . . . . . . Analyses of Covariance for Race x Sex Interaction: Communication Process. . . . . . . . . . . . . . Analyses of Covariance for SES Main Effect: Communication Process. . . . . . . . . . . . . . Analyses of Covariance for Sex Main Effect: Communication Process. . . . . . . . . . . . . . Analyses of Covariance for Race Main Effect: Communication Process. . . . . . . . . . . . . . Page 57 58 59 61 62 63 75 90 92 93 95 96 97 104 105 Table 4‘9 0 4-10. 4-11. 4 12. 4-13. 4-14. 4-15- 4-16- 4-18. 4-19. 4-22. 4-23. Analyses of Covariance for Grand Mean: ~Process. . . . . . . . . Tests for Significance of Covariates in Multivariate Tests of Differences between Mothers and Fathers on Communication Process Dimensions of Parent- Child Interaction. . . . Summary of Multivariate Tests of Differences between Mothers and Fathers on Verbal Communication Dimensions of Parent-Child Interaction Analyses of Covariance for Race x Sex x SES Verbal Communication Interaction: Communication Analyses of Covariance for Sex x SES Interaction: Verbal Communication . . Analyses of Covariance for Race Verbal Communication . . Analyses of Covariance for Race Verbal Communication . . X Sex Interaction: Analyses of Covariance for SES Main Effect: Communication. . . . . . Analyses of Covariance for Sex Main Effect: Communication. . . . . . Analyses of Covariance for Race Main Effect: Communication. . . . . . Analyses of Covariance for Grand Mean: Communication. . . . . . Tests for Significance of Covariates in Multivariate Tests of Differences between Mothers and Fathers on Verbal Communication Dimensions of Parent- Child Interaction. . . . Summary of Multivariate Tests of Differences between Mothers and Fathers on Nonverbal Commu- Verbal Verbal Verbal Verbal nication Dimensions of Parent-Child Interaction. Analyses of Covariance for Race x Sex x SES Nonverbal Communication. Interaction: Analyses of Covariance for Sex x SES Interaction: Nonverbal Communication. xi Page 106 108 109 111 113 114 115 117 118 119 120 123 124 126 128 Table 4-24. 4-25. 4-26. 4-28. 4-29. 4-32. 4-33. 4-36. 4-39. Analyses of Covariance for Race x SES Interaction: Nonverbal Communication. . . . . . . . . . . . . . Analyses of Covariance for Race x Sex Interaction: Nonverbal Communication. . . . . . . . . . . . . . Analyses of Covariance for SES Main Effect: Nonverbal Communication. . . . . . . . . . . . . . . . . . . Analyses of Covariance for Race Main Effect: Nonverbal Communication. . . . . . . . . . . . . . Analyses of Covariance for Sex Main Effect: Nonverbal Communication. . . . . . . . . . . . . . . . . . . Analyses of Covariance for Grand Mean: Nonverbal Communication. . . . . . . . . . . . . . . . . . . Tests for Significance of Covariates in Multivariate Tests of Differences between Mothers and Fathers on Nonverbal Communication Dimensions of Parent— Child Interaction. . . . . . . . . . . . . . . . . Summary of Multivariate Tests of Differences between Mothers and Fathers on Teaching Procedure Dimensions of Parent-Child Interaction . . . . . . Analyses of Covariance for Race x Sex x SES Interaction: Teaching Procedure . . . . . . . . . Analyses of Covariance for Sex x SES Interaction: Teaching Procedure . . . . . . . . . . . . . . . . Analyses of Covariance for Race x SES Interaction: Teaching Procedure . . . . . . . . . . . . . . . . Analyses of Covariance for Race x Sex Interaction: Teaching Procedure . . . . . . . . . . . . . . . . Analyses of Covariance for SES Main Effect: Teaching Procedure. . . . . . . . . . . . . . . . . . . . . Analyses of Covariance for Sex Main Effect: Teaching Procedure. . . . . . . . . . . . . . . . . . . . . Analyses of Covariance for Race Main Effect: Teaching Procedure. . . . . . . . . . . . . . . . . . . . . Analyses of Covariance for Grand Mean: Teaching Procedure. . . . . . . . . . . xii Page 129 130 132 135 136 137 138 140 142 143 144 145 152 153 154 I-3- II-l. II-2. II-3. II-4. II-S. II-6. II-7. II‘8. 11.9 0 II-lO. II-ll. Tests for Significance of Covariates in Multivariate Tests of Differences between Mothers and Fathers on Teaching Procedure Dimensions of Parent-Child Interaction. . . . . . . . . . . . . . . . . . . . Summary of Significance and Marginal Significance for Individual Variables . . . . . . . . . . . . . Minimum Suggested Observer Reliability Indices for Parent-Child Interaction Rating Procedure. . . . . Category Descriptions for General Information Section of Parent-Child Interaction Rating Procedure . . . Category Descriptions for Sections 1 and 2 of Parent-Child Interaction Rating Procedure. . . . . Demographic Group Means for Initiation Categories on the Parent-Child Interaction Rating Procedure. . . Demographic Group Means for Response Categories on the Parent-Child Interaction Rating Procedure. . . Demographic Group Means for Response Object on the Parent-Child Interaction Rating Procedure. . . . . Demographic Group Means for Reinforcement on the Parent-Child Interaction Rating Procedure. . . . . Demographic Group Means for Feedback on the Parent— Child Interaction Rating Procedure . . . . . . . . Demographic Group Means for Reward Categories on the Parent-Child Interaction Rating Procedure. . . . . Demographic Group Means for Verbalization Categories on the Parent-Child Interaction Rating Procedure . Demographic Group Means for Fantasy Verbalization on the Parent-Child Interaction Rating Procedure. . . Demographic Group Means for Voice Tone on the Parent- Child Interaction Rating Procedure . . . . . . . . Demographic Group Means for Verbal Specificity on the Parent-Child Interaction Rating Procedure. . . Demographic Group Means for Time Orientation on the Parent—Child Interaction Rating Procedure. . . . . xiii Page 160 162 230 233 234 244 246 248 249 250 251 254 257 258 259 260 Table Page II-12. Demographic Group Means for Task Orientation on the Parent-Child Interaction Rating Procedure. . . . . . 261 II-l3. Demoqraphic Group Means for Interference (Initiator) Categories on the Parent-Child Interaction Rating Procedure I O O O O I C C C O O O O O O O O I O O O O 262 II-l4. Demographic Group Means for Interference (Task Remoteness) on the Parent-Child Interaction Rating Procedure. . . . . . . . . . . . . . . . . . . . . . 264 II-lS. Demographic Group Means for Affective Tone on the Parent-Child Interaction Rating Procedure. . . . . . 266 II-l6. Demographic Group Means for Anxiety on the Parent- Child Interaction Rating Procedure . . . . . . . . . 267 II—l7 Demographic Group Means for Level of Involvement on the Parent-Child Interaction Rating Procedure. . . . 268 II-lB Demographic Group Means for Physical Behavior Categories on the Parent-Child Interaction Rating Procedure. . . . . . . . . . . . . . . . . . . . . . 269 II—19. Demographic Group Means for Concept Specification Categories on the Parent-Child Interaction Rating Procedure. . . . . . . . . . . . . . . . . . . . . . 272 II-20. Demographic Group Means for Teaching Method Categories on the Parent-Child Interaction Rating Procedure. . . . . . . . . . . . . . . . . . . . . . 274 II-21. Demographic Group Means for Length of Session on the Parent-Child Interaction Rating Procedure. . . . . . 277 II-22. Demographic Group Means for Age of Child at Time of Parent's Assessment on the Parent-Child Interaction Rating Procedure . . . . . . . . . . . . 278 xiv d}. ..Au.n UL»). ”Shana... .,.. . . 2.... N HM.— \ LIST OF FIGURES Schematic representation of spheres of influence on the socialization of the individual. . . Bales' Interaction Process Analysis . . . . . . . . . . Comparison between mothers and fathers in relation to socioeconomic group for use of unqualified power assertion. O O O O O O O O O O O O O O O O O O O O 0 Comparison between mothers and fathers in relation to socioeconomic group for use of rejection responses . Comparison between mothers and fathers in relation to socioeconomic group for initiations to which there is no opportunity to respond . . . . . . . . . . . . Comparison between mothers and fathers in relation to socioeconomic group for ignoral responses. . . . . . Comparison between mothers and fathers in relation to socioeconomic group for relative distribution of response object categories . . . . . . . . . . . . . Comparison between mothers and fathers in relation to socioeconomic group for absolute amount of response object categories. . . . . . . . . . . . . . . . . . Comparison between mothers and fathers in relation to socioeconomic group for affective tone . . . . . . . Comparison between mothers and fathers in relation to socioeconomic group for anxiety level. . . . . . . . Comparison between mothers and fathers in relation to race and sex of child in relative amount of teaching activity devoted to general summarization . Comparison between mothers and fathers in relation to race and sex of child in relative amount of teaching activity devoted to the introduction of new concepts . . . . . . . . . . . . . . . . . . . . XV Page 15 22 99 99 101 101 102 102 133 133 147 148 4-12. I-6. Comparison between mothers and fathers in relation to race and sex of child in relative amount of teaching activity devoted to use of illustration as a teaching method . . . . . . . . . . . . . . . Comparison between mothers and fathers in relation to socioeconomic group in relative amount of teaching activity devoted to use of demonstration as a teaching method. . . . . . . . . . . . . . . . . Scoring form for Mother-Child Interaction: Toy Sorting Task . . . . . . . . . . . . . . . . . . . Scoring form for Father-Child Interaction: Nine Block Sorting Task . . . . . . . . . . . . . . . . General information form: Parent-Child Interaction Rating Procedure . . . . . . . . . . . . . . . . . Rating frame for Section l--Parent observation: Parent-Child Interaction Rating Procedure. . . . . Format for Section l--Parent observation: Parent- Child Interaction Rating Procedure . . . . . . . . Rating frame for Section l--Chi1d observation: Parent-Child Interaction Rating Procedure. . . . . Format for Section l--Child observation: Parent- Child Interaction Rating Procedure . . . . . . . . Rating form--Section l: Parent-Child Interaction Rating Procedure . . . . . . . . . . . . . . . . Rating frame for Section 2--Parent observation: Parent-Child Interaction Rating Procedure. . . . . Format for Section 2--Parent observation: Parent- Child Interaction Rating Procedure . . . . . . . Rating frame for Section 2--Chi1d observation: Parent-Child Interaction Rating Procedure. . . . . Format for Section 2--Chi1d observation: Parent- Child Interaction Rating Procedure . . . . . . . Rating form--Section 2: Parent-Child Interaction Rating Procedure . . . . . . . . . . . . . . . . . Indication of nonapplicable P-CIRP categories . . . . xvi Page 150 158 206 211 215 217 218 219 220 221 222 223 224 225 226 228 Figure Page I-lS. Assignment of points for P-CIRP observer reliability. . 231 xvii socialization of the individual. fluences. as formative forces on the child's behavior. CHAPTER 1 INTRODUCTION If America's parents are given the place, power, and prestige to enable them to function as guides, companions, and sources of love and discipline for their children, and to have a decisive role in de- termining the environments and programs in which their children live and grow, the great majority of them will be able to take full advantage of the opportunity to enhance the quality of life both for their children and themselves. . . . The cru- cial factor is not how much time is spent with the child but how the time is spent. A child learns, he becomes human, primarily through participation in a challenging activity with those he loves and admires. It is the example, challenge, and rein- forcement provided by people who care that enable a child to develop both his ability and his identity. --Children and Parents: Together in the World. In Report to the President: White House Conference on Children. Forum 15. Washington, D.C.: _U.S. Government Printing Office, 1970, p. 241. Statement of the Problem The family has traditionally been considered as a key force in the "widening world of childhood" pictures a spiral out from the parental home, wherein the child initially experiences his chief social in- Clausen's (1968) description of the The techniques which the parents employ in their treatment of the child, toqether with their general attitude toward him, serve years, the particular quality of the parents' roles in parent-child 1 Throughout the dependent relationships is of paramount importance in the establishment of perma- nent motivational and personality attributes (Merrill, 1946). The preschool period is a particularly critical period in the child's education and general socialization. His rapidly developing cognitive skills make this a prime learning time, and his widening range of contacts provides the experiential base for further develop- ment of these skills. At the same time, the family provides the pri- mary influence on the child's socialization, a position of continuing qualitative and quantitative impact in the child's total development. Unfortunately, empirical information concerning the nature of parent-child relationships has been sparse at best. Only within this century have the importance and feasibility of conducting scientific investigations of parent-child relationships even been accepted (Gildea, et a1., 1961). Systematic observation of parent-child relationships began to appear only after World War II, and the first use of preceded behavior categories to record interaction was not until 1951 (Lytton, 1971). Most studies still have tended to employ either parental reports or summary observations of molar variables, both of which present obvious psychometric problems. Hoffman and Lippitt's (1960) description of more than a decade ago still summarizes the present situation quite well: Increasing attention is being paid to the importance and complexity of measuring parent behavior and parent-child interaction. In part, this growing interest probably stems from the rather low level of success in making direct link- ages between parental personality, values, and attitudes and child personality and adjustment; and, in part, it probably reflects the theoretical and methodological inter- est in moving from correlational studies to more process- oriented studies of socialization. Direct observation of parent—child interaction is still rare in the research literature, as compared to the number of studies in which categories of parental behavior are derived from parent reports or from child reports (p. 976). Studies concerned with parent—child relationships have generally focused only on the mother-child dyad. Psychologists and educators have typically ignored fathers, as pointed out by Weiss (1970) in a very pertinent article. He cited a continuing trend for emphasis being placed on the mother: Checking the number of articles in Psychological Abstracts . . . in 1958 there were 20 entries for all categories con- cerning mothers, while fathers accounted for only 2 studies. For 1968 the corresponding entries were 144 and 29, respec- tively. While the number of studies involving fathers in— creased during the decade, it was still out of proportion to the psychologists' penchant for studying mothers (p. 6). Furthermore, most of those studies which have acknowledged the father's role have done so as reported by the mother, from a psychoanalytic perspective, or in very general terms. In a decade review of research in the area of parent-child rela- tionships, Walters and Stinnett (1971) recommended not only that the relationship of the father to the children should be explored to a much greater extent but also that many assumptions concerning parent-child relationships should be explored. They cited the following example: Because we have believed that the impact of mothers upon the development of children is greater than the impact of fathers, we have investigated maternal impact to a far greater extent than we have examined the impact of fathers. Yet, much of the evidence of the past decade suggests that the variability of children's behavior is more closely associated with the type of father one has than the type of mother (p. 102). A similar plea for increased study of family interaction through direct observation and for study of interactions involving fathers was made in a recent report on early childhood research and development needs (Chapman, 1972). Information about parent-child interaction patterns is of particu— lar importance to developmental, educational, and social concerns at a time when societal patterns are undergoing relatively rapid changes. For example, questions about the interchangeability of maternal and paternal roles, sex-role identification, father absence (either perma- nent or temporary), and alternatives to traditional patterns of sociali- zation emphasize the need for empirical information in this area. It is important, therefore, to obtain comparative as well as descriptive information concerning the interaction patterns of both mothers and fathers with their preschool children. Objectives The basic purpose of this study was to compare mothers' interac- tions with their preschool children and fathers' interactions with these same children. The situation of interest was defined as an un— structured teaching situation in which each parent was paired individ— ually with his child. More specifically, this study was designed to identify potential similarities and/or differences between mothers and fathers on the following aspects of their didactic interactions with their preschool children: communication process, verbal communication, nonverbal com— munication, and teaching procedure. A secondary intent was to investi- gate these comparisons in relation to socioeconomic group, race, and sex of the child. Conceptual Definitions For purposes of this study, interaction in the dyadic didactic situation was considered to be indicated by communication process, 4..” verbal communication, nonverbal communication, and teaching procedure dimensions. These four components are each comprised of several aspects, which are delineated as follows: (a) Communication process--initiation, response, response object, reinforcement, feedback, and reward; (b) Verbal communication--verbalization, verbal fantasy, voice tone, verbal specificity, time orientation, task orientation, and nature of interference; (c) Nonverbal communication--affective tone, anxiety, level of in— volvement, and physical behavior; (d) Teaching procedure-—concept specification and teaching method. Conceptual definitions of each of these aspects are as follows: (a) Initiation is the first unit in an interaction sequence. It describes the stimulus portion of a communication message, or that which provides the circumstances to which the other person may respond. This call for response may be neutral in nature or it may carry a par- ticular psychological thrust (Moustakas, Sigel, and Schalock, 1956). This may be in the form of a power assertion, either qualified or un- qualified in nature. The unqualified power assertion is defined as an attempt to make use of personal jurisdiction by physical punishment, isolation, or insistence upon a specified behavior, while the qualified power assertion makes use of the same technique, but with the addition of a reason or a cushion or both to this exertion of authority. Another approach is the use of a bribe or a bargain, appeal, or suggestion for behavior with the offer of a reward (either tangible or intangible) for compliance. Still another form of direction is that in which a sugges— tion is made, but the other person is given an overt choice of compliance. L' v . ,.. I I); D \ The probable impact of choosing one alternative over the other is irrele- vant in this case, since the only consideration is whether a choice is offered, whether this be nominal or actual. Finally, the situation may arise in which a communication message does not contain an initiation segment. This may be because no opportunity for continuing the communi— cation is allowed by the other person or because the interaction sequence is an ongoing one from a preceding time interval. The failure to con- tinue the interaction may also be an indication of a particular behav- ioral style which might be described as egocentric in nature, or repre- sentative of a lower level of social interaction. Thus this category contains both a style factor and a developmental one. (b) Response is the portion of a communication message which fol— lows from communication by another and derives its meaning from that process by continuing the sequence (or failing to do so). Its presence and the type of response presented represent both a style dimension and a level of development in communication skill. The primary distinction to be made is whether a response actually occurs or if it fails to occur. If it does occur, this response may be either an acceptance of the communication of the other person or a rejection of this communica- tive interaction. If a response to the other person does not occur, this may be because the other person does not give an opportunity for the response, even though the situation for a response has been created, or it may be that there is no immediately preceding communication by the other person (as in the case of the beginning of a new interaction sequence). The failure to respond may be the case of ignoring, which may be either intentional or unintentional, but represents a failure to respond to the message presented by the other person, even though the opportunity to do so is presented. Finally, the communication may be the continuation of an ongoing communication message from a previous interval. (c) Response object designates the focus of a response (either ac- ceptance or rejection) toward the person as an individual or toward his behavior or performance. Interpretation of the response and its subse- quent impact are a function of this response dimension. (d) Reinforcement indicates the appropriateness of a response (either acceptance or rejection) in consideration of the stimulus pro- vided by the other person. The correctness of the response is theoreti- cally as important as its occurrence, since cues which are in conflict may be given, and different types of behaviors may be reinforced differ- entially. Therefore, the consistency of cues is considered, as well as the kinds of behavior which are reinforced and how this is done. (e) Feedback indicates the specificity with which a response is made. Each acceptance or rejection may be so determined because of its explicit nature or by implication. Language forms (or nonverbal language substitutes) are used as the distinguishing criterion for the eXplicit or implicit nature of the feedback. This characteristic of a response is important not only in terms of its clarity in communicating meaning, but also because of the significance of this factor for the child's learning. Since language has been shown to be an important mediator of meaning, the specificity of feedback concerning the previous communica- tion is a significant indication of both communication style and the extension of that, the efficiency of general teaching behavior (Hess, et a1., 1968). (f) Reward describes the extent to which either praise or criticism is included in a feedback message. The importance of the use of this supplementary form of communication may be seen in terms of styles of communication and also of consistency of cues provided to the other person. Both affect and content are criteria for designation of either praise or criticism. (g) Verbalization is characterized as the quantity of verbal com— munication which occurs and the quality of that communication. Quality is described in terms of the following six mutually exclusive categories: process of orientation, evaluation, control over others, decision, ten- sion management, and integration (Bales, 1951). (h) Verbal fantasy is defined as the extent to which imaginary verbalizations, such as dramatic play or obvious exaggerations, are used. Its psychological interpretations vary, but the use of fantasy as such represents a particular approach to reality in both task and nontask situations. In particular, a parent's use of fantasy in the instruction of a child indicates a method of relating to that child and, in many cases, a diversity in approach that is important in con- sidering teaching style. (i) Voice tone is the affect conveyed through tonal patterns in verbal communication, regardless of the literal content of that commu- nication. This characterizes the connotative in contrast to the deno- tative aspect of the verbal message. (j) Verbal specificity is described as the extent to which verbal communication uses a concrete, differentiated structure rather than a global or abstract one. This concept has received theoretical atten- tion as an indicator of general perceptual style (e.g., Hess, et a1., 1968; Bernstein, 1961). The role of language as an indicator of thought modes can be tapped to explore the conceptual orientation which appears to distinguish between the global and the differentiated thinkers in a broad sense and which is an important component of efficient informa- tion transmission in a more limited sense. (k) Time orientation is the extent to which verbal communication reflects a focus on past, present, or future perspectives. This socio- logical concept has received attention as an indicator of value orienta- tion and thus is an important behavioral style (Kluckhohn, 1961). While verb tense is one indicator of this orientation, the intent of the time reference is more important than the grammatical construction as such. (1) Task orientation describes the extent to which verbal communi- cation reflects a focus on task (in contrast to non-task or social) goals. This focus has been studied in relation to managerial style and resulting affective environment (e.g., Bales, 1951), elements which are critical in the communication and interaction processes. (m) Nature of interference designates the source of diversion from a task orientation as reflected by verbal communication. For example, the control of a diversion may stem from one of three sources--the parent, the child, or some outside factor. For a diversion initiated by either the parent or the child, the focus may be shifted to some non-task use of the task materials, to some other factor in the imme— diate environment, or to some nonsituational factor. Outside inter- ference is used to designate those situations which are not under the control of either the parent or the child. (n) Affective tone is the emotional characterization of a situa- tion based on general outlook and the resultant structuring of the environment. 10 (0) Anxiety reflects an affective level of an individual in terms of apprehension or uneasiness, but without necessarily having any defi— nite object of concern. A somewhat amorphous general personality trait, anxiety provides a situational context for other behavior which occurs and which must be considered for its appropriate interpretation. (p) Level of involvement indicates the amount of psychological in- volvement of an individual in his immediate situation. This behavior may be an indication of general personality style, and it is also an important determinant of the affective environment of the situation. As such it is related to the effectiveness of communication. Indirectly, this factor might also be considered a communication in itself of the importance of the situation and/or the other person. (q) Physical behavior represents nonverbal behavior involving some type of overt physical movement and representing an intentional commu- nicative act. Both quantitative and qualitative components are con- sidered within the scope of physical behavior. Quality is described using a multiple classification approach across two dimensions: (1) presence or absence of physical contact, and (2) positive or negative quality. (r) Concept specification is described by the nature of the behav- ioral units in a teaching process as reflected by organization of the various aspects of concepts presented and of their component units, as well as the pattern of approaches used to communicate the information. This system includes introduction of a new concept, introduction of a new approach to teaching the same concept, a focus on a different aspect of the same concept, the continuation of the same method of teaching a concept, general orientation, and general summary. These methods of 11 presentation are important both for the child's learning of a desig- nated task and in a more general sense for his logical acquisition of knowledge of the world around him and his own conceptual development. Both continuity and diversity are important in communication patterns, and the organization of these communication messages reflects the manner in which thought patterns are learned and practiced. (5) Teaching method indicates the type of technique (i.e., verbal, nonverbal, or both) used in presenting a unit of communication in the teaching process. This mode, as well as the pattern of methods, is an important indicator of the parent's communicative style and also of effectiveness in teaching particular concepts. Assumptions The conceptualization and interpretation of this study were based on several theoretical assumptions. These included the following: (a) The preschool child's parents are his most important socializ— ing agents, and through his interactions with them he learns both general behavior patterns and specific modes of relating to the world around him. (b) Interaction is a dynamic process, but it can nevertheless be systematically dissected into relatively distinct units for purposes of analysis. (c) Parent—child interaction involves reciprocal influences be- tween the parent and the child, but parental behavior can be isolated analytically as a meaningful individual variable. (d) Didactic interaction within the parent-child dyad is adequately characterized by dimensions of communication process, verbal communica- tion, nonverbal communication, and teaching procedure. If If If If If If If If If If If WWWWWWWQ’WWQ’ child child child child child child child child child child child CHAPTER 2 REVIEW OF LITERATURE lives lives lives lives lives lives lives lives lives lives lives with with with with with with with with with with with criticism, he learns to condemn. hostility, he learns to fight. ridicule, he learns to be shy. shame, he learns to feel guilty. tolerance, he learns to be patient. encouragement, he learns confidence. praise, he learns to appreciate. fairness, he learns justice. security, he learns to have faith. approval, he learns to like himself. acceptance and friendship, he learns to find love in the world. -—Dorothy Law Nolte, Children Learn What They Live. Socialization Theory One of the most basic tasks of a society is socialization, or the problem of how to prepare individuals so that they will become adequate members of that society. It is not surprising that this "concern with the preper development of offspring-~with their acquisition of needed skills, the curbing of aggressive tendencies, the directing of their feet to paths of righteousness--is as old as man (Clausen, 1968, p. 20)." Typically, a dual foundation for this concern has been evident--first, as an inevitable consequence of being a parent, and secondly, as a social concern, since cultural units have traditionally wished to perpetuate those values and behaviors on which they are based. Documents dating back to the early philosophers, including litera- ture in a wide range of disciplines, have dealt with various aspects of 12 13 the concept and process of socialization. Among those disciplinary con- siderations summarized by Dager (1964), Clausen (1968), and Zigler and Child (1969) were anthropology (social and cultural), sociology (family and cultural), and psychology (social, developmental, and personality). In addition to these basic disciplines, such applied areas as education, child development, family relations, and home management have devoted considerable attention to the study of aspects of socialization. So broad has been the treatment, in fact, that Dager (1964) described socialization, "in addition to biological and hereditary makeup, (as) the common meeting ground of all the sciences dealing with man (p. 740)." Since the range of disciplinary considerations has been so broad, it is not surprising that the range of theoretical perspectives for viewing socialization has also been great. Socialization has been described as the twin problem of personality, a focus which itself in- troduces a large and diverse base of theoretical positions. Fortunately, these positions have more often been complementary than competing, so it is possible to consider elements of several in deriving an appropriate framework. Unfortunately, however, much of the vast amount of litera- ture dealing with socialization has been speculative in nature; empiri- cal evidence which does exist has characteristically focused on a single behavior or dimension viewed from a single perspective. Emphasis on the socialization process has variously been placed on the individual and on society. Elkin (1960), for example, and Child (1954) focused on learning by the individual of a particular set of behaviors. Aberle (1961), however, stressed the social apparatus which influences the individual's learning and defines for him what is ac- ceptable. Still other authors have emphasized the two-directional a l I“ 1‘ 1.‘ 14 influences, or reciprocal aspects of the socialization process (e.g., Bell, 1968, 1969; Fields, 1969; Osofsky, 1970). As a process, socialization must be seen as continuing throughout the life cycle. Clausen (1968) stressed this dynamic dimension in his characterization of the "continuing interaction between the individual and those who seek to influence him, an interaction that undergoes many phases and changes (p. 3)." Despite many inter- and intra-individual, as well as inter- and intra-societal, variations, certain basic patterns can be identified which characterize the process. From a developmental perspective, the influences represent a consistent--though not neces- sarily uniform--decrease in proximity in the social structure from the individual. The increasing spheres of influence on the individual are repre- sented schematically in the conceptual framework in Figure 2-1. As indicated here, the environment of the young child centers in the home, Where the family acts as the primary agent of socialization, imparting the child with the skills, knowledge, attitudes, values, and motives CUrrent in the group (Sewell, 1963). The mother is the primary focus ‘3f the infant's very early world, which soon expands to include the father, and then other family members. As the child's range of experi- er'ICe broadens, informal and then formal groups assume roles of increas- ifug importance in the developmental process. Emphasis ultimately spi- rals into the abstract dimension of cultural institutions. At every Stage of development, of course, dominant cultural themes (e.g., ideolo- ‘gies, values, norms, and attitudes) act as mediators of the total process. Despite increasing realms of influence from external sources and Shifting patterns of saliency, the long—range importance of early Developmental Phases Adulthood Adolescence Later Childhood/ Preadolescence Middle Childhood Early Childhood Infancy FIGURE 2-1. /“/“‘I“‘\' \ 15 Cultural Themes I l I I I I Inbtitution% I I I I I I IFormal I Organizatiohs I I I Informal' I . . I Organizations .Gangs, ICliques Dimensions of Influence Social Systems Extra-familial Groups Family Schematic representation of spheres of influence on the socialization of the individual (Adapted from models in other sources, e.g., Dager, l6 influences cannot be denied. Bloom (1964), looking at stability and change in human characteristics, suggested that the tremendous impor- tance of the early environment may be due to the rapidity of growth during this time. The continuity of family interactions, as well as absolute amount of impact over time, establishes a pattern of influence which has been conceptualized in two major areas: one, sources and types of parental reinforcement, or the type, severity, consistency, and direction or distribution of parental reward and punishment; and two, parental exercise of power (McCandless, 1969, p. 801). Interaction Theory Communication Process Overview. Interaction was suggested by Hall (1959) as one of the ten Primary Message Systems through which one may begin to study a cul- ture. The interactionist approach to the study of human behavior in- cludes the concept that "human nature and the social order are products of communication (Shibutani, 1961)." While the meaning of interaction itself is a legitimate and important subject of study as a specialized phenomenon in its own right, the emphasis more functionally lies on interaction phenomena as they underlie man's organization of himself to himself, to others, and to his environment (Bean, 1967). The phenomenon of interaction has been described by Loomis (1960) as characterized by patterned social relations that display in their uniformities social elements, articulated by social processes, the dynamics of which account for the emergence, maintenance, and change of social systems. As a basis for study, Newcomb (1953) proposed that . . . many of those phenomena of social behavior which have been somewhat loosely assembled under the label of 'inter- action' can be more adequately studied as communicative 17 acts. . . . Just as the observable forms of certain solids are macroscopic outcomes of molecular structure, so certain observable group properties are predetermined by the con- ditions and consequences of communicative acts (p. 393). A number of different models have been proposed for analysis of the communication process. While the components of each differ, several elements described by Berlo (1960) appear to be relevant to the systems in general. First, a dynamic state or ongoing process is involved, rather than a static situation. Secondly, communication can be described as the performance of a set of behaviors involving the transmission or reception of messages. These messages, then, are described as behavioral events related to internal states of people. Homans (1958) was one of the first to suggest general adoption of an exchange model of social behavior. Actually one of the oldest theories of behavior, this view considered social behavior as an ex- change of goods, including both material and also nonmaterial ones such as symbols of approval or prestige. This model led to the suggestion that persons that give much to others try to get much from them, and persons that get much from others are under pressure to give much to them. This process of influence tends to work out at equilibrium to a balance in the exchanges. Using the social exchange model, Longabaugh (1963) developed a coding system for categorizing interpersonal behavior. This system described acts on two dimensions: (1) the resource salient--that of value to the interactors and prominent in their interaction, and (2) modes--the ways in which resources become salient and are dealt with by the interactors. Examples of resources are information, support, and control, while modes include seeking, offering, depriving, accept- ing, and not accepting. This coding system differs from all other 18 category systems in two ways. First, categorization is explicitly and solely made on the basis of presumed contribution of the action to the interpersonal relationship rather than on its meaning for the actor or its effect on the target. In addition, once an action creates a new focus for the participants, the immediate and direct implications of the action are followed until it is determined whether or not the salient resource has passed from its initial possessor. This model provides a basis for division of an interaction situation into units which have meaning but which are logically discriminable from each other in the total sequence. Initiation. The type of initiation which is made by an individual in an interaction situation reflects his involvement in that situation and also his assertion of power or control. If he chooses, for example, to discontinue his involvement, the implication is probably a lack of concern for the situation, the other individual, or both (Longabaugh, 1963). If, on the other hand, he maintains the interaction by attempt- ing to influence the behavior of the other individual, as is likely in a teaching situation, he typically manifests power through one of a variety of influence techniques (Hoffman, 1960; Sigel, et al., 1957; Sigel, 1960). These techniques may take the form of verbal or gestural communication or of physical or environmental communication. As Sigel, et al. (1957) pointed out, such influence techniques are typically used by parents as a means of modifying children's behavior and are of funda- mental importance in affecting the psychological development of the child. Response. The complement of the initiation in an interaction se- quence is the response. Using the Longabaugh (1963) model, alternatives 19 to the offered resource--or power assertion-—are suggested. Again, failure to respond may be interpreted as lack of concern for the other individual, the situation, or both. The relative effects of responses which are rejecting in nature in contrast to ignoring initiations may be suggested by sociometric analyses. Research (e.g., Moreno, 1960) has suggested that the individual who is not chosen at all may in fact have lower status than those who are chosen either positively 9£_nega- tively. Such an interpretation could be seen as consistent with behav- ioral theory, which has suggested that modification of behavior in an educational setting may be accomplished most effectively by positively reinforcing those behaviors which are desired and by ignoring those which are not desired (e.g., Lesser, 1971; Read, 1971; Wilson, et al., 1969). Response object. A time-honored educational principle has been that responses should be made to the behavior of the child rather than to his worth as an individual (e.g., Wilson, e£_alp, 1969). This pro— cedure has been suggested as critical in the development of a positive self-concept. The tendency to belittle an individual (in contrast to responding to his behavior) typically has been thought to signify de- fensiveness and insecurity (Read, 1971). Reinforcement. An important part of information transmission in the teaching situation is the feedback given about the correctness of the previous response. An implication of both the Hess-Shipman re- search (Hess, et a1., 1968) and the Boger-Kuipers research (Boger, eg 31., 1969) was that a child's learning was facilitated by appropriate feedback, whether it was positive or negative in nature. Again, specificity of information is related to this concept, as is consistency 20 of cues. A related problem which has typically plagued teacher edu- cators, for example, has been the teacher whose face or voice tone gives a different message than the words which are used in responding to a student. Feedback. Another technique used by both parents and teachers to "keep the child guessing" is the failure to give overt or explicit feed- back concerning the acceptability or correctness of behavior or per- formance. Like other indications of specificity, the use of explicit responses providing either confirmation or correction were indicated by Hess, et al. (1968) to be important factors for the child's learning. Through their reinforcing effect, such responses apparently act as social rewards or punishments to the child. Reward. Still another way in which teacher or parent reactions to children's behavior is given is with the use of either praise or criti- cism. As defined by Hess, et al. (1968), these behaviors are of greater intensity than simple affirmation or negation responses. Not surpris— ingly, these variables were found to be rather highly related to other variables which discriminated among mothers with respect to quality of teaching behavior. Verbal Communication Overview. The most frequently studied aspect of the communication process is that of verbal communication. The importance of language use has been emphasized by the fact that it represents a key distinction between human beings and all other forms of life. In addition, language has been pinpointed as a critical variable in the development of in— telligence and in general adaptive behavior (Bloom, 1964). Studies of language have tended to focus on several areas dealing variously with 21 such concerns as content, structure, and develOpmental aspects. Psycho- linguistics was developed relatively recently as a discipline with the intent of relating many of these diverse approaches into a single co— herent picture of language. As Miller (1951) has pointed out, the task of analyzing meaning can make operational sense only when we realize that it involves discovering the aspects of the total situation which are relevant to the speaker's behavior. Verbalization. Klineberg (1963), Deutsch (1963), and Brunner (1965) have all investigated the role of language in development and have found it to be a critical factor in the "cumulative deficit" which tends to characterize children from disadvantaged environments. Bereiter and Engelmann (1966) emphasized the verbal experience deficit hypothesis, sug— gesting that academic achievement is primarily affected by the verbal ex- perience of the child. Gordon (1965) also reported a number of language deficiencies in disadvantaged children, including limited vocabulary range, restriction in number of grammatical forms used, lack of experience with prolonged speech sequences, and restricted language usage with most communication through gestures and other nonverbal means. Hess, et_al. (1968) also focused on language usage as a primary indicator of learning and of interaction in the mother-child dyad. Psychologists have, in general, devoted considerable attention to the importance of the spoken (or written) word in communication and interaction. One of the most highly developed and frequently used systems for the analysis of such com- munication has been Bales' (1951) Interaction Process Analysis. Origi- nally designed for groups facing a common task, this system has been seen as appropriate for less structured groups as well. As shown in Figure 2-2, both dimensions of affective direction and of nature of content are 22 1. Shows solidarity, raises other's F. status, gives help, reward ie-e . . o o m-a 2. Shows tenSion release, jokes, -a-a w u A - - 8 46' 2 .3 laughs, shows satisfaction m e a 3. Agrees, shows passive acceptance, understands, concurs, complies __ 4. Gives suggestion, direction, implying autonomy for other 5. Gives Opinion, evaluation, anal- B ysis, expresses feeling, wish 6. Gives orientation, information, PI . . . ‘ .x H 3 L__ repeats, clarifies, confirms ea: — . . . abodef a d g 7. Asks for orientation, informa- y tion, repetition, confirmation C 8. Asks for opinion, evaluation, analysis, expression of feeling 9. Asks for suggestion, direction, possible ways of action 10. Disagrees, shows passive rejec- F' tion, formality, withholds help I m m .4 c u > . fi.2 3-3 D 11. Shows tenSion, asks for help, 8 g 2 g‘ withdraws out of field ‘“ E g 12. Shows antagonism, deflates other's status, defends or asserts self A - Positive reactions a - Problems of communication B - Attempted answers b - Problems of evaluation C - Questions c - Problems of control D - Negative reactions d - Problems of decision e - Problems of tension management f - Problems of reintegration FIGURE 2-2. Bales' Interaction Process Analysis (Bales, 1951, p. 59). 23 considered, including processes of orientation, evaluation, control over others, decision-making, tension management, and integration. Fantasy verbalization. The use of fantasy has typically been con- sidered within psychoanalytically-based literature as a defense mechanism (e.g., White, 1956). However, it has also been considered on a somewhat different level by educators, particularly in the area of early childe hood, as a teaching technique which can be used effectively in communi- cating both task and nontask dimensions (e.g., Spodek, 1972). The use of dramatic play, for example, has been encouraged as a means by which the limits of the child's world are expanded (Smilansky, 1968) and as a means of developing the mental, physical, creative, social, and emotional components of the child's being (Hildebrand, 1971). Voice tone. An important element of communication is not only what is said but also hpw_it is said. Eisenberg and Smith (1971) have pointed out that voice set is determined by the speaker's physical and psycho- logical condition and that it may in fact become the basis for social categorization. Indicators of the quality are vocal intensity (volume), pitch, resonance, rate, and rhythm; the composite of these factors can be used as a basis for relatively accurate assessment without the assist- ance of complex instruments, however. An interesting demonstration of this phenomenon was an experiment by Davitz and Davitz (1959), in which content-free speech (recitation of the alphabet) was used to convey par- ticular emotions. Identification of the intended affect was accomplished by a group of subjects with very high accuracy. Verbal specificity. Relative specificity of verbalization has re- ceived theoretical attention as an indicator of general perceptual style. Although, as Hess, et al. (1968) have pointed out, the specificity 24 dimension is a continuous one which includes both verbal and nonverbal components, a critical distinction has been the use of specific labels to refer to the relevant attributes under consideration. Analysis of maternal teaching behavior indicated that the most important variable in the relationship of the behavior of the predominantly working-class mothers to the performance of their children was the amount of speci- ficity which they used (Brophy, 1967, 1970). Time orientation. The sociological concept of time orientation has been considered as an indicator of the individual's value orientation (e.g., Kluckhohn, 1961) and therefore as an important determinant of be- havioral style. Hall (1959) has pointed out that not all people experi- ence time in the same way. The basic orientation of our culture is toward time as a stable point of reference, basically unchanging, so that time can be used as a guide to, or a measure of, behaviors. Within our culture, however, different social classes have been identified which, though having factors in common by virtue of being part of the same culture, also have differences among them. There has been evi- dence, for example, that temporal goal orientations are different for lower socioeconomic group children than for their middle socioeconomic group peers (e.g., Mischel, 1961). Members of the middle socioeconomic groups have tended to be more oriented toward the future, while the upper socioeconomic group has typically used a temporal orientation of several generations (Bonner, 1965). In addition, because the mode of time orientation is culturally transmitted, it has been suggested that differences between people occur to some extent as a function of age (May, 1967). Anticipation and intentionality have also been suggested as dimensions which would theoretically be reflected in verbal time 25 orientation. Thus individual style as well as possible cultural or subcultural differences might be expected to occur relative to this dimension. Task orientation. The relative focus on task and non-task factors has been considered in relation to managerial style and the resulting affective environment (e.g., Bales, 1951). Tests of the "great man" theory of leadership (e.g., Borgatta, et a1., 1954; Theodorson, 1957) have indicated that maximum effectiveness in a group situation is achieved by an individual who has a high sociability rating (i.e., is popular with the other group members) but who also has a high task orientation and competence. Again, the application of these principles to the educational setting has produced consistent results with those found in the management studies. For example, studies of teacher ef- fectiveness (e.g., Beecher, 1961; Getzels and Jackson, 1963; Ryans, 1960) have indicated that the most effective teachers have been ones whose primary orientation was on the cognitive learning of the students and who were at the same time rated most highly by the students in terms of popularity. These findings have not, however, been directed toward a content versus process orientation of learning, but rather toward pri— macy of learning versus feeling, a still unresolved debate in the edu- cational literature. In summary, though, the implications of task ori- entation appear to suggest that task accomplishment (e.g., learning) is facilitated by a direct focus on that task, but that the manner in which the task accomplishment is approached is a critical dimension which must be considered simultaneously. Nature of non-task interference. The initiator of non-task orien- tation is one indication of the individual who is in primary control in 26 the situation as well as the relative task orientations of the indi- viduals involved. The focus of this verbalization may also indicate the extent to which it departs from the task at hand-~which again indi- cates a degree of task orientation. There has been little previous re- search focusing directly on this dimension of verbal communication, al- though the previously cited literature on task orientation is again somewhat relevant. In addition, literature on extent of adult control in a task situation might be considered. Studies of teacher attitudes and behavior (e.g., Levin, 8t al., 1957; McGee, 1955) and of parental attitudes and behavior (e.g., Baldwin, et a1., 1949; Schaefer, 1959; Sears, et al., 1957; Straus, 1962) have suggested that in general, the optimal situation is one which is consistent and is one in which adult control is maintained but in which the child has the freedom to exer- cise his own autonomy within the limits of the situation presented. Nonverbal Communication Overview. Despite the old cliche that "actions speak louder than words,‘ the formal study of communication has traditionally emphasized the use of language to the exclusion of all other forms of communica- tion. However, the nonverbal aspects of interaction convey meaning also; as Eisenberg and Smith (1971) have stressed, "analysis of inter- personal communication is incomplete without a consideration of more than mere use of language (p. 4)." It has been only within the past fifty years that behavioral scientists have concerned themselves with a systematic attempt to ana— lyze all_levels of communication. The void in this area was quite evi— dent still from the extreme sparsity of relevant material in the litera- ture. Perhaps the best generalization about knowledge of nonverbal 27 communication was made by Sapir in 1949 but is applicable still; he wrote that "we respond to gestures with an extreme alertness and, one might say, in accordance with an elaborate and secret code that is written nowhere, known by none, and understood by all (p. 556)." Affective tone. Both pervasive and transitory emotional states establish the base undergirding any interaction by indicating the af- fective dimension (Krech and Crutchfield, 1948). Relatively little re- search has been directed at the specific conditions under which differ- ent moods appear, change, and disappear. Accuracy with which such emo- tional states can be judged has been rather widely investigated, however, particularly through analysis of facial expressions. By the 1940's, though, psychologists had agreed that emotion could not be judged from facial expression alone, for the apparent ability to do so rested on an interpretation of all the information in a situation (Ekman, et_al., 1971). Williams and Sundeme (1965) did find that both the face and the voice tended to transmit consistent information about emotional state, although it cannot be inferred that the effectiveness would be the same for the two modes. Recent research has, in general, provided confirma- tion for the ancient idea that the human body transmits signs which can be interpreted with some confidence by an observer (Eisenberg and Smith, 1971). Anxiety. One specific emotion which has been thought to be of particular importance to communication is that of anxiety. Three types of anxiety which have been identified are reality or objective anxiety, neurotic anxiety, and moral anxiety (Hall and Lindzey, 1954). It is the first of these to which concern is directed as a component of the cHmmmunication process for all individuals. One effect of anxiety which 28 has been demonstrated is a detrimental influence on the learning pro— cess (Bugelski, 1956; Solomon and Wynne, 1954). This has been shown to be true especially in an unstructured situation and also for women (Smith, et al., 1956). Associated defensive measures have also been shown to be detrimental to the interactive process by placing severe constraints on the individual's functioning (Eysenck, 1961). Level of involvement. One might logically suspect that an indi- vidual's level of psychological involvement in a situation would be an important component in interaction, but again, there has been'little empirical research in this area. Consideration of areas which tinge on this factor, however, support the position. Studies of teacher charac- teristics (e.g., Getzels and Jackson, 1963; Ryans, 1960) have consis- tently indicated that the teacher who is highly involved, alert, enthu- siastic, and generally "with it" was both more highly rated in popularity by students and at the same time more effective as a teacher in promo- tion of student learning. Social psychological research on the "great man" theory of leadership (e.g., Borgatta, et_al,, 1954) have also indi- cated that the most effective group members are those showing high de- grees of involvement--which includes involvement both on task criteria and on sociability functions. Finally, studies of mother-child rela— tionships (e.g., Becker, gt_al., 1962; Boger, gt_al., 1969; Hess, et_al,, 1968; Yarrow, 1964) have suggested that the nature of the mother's in- volvement with the child is a critical factor in the positive social and cognitive development of the child. Physical behavior. Body actions with expressive content have been popularly referred to as "body language." Hand gestures are a familiar type of such expression, but many others are possible. Ekman and 29 Friesen (1969) suggested a fairly elaborate scheme for classifying body action. Their system includes five types of body expressions: (1) em- blems, (2) illustrators, (3) regulators, (4) affect displays, and (5) adaptors. Emblems are usually gestural equivalents of a word or phrase, while illustrators are directly linked to speech and "illustrate" what is being said orally. Regulators are actions which control oral inter- action, such as head nodding or raising the eyebrows, whereas affect displays are body expressions which indicate the emotional state of the communicator. Finally, adaptors are movements which are theoretically performed without one's awareness; they are carryovers from movements learned earlier in life which were part of a patterned activity with an instrumental purpose, such as wiping the mouth with the hand. While there has been little attention given to the use of physical behaviors in communication, one exception was a study by Carmichael, eg_al. (1937) on gestures as emotional indicators. Unfortunately, this re- search did not relate the behaviors noted to specific contexts or ac- tivities, while the reverse extreme of specificity has characterized studies of physical aggression. Primarily developmental in nature, such studies have tended to indicate decreases in amount of negative physical behavior with increasing age (e.g., Walters, e£_al., 1957). It might be suspected, however, that while the amount of gross physical behavior decreases, more subtle physical behaviors increase as either intentional or unintentional communicative acts. Observational studies of classroom behavior in particular (e.g., Anderson, gt_§l,, 1946; Thomas, 1929) have also included physical contacts of the subjects and have generally reported high observer reliability using behavioral categories. 3O TeachingfiProcedure Overview. There has been considerable philosophical debate on edu- cation as a process and somewhat more specifically on what is meant by teaching. In the simplest case, however, that of specific concern for instruction in a particular task, the definitional aspect is much less complicated. On the other hand, though, the problems of defining the process and procedures are as difficult as in any other situation. Un- til the l950's, most attempts to describe patterns of teacher behavior were fairly limited and dealt primarily with such areas as permissive- ness or subject-matter orientation. More recent research (e.g., Getzels and Jackson, 1963; Ryans, 1960) has suggested that there exists no single prescription for a set of characteristics comprising good teach- ing or a good teacher. Teaching obviously involves many aspects-- verbal and nonverbal, cognitive and affective, direct and indirect—~but the composite of the dynamic process is probably the most important con- sideration. Wallen and Travers (1963) have suggested that such behav- ioral patterns are derived from several sources: (1) teaching tradi- tions; (2) social learnings in the teacher's background; (3) philosophi- cal traditions; (4) the teacher's own needs; (5) conditions existing in the school and community; and (6) scientific research on learning. Concept specification. The organization used by the teacher in the presentation of the material has been thought to be an important determinant of the efficiency with which the desired learning is ac— complished. Carroll (1964) has pointed out that a search of the litera- ture for any comprehensive treatment of concept teaching is in vain, however, but his analysis reinforces the necessity of appropriate se- quencing by the teacher in presentation of material for concept learning. 31 Literature on the inductive and more frequently used deductive processes (e.g., Beberman, 1958; Michael, 1949) have not been conclusive in their results, although a commonality of both methods appears to be the need for an organization toward the specified learning (Henderson, 1963). The functions of orientation (e.g., Hess, g£_al,, 1968), summary (e.g., Reynolds and Glaser, 1964), and variety in teaching methods (e.g., Hughes, E£_3l:' 1959) have been suggested by educators and researchers. A basic principle of education which must be considered as well is that "since learners differ in their capacity to make the responses to be acquired, learning will be most efficient if it is planned so that each learner embarks on a program commensurate with his capacity to acquire new responses (Wallen and Travers, 1963, p. 496)." The situational con- text is obviously relevant, as are individual differences in both teach- ing and learning. Teaching method. The generalization that one picture is worth a thousand words has been often repeated but less frequently applied in education. Considerable research has focused on lecture versus dis- cussion methods of teaching (e.g., Casey and Weaver, 1956; McKeachie, 1962), with the discussion method typically coming out on top in terms of student attitudes if not knowledge of content as well. The rela- tively recent development of the media field, however, has increasingly focused attention on the importance of both aural and visual learning (e.g., Lumsdaine, 1963). Individual differences in learning styles and the relative effectiveness of these modes have also been noted (Dwyer, 1967). As Eisenberg and Smith (1971) pointed out in their analysis of nonverbal communication, transmitting any communication message involves lboth verbal and nonverbal components. It would appear to be logical, 32 then, that maximization of the teaching situation would involve the use of both verbal and illustrative modes of instruction. This principle has been suggested in research such as that of Hess, g£_al, (1968), in which a more refined level of teaching strategies was used by those mothers employing both verbal and nonverbal indicators in their teach- ing with their preschool children, with a subsequently higher level of learning by those children. Methodological Considerations A persistent concern of parent-child research has been with method- ology. Lytton (1971), in his methodological review of observational studies of parent-child interaction, described the range of methods which have been used along a continuum ranging from those farthest removed to those most closely involved with the primary object of study--normal parent-child interaction. The methods ranged as follows: (a) Questionnaires for parents or adolescents, (b) Interview of parents, (c) Observations of structured interaction in the laboratory, (d) Observations of unstructured interaction in the laboratory, (e) Naturalistic observations in the home. Throughout the history of parent-child research, the interview has been the most widely used method. This procedure still dominates the field, although increasingly sophisticated quantification techniques have been developed. After World War II, systematic observations of parent-child relationships began to appear (Lytton, 1971), but it has been primarily within the last decade that a significant number of such studies have been reported. 33 Two landmarks in the methodological history of parent-child re- search are particularly noteworthy. The first was the Fels study (Baldwin, gt_al., 1945, 1949), which was the first study in which sys- tematic observations of parent-child relationships were reported. A second development which seems particularly significant was the use of precoded behavior categories in data collection. This procedure was apparently used for the first time in 1951 by Merrill Bishop. Several considerations are important in evaluation of parent-child research methodology. Lytton (1971) described the following as par- ticularly important: (a) Amount of control of behavior and stimuli, (b) Methods of recording behavior, (c) Conceptualization of variables, (d) Range and type of behavior sampled, (e) Reliability of data, (f) Validity of data, (9) Acceptability of procedures to subjects. Comparison of the two basic procedures which have been used in parent-child research--interview and observation--suggests several im— portant differences. The degree to which each becomes relevant is a function, of course, of the unique circumstances of a particular study. The interview is typically the most convenient and economical means of collecting data concerning a wide range of behaviors (both past and present) for both parent and child. It permits acquisition of informa- tion about relatively inaccessible events in the life of the family. In addition, the interview can yield detailed descriptions of the child's behavior at previous points in time, his response to those situations, 34 and a variety of contextual data (Hoffman, 1957). The interview method has several potential handicaps, however, which can be summarized by Yarrow's (1963) observation that "mothers' interview responses represent self-description by extremely ego-involved reporters (p. 217)." Invalid reporting may occur through inaccurate re— call, misperception of the phenomena of interest, deliberate distortion to enhance social acceptability, or unconscious distortion due to psy- chological defense mechanisms (e.g., repression). The validity of con- cern for such errors has been confirmed by several studies (e.g., Haggard, et al., 1960; Yarrow, et al., 1968). Both unreliability for individual parents (correspondence of recall to earlier measures) and extreme variability across parents in degree of consistency have been noted (Yarrow, 1963). Errors of the first type can be appreciated, however, in view of the considerable demands on the mothers' abilities to make difficult discriminations and generalizations, as well as to exercise rather highly refined recall skills. Variability across sub— jects is also logical, since reports are essentially the nonstandardized observations of untrained raters. Several interesting anomalies have been reported in validity studies of parent interviews. For example, Yarrow, et al. (1968) found that mothers tended to recall boys as BEES and girls as lg§§_aggressive than described on initial measures. They also recalled more sharing by the father in the discipline role for boys but less sharing for girls than was reported on the initial measures. While this directional bias is psychometrically noteworthy for interpretation of results, it also bears implications for underlying theoretical issues. Perhaps even more disconcerting to the parent-child researcher is Zunich's (1961) 35 ‘finding that when comparisons were made between questionnaires on atti- ‘tudes and observations of behavior, there were even fewer significant (:orrelations than would be expected by chance. Observational studies have not been without their drawbacks either. 'The potential artificiality of the laboratory setting and/or the pres- ence of an observer cannot be ignored. Further, even trained observers Inay destroy, miss, or misinterpret the phenomena of interest in their reporting. Yarrow (1963) summarized a significant set of cautions for the researcher using observational procedures: The avoidance of such calamities requires awareness by the investigator of new kinds of data-sampling problems, judi- cious selections of the categories and framework for ob- serving, and the working out of a relation with the observed that is neither destructive nor unduly confining for the research. . . . (The researcher) must experiment with the appropriateness of different kinds of observational lens, from one extreme of attempting recordings of minute move- ments in nonpsychological terms to the other extreme of psychodynamic interpretations that are relatively inatten- tive to the details of the interactions themselves (p. 223). Obviously, no method is without its limitations in research as Complex as that focusing on parent-child interaction. Shakow's (1959) delineation of the central question remains: "How can one study human IPsychological phenomena scientifically with a minimum of distortion and EEthically with a minimum of trespass (p. 51)?" The best guidance is that 1Erom Lytton (1971) for the researcher to decide, in light of his basic Elims and opportunities for research, what kind of distortion he is will- i rig to tolerate . Family Relationships Sierent-Child Relationships Much of parent-child research is concerned with relations between IParent and child behavior, but there has been little research which 36 sspeaks to any cause and effect relationships. There has been an in- czreasing body of literature dealing with the reciprocal influences of }parents and children (e.g., Bell, 1968, 1969; Osofsky, 1970). Although few generalizations can be reached, a number of important studies have Ibeen reported which provide a basis for future research. Sears, et al. (1957) conducted a classic study of child rearing in which a large group of mothers were interviewed concerning the vari- ous typical childrearing dimensions. One of the major variables was the xnother's warmth, a factor which was pervasive in its effects on the development of the child's behavior. Punitiveness, in contrast with rewardingness, was found to be an ineffectual quality for a mother to ‘use. Although differences were found between the socioeconomic groups, these were interpreted in terms of circumstances rather than as neces- sary consequences. Working class mothers were found to be more puni- tive and more restrictive than were middle class mothers. The various (dimensions of this study were global in nature; as a basis for further research, however, they established a worthwhile foundation. A somewhat similar framework was used by McKinley (1964), who (Rallected data from high school boys about their parents as well as Ixeanalyzing a portion of the Sears, et al. (1957) data. Social class Variation in socializing patterns was of particular interest, a phenome- ruon that was suggested to be related to variation in satisfaction, ‘NTlich in turn would be related to aggression and to compensating achieve- ment in other areas of life, as well as to several secondary responses—- e.g., regression, anxiety, and submission. The socializing techniques Llsed by the lower class parents were found to be more severe and aggres- SSively oriented than those used by parents from higher status groups. 37 The father's role was more highly related to social class than was the ‘mother's, possibly because fathers were exposed to more variability in resources received and aggression experienced. These data suggested that fathers tended to lose authority in the family at the lower levels as a consequence of their social and occupational inadequacies. More recent research has also reinforced the idea that children learn interactional modes of response primarily from their parents. For example, academic achievement, leadership, and creative thinking of chil— dren have been positively related to warm, accepting, understanding, and autonomy-granting parent-child relationships (Walters and Stinnett, 1971). Elder (1963) also found that children more often model their roles after parents who are democratic than after parents who are either permissive or authoritarian. Winch (1962) has suggested that males tend to identify with the more functional parent, although this finding did not generalize to females. The traditional theory of sex-role identification has often been used to explain differences between masculine and feminine role behavior. As summarized by Lynn (1966), this theory suggests that the initial .parental identification of both male and female infants is with the Inother. Boys then shift from this initial mother identification to a Chalturally defined masculine role. Thus both males and females tend to inentify more closely with the mother than with the father, and both 1'iave more ongoing interaction with their mothers. While this theoreti- <2al orientation is focused on sex-role behavior, its relevance extends tlo broader consideration of the socialization of children in a culture, IParticularly as related to interaction between parents and their same- Sex or opposite-sex children. 38 Another study by Sears, e£_al, (1965) suggested that children of Jboth sexes initially adopt feminine—maternal types of behavior, and when the parent of the opposite sex rewards dependency the child tends to develop behavior qualities characteristic of that sex; finally, behavior considered to be masculine or feminine was reported to be more influenced by parental attitudes toward the control of the child's behavior than by any aspect of the availability of masculine or feminine models. In a study of parental antecedents of social aggression in young children, Delaney (1965) found that parental restrictiveness, rather than permissiveness, was positively related to child aggression, espe- cially in the relationship between paternal restrictiveness and aggres- sion in boys. Delaney also found that parents disagreed markedly con- cerning which parent assumed responsibility for disciplinary control of the child. Baumrind (1967) examined the relationship between certain types of preschool behavior and childrearing practices of parents. It was found that parents of the preschool children who exhibited the most positive behavior tended to be more consistent, more loving, more secure in the laandling of the children, and they were also more likely to accompany a (iirective with a reason. These parents also tended to be more support- iuve and to communicate more clearly with their children than the parents (bf children characterized by discontent or lack of self-control. The IDarents of the positive group tended to enforce the directives that tZhey gave to their children and to resist the child's demands. However, 1they did not overprotect the child. The parents of those children EShowing little self-control or self-reliance, however, tended to use Vvithdrawal of love and ridicule rather than power or reason as incentives 39 for their children. Fathers in this group were also characterized as weak reinforcing agents. Social class differences among parents have indicated that middle socioeconomic group parents tend to be more controlling and supportive of their children than lower socioeconomic group parents; they are also more likely to discipline their children by utilizing reasons and appeals to guilt and less likely to use physical punishment. Bronfenbrenner (1961) indicated that it was primarily at lower-middle class levels that boys tended to get more punishment than girls, and that girls got greater warmth and attention. With an increase in the social position of the family, direct discipline dropped off for boys and indulgence and protec- tiveness decreased for girls. Different types of discipline according to social class were also reported by Kohn and Carroll (1960). They found that middle socioeco— nomic group parents regarded it as of primary importance that a child be able to decide for himself how to act and that he have personal resources to act on his decisions. Working class parents, however, were more con— cerned that a child act reputably and that he not break proper rules. .Rcsen (1964) reported in addition that middle socioeconomic group parents ‘Vere more likely to discipline their children by using reason and appeals tr) guilt, at the same time using less physical punishment than did lower Scncioeconomic group parents. While the ways in which contingent conditions affect parent-child IRelationships are not evident, current research has suggested a number C3f'important dimensions in this area. Unfortunately, most of them are Silobal in nature, and both methodological and theoretical problems have Amended to make conclusive results impossible. Even at this level, 40 however, the research consistently attests to the importance of the parent-child relationships as factors in the children's development. P40 Cher-Child Relationships The overwhelming majority of research in the area of so-called parent-child relationships has actually been concerned with the child and his mother. At least in part, this has undoubtedly been because it is the mother with whom the child most frequently interacts and who is therefore his primary mediator of the world around him. Research has fOCUSed on such dimensions as teaching behavior, control mechanisms, la:1’1c_;l:lage, and personality characteristics. One of the most comprehensive studies to be undertaken was the Chicago study of Hess, et al. (1968) . Among the primary objectives of thi 8 study were the identification and measurement of the cognitive aSpeczts of mother-child interaction and the identification of maternal teae1‘1ing styles along with a study of their effects upon the children's c:c>gP‘=§2'Ltive behavior. Attention was focused on types of control strate- giQS . rather than the degree of restriction or regulation used by the m{D‘tl'lers to guide behavior. Results indicated that mothers who taught "10% t effectively used techniques helping the child to learn while enjoy- ing the task. Orienting, labeling, nonverbal focusing, high specificity, 130% t~response feedback, and praise were qualities characteristic of the “16% 7': effective mothers. Less effective mothers, in contrast, tended to Ca§t themselves in taskmaster-authority figure roles; poor organization anQ failure to give specific information characterized their teaching. (1968) examined the relationship between ratings of Milmoe , et al . Qt“ tions in mothers' voices and measures of their children's behavior. 41 Their results indicated that the ratings of anxiety and anger in the ‘mother's voice were associated with various signs of irritability and insecurity in the children, that daughters of the mothers with anxious voi ces tended to be more attentive and cautious in the various test 51 tuations at three different ages, and that ratings of warmth and pleasantness in the filtered condition of the mothers' voices were related to inattention to a human voice at thirteen months of age. This research suggests that the reactions to voice quality may be an early signal to the child and may have important implications in his de Ve lopment. Lelar (1970), in a study of mother—child interaction and language per formance in black preschool children, found that children who scored comparatively high in language performance tended to come from families in which the mothers were rated high in affectionateness, acceptance of the child, use of praise, rewarding of the child's independence, use of reasoning with the child, and encouragement of the child's verbaliza- tiQr‘S. Children high in vocabulary also were rated high in indepen- dence and initiative in verbal communication. Trends also indicated that those children who scored comparatively high in vocabulary came erm families in which the mothers were rated low in such factors as 118% of criticism and restriction of the child's independence. Bing (1963) also sought to examine differences in mothers' child- reQ‘ll‘ing practices and in their behavior with their children in relation t6 Qifferences in children's cognitive development. A comparison of mghh . . . . ers whose children were rated as high verbal Wlth those whose chil- de . . - l‘1 were rated as low verbal indicated that mothers in the former group 9Q VQ their children more verbal stimulation during infancy and early 42 childhood, remembered a greater number of their children's early accom- plishments, let their children participate in more conversations, pun- ished them less for poor speech, bought more story books for them, and criticized them more for poor academic achievement; they used anxiety arousal more in cautiousness training, showed less permissiveness with 01:) j ect experimentation, had more restrictions, and perceived their hus- bands as stricter than themselves. Mothers of the high verbal group of Chi Jdren were also found to be higher than low verbal mothers in all Categories of helping behavior, in pressure for improvement, in giving he lp after request by child, in asking the observer more questions, in giving more physical help, and in giving such help sooner. They were, however, also higher on withholding help and disapproval than were the low verbal mothers. These results indicated that the level of inter- action between the mother and the child was important in the child's de ve lopment of verbal abilities; discrepant nonverbal abilities seemed to be enhanced by allowing the child a considerable degree of freedom to ea-cperiment on his own. In a study of maternal reactions in the development of independence and achievement behavior in young children, Crandall, £31; (1960) fouhd that children who behaved independently with their mothers also tel—)‘ded to behave independently toward nursery school staff. Positive maternal reactions to the children's achievement efforts were not asso- Ci Q‘T‘Zed with general expressions of affection of the mothers. Mothers “'er frequently rewarded achievement in their children were less nurtur- an 7.: and less accepting and rewarding of help-seeking and emotional sugbort-seeking. Independence training and the rewarding of achievement V0 : Q positively related. The mothers who responded positively to their 43 children's approval-seeking in the home tended to have children who evi- denced more achievement efforts. Also, the mothers who more often re- warded the children's achievement efforts when the children did not seek approval for their performances had children who displayed strong achievement strivings outside the home. Several studies of mother-child variables have shown relationships between specific maternal behavior and dependency (e.g., Sears, et al., 1953; Symonds, 1939) . In addition, Hart (1957) found that authoritarian mothers tended to use more non-love-oriented behavior control techniques than did a comparison group. There were significant variations in the proportion of such responses selected as a function of the area involved. A number of other studies have considered the relationship between sslpecific childrearing variables and children's behavior. In many cases these also included reports of paternal attitudes or practices; the fact tJ'lat mothers are thought to serve as the primary model for their daugh- ters, however, has kept the focus on the mother-child dyad in particular. E‘a-\ther-Child Re lationships Very little research has been done in the area of father-child ll'1‘teraction. Most of the literature which has acknowledged that children h53-»‘9'e a parent other than the mother have done so either by bemoaning the SI-:’<‘3I.rsity of research involving fathers and/or by suggesting along with the conclusions of the study that more research needs to be done which il"“rolves fathers as well as mothers. LeMasters (1970) outlined a role analysis of the father in modern ““9 rica as follows: (a) The parental role is a peripheral role for the American male; t . his is not typically the case with females. I. “2 44 (b) There is no biological base or imprinting for the father role at: the human level as there is for the mother. (c) The human father is a mammal--and in the entire mammalian series there are only two fathers (human and wolf) that assume any major responsibility for their offspring. (d) The human father is also a primate; such animals assume the role of protecting the females and their offspring, but do not typically assume any responsibility for their daily care. (e) The father's parental role in the United States is peculiarly tied to the success or failure of the pair-bond between himself and his Wife - (f) There is a tendency to think that men and women are only super- ficially different-~and this difference is seen in their diverse reac- tions to parenthood. (g) The American father is poorly prepared for his parental role. Along the same line, Davis (1949) suggested that "the weak link In the family group is the father-child bond. There is no necessary a . . . . . . Ssc>CZLation and no easy means of identification between these two as t here is between mother and child (p. 400) ." In addition, however, he 0 ‘ . p 11'1th out that almost all human societies have evolved complex soc1al de - vices for binding the father to his children. While these devices do e - a(lst in our society, a review of the parent-child literature would 3‘19 . . . g’est that they are nominal rather than functional in many cases. Although father presence has received little attention by re- Sea“): Qhers, father absence has been a much more widely studied topic. The literature in general has indicated that father-absence is asso- Qi ated with lower masculine identification of male children. The longer r' . I’T' 45 the father absence and the younger the age of the male child when the father leaves, the greater the impact the father absence seems to have upon the child's masculine identification. There is some indication from the research that the responses of the mother to the father-absent boys play an important role in the son's masculine identification. Wt'lile studies of father absence have dealt almost exclusively with ef- fe cts upon male children, there has been some evidence to suggest that girls are also affected by the absence of the father figure as well (e - 9., Kagan, 1964) . Bell (n.d.) , in reviewing the empirical evidence Con cerning the significance of the father in the socialization process, concluded that the many variations in family forms in our society must be more adequately studied before definitive answers can be reached con cerning the importance of the father in that role. Ostrovsky (1959) has analyzed the changed role of the father in the present-day family by observing what happened to children who were to Varying degrees deprived of the father's presence. He suggested that, because of the infrequent presence of the father, the child t3’91ca11y finds that he is not able to rely on him as he can on his It"(D‘CIher for the fulfillment of his everyday needs. The father thus loses status in the child's mind as a representative male figure and as the symbol of authority. This role is in contrast to the traditionally do‘TL‘iJiant role of the male adult in other phases of society. Another problem suggested by Ostrovsky (1959) was that the nature of the typical father-child relationship often leads to a remote and al- rugst. embarrassed attitude of fathers, and men generally, toward young Q115~Il.dren, an attitude which is caused in part by the lack of sufficient Q *Derience in being around children. Another factor to which this 46 situation was attributed is the tendency in our society to repress af- fectionate playfulness and tender feelings on the part of men. Though both these situations may have changed to some extent in the past decade, the contrast between males and females in the general population is sti ll obvious. LeMasters (1970) suggested on the basis of general observations that variations in father performance exist between different socio- economic groups in our culture. He noted, for example, that there is Considerable evidence that the lower socioeconomic group father (the twenty or thirty percent at the poverty level) finds his parental re- SPOnsibilities overwhelming. Not only does he lack the money required to Support his family, but he also suffers from inadequate education, poor health (both physical and mental), slum housing, high rates of ma1‘5.‘tal instability, and a variety of other problems. It is also sug- ges‘Ced that, although no systematic studies of the upper class father in Oui‘ society have ever been made, their outside commitments force them to delegate much of their parental role to other persons. The Conclusion which is reached, then, is that the best American fathers are found at the stable blue—collar level and in the vast white-collar mldd le class. Cc>\Ir‘&rison of Mother-Child and Father—Child Interaction Although many studies dealing with parent—child relationships have concluded that there is a need for investigating the differences be- tween relationships of fathers and those of mothers with their children, thQ be has been little heed given to that advice. The majority of the S“Elatiies which have even included both parents (and the number 0f SUCh St udies has been scandalously small) have been concerned only with 47 relationships between the attitudes and/or behavior of the parents in relation to their child's behavior (e.g. , Baumrind and Black, 1967) or with agreement between the parents on either childrearing practices (e - g . , Delaney, 1965) or assessment of their children's behavior (e.g., Bron , et al., 1961) . The data from such studies were not even typically pre sented in such a way that parent comparisons could be made from them. In one of the few studies designed specifically to assess differ- ences between mothers and fathers, Emmerich (1962) used a mailed ques- tionnaire to which both parents in a middle socioeconomic group sample were asked to respond concerning the nurturance-restriction and power dimensions of the parental role. It was found that mothers tended to be more nurturant and less restrictive than did fathers. Parents also tended to exert more power toward their same-sex children than toward their opposite-sex children. In addition, parental nurturance-restriction tended to vary in a cyclical fashion as a function of the age of the child, particularly with male children. No other significant differences were found between the parents, but no recognition was given to the pos— Sibly confounding effects of the sex of parent by sex of child interac- tion or the sex of child by age interaction on the other comparison of i1'7";e.‘.l:‘est. In addition, the behaviors of the parents were self-reports only . with the additional limitation that there was no guarantee of the incl'SEDendence of responses for the parents from each family. Another study which compared mothers' and fathers' behaviors was conducted by Rothbart and Maccoby (1966) , with a focus on differential reaq‘tions to sons and daughters. The parents were asked to react to a Child‘s voice in a hypothetical situation; the same child, unidenti- fifib 1e by sex, was used for all trials, but some of the parents were 48 informed that it was a girl's voice and others that it was a boy's voice. The results indicated that mothers tended to respond more per- mi ssively toward the boy's voice and that fathers tended to respond more permissively toward the girl's voice. An unexpected finding was that mothers were more accepting of comfort-seeking in their sons than ’in their daughters. While the results of these hypothetical situations we re generalized to parental behavior toward their own children, an obvious limitation is the absence of any true behavioral measures as Validation. Jackson's (1956) study of verbal solutions to parent-child prob- lems also sought to compare mothers' and fathers' responses to a parental r0 1e questionnaire, using hypothetical problem situations. It was found tlll'lat mothers suggested methods of control that were more coercive than those suggested by fathers, but they also tended to vacillate more be- t‘veen mild and severe methods than did fathers. Again, there was no validity check of the responses to the hypothetical situations. In ac3~<3ition, no information was given concerning the independence of the I>aI‘ents' responses. Since there were quite unequal numbers of mothers and fathers in the sample, it is apparent that the parents were not in e"el:~y case paired, but there was no indication concerning the extent of it‘<3ependence of their responses. In his study of power assertion by the parent and its impact on the child, Hoffman (1960) found that working class fathers used more initial and unqualified power assertions than did middle class fathers, while differences between working and middle class mothers were indi- Qé-‘zed only in initial unqualified power assertion. Failure to find a d - . q~3li‘ect relationship between the father's use of unqualified power 49 assertion and child behavior variables led to the conclusion that the father functions like a catalyst in affecting the mother's influence kpeaikiavior toward the child, this behavior in turn affecting the latter's Epeauzrsonality development. Although the sample for this study was limited jgr). size, the behavioral nature of the data and the treatment of the fieazrtily as an interacting system make the research particularly relevant and unique in the area of family-child relations. A difference was noted between mothers and fathers in Hurley's (11.E)€35) study of parental acceptance-rejection in relation to children's Sirirtzeelligence. A mailed questionnaire and two follow-up interviews were used; the sample consisted of approximately two-thirds mother-father I?Eifii.zrs and one-third individual parents. An inverse relationship was found between the I.Q. scores of third-grade children and their parents' re Sponses to the acceptance-rejection measures, a trend which was most prominent between the scores of mothers and their daughters. Although fa~1i‘--1“u.==:rs' and mothers' acceptance-rejection scores were not differen- tifilly related to their children's I.Q. scores, daughters' I.Q. tended t“:’ IDe more closely linked to parental acceptance-rejection than did Sons ' I.Q. Within the framework of parental identification, Livson (1966) srt:\1fiuo xosue .. wouHCMm .uwxuo3 cofluosuumcoo .wau03 wuocsmq .. ucmumflmmm mumsomum .Hmnommu ~Houooo ..m umaflmumu .Hmesmnmoum Housmsoo .Hmc3o monw ..m cmaoficcomu nma .umflawum Hams .umucflmm ..m.w m m 59 OH mamzdm A4909 mmHmEmm oufl£3 mmmamaoofiz mmHmE mufisz mmmlmaovfiz mmamew xomHn mmmnmaoofiz mmame xoman mmmlmavoflz mmamsmm muHQS mmmlum3oq mmHmE muwn3 mmmIHmBOA mmHmEmm xomHn mmmlum3oq mmame xoman mmmluw3oq OHMCOMWWGMOHQ Hafiummmcmz Ho Hmcoflmmmuoum Iflfimm umxuoz cmHHme uwxuoz omwauxmuflsmm meuoz omaaflxmca vm>onEmcD Amm zv Monumm mo coaummnuoo macho ownmmanEmo m.oafl£o omamemw mumsumm mo coaummsooo VIM mqmflfi 60 Table 3-5. (g) The average number of children in the families represented in this sample was slightly less than three. Smaller families character- ized middle socioeconomic group families, while much greater variability in family size characterized lower socioeconomic group parents. The distribution for each demographic group is given in Table 3-6. (h) The average ordinal position in the family of the sampled children was slightly over two. The distribution for each demographic group is given in Table 3-7. (i) None of the sampled children had previous preschool experience at the time of the mother-child session. No child had as much as one term's preschool experience at the time of the father-child session. Data Collection Procedures All data collection was conducted as part of a series of four sessions designed both for orientation of children entering the pre- school program of the Institute for Family and Child Study at Michigan State University and for concurrent collection of behavioral data from the children and their parents. All data were collected between September of 1969 and December of 1970 and were therefore from families with children entering the three-year-old unit during those two years. The mother-child interaction was measured during the third session and the father-child interaction during the fourth session for each child. An attempt was made to have all four sessions complete before a child began preschool classes, but this was not possible in all cases. The primary exception was the fathers who participated during the first year, when the father-child interaction assessment was not initiated the same time as the other measures. The period of time 61 TABLE 3-5 Age of Sampled Children at Time of Parent-Child Assessment Session Sampled Child's Mother-Child Interaction Father-Child Interaction Demographic Group N Meana Standard N Mean Standard DeViation Dev1ation Lower-SE5 black males 4 38.8 2.9 4 39.3 2.5 Lower-SE8 9.8 . 4 . . black females 4 3 3 5 40 O 3 7 L°Ver'sgs 4 41.0 2.6 4 41.5 2.1 white males Lower-SE5 4 .O . 4 . . white females 39 l 6 39 8 l l Middle-SE5 38. . 4 . . black males 4 3 3 6 38 5 3 6 Middle-SE5 8.8 . 4 . 4.4 black females 4 3 3 8 39 3 Mlédle'SES 4 36.3 0.7 4 37.3 0.9 white males Mlqdle'SEs 4 36.8 2.0 4 38.0 1.0 white females - TOTAL SAMPLE 32 38.6 3.1 32 39.2 3.0 a . Age in months 62 TABLE 3-6 Family Size of Sample Families Number of Children in Family Sampled Child's Demggraphic N Mean Standard roup Deviation Lower-SE8 black males 4 3'8 2.0 Lower-SE8 black females 4 3-8 1.8 Lower-SE8 white males 4 3'0 1.0 Lower-SE5 white females 4 2-8 1-5 Middle-SE8 black males 4 1-5 0.5 Middle-SE5 black females 4 2-0 0-7 Middle-SE8 4 2.8 007 white males Middle-SE8 . .4 white females 4 2 3 0 TOTAL SAMPLE 32 2.7 1.4 63 TABLE 3-7 Ordinal Position of Sampled Child in Family Ordinal Position Sampled Child's p Deviation Lower-SE5 black males 4 2'8 2'5 Lower-SES black females 4 3'3 1'6 Lower-SE8 white males 4 2'3 0'8 Lower-SE8 white females 4 2'3 1'1 Middle-SES black males 4 1'3 0'4 Middle-SE5 black females 4 1'5 0'5 Middle-SE8 white males 4 4'3 1'3 Middle-SE8 white females 4 1'8 0'8 TOTAL SAMPLE 32 2.2 1.4 64 between the mother-child session and the father-child session ranged from one day to two and one-half months for the entire sample, with an average interval of 18 days (mean = 17.78; standard deviation = 20.80). All testing was conducted in a mobile laboratory unit equipped with videotape and one-way observation facilities. The room in which the session was held was 8' x 11' and contained no furniture other than the table and chairs used for the parent, child, and examiner. All sessions were videotaped for subsequent rating. Although the parents were not specifically informed prior to either session that it was being taped, a general orientation to the preschool program and the larger research project at a preceding session included an explana- tion that many of the sessions through the program would be videotaped for research purposes. A general release form for the family's par— ticipation in the program was signed at that time. All testing of both mothers and fathers was done by an experienced psychometrist who was working full-time with the larger research pro- ject through which these data were collected. She conducted not only the two parent-child interaction sessions but the two preceding sessions as well for each parent—child subject unit. Rating of the videotapes was accomplished by six graduate and undergraduate research assistants, among whom the total amount of rating was randomly distributed. All seven of the individuals involved in these assessment procedures were white females, and all received extensive training in the respective measurement techniques. 65 Measurement Mother-Child and Father-Child Interaction In order to obtain a natural yet standardized sample of dyadic interaction in a teaching situation, an unstructured setting with a de- fined task was selected. Criteria for the task included suitability for preschool children from diverse subcultural groups as well as a task level simple enough to be easily understood by the parents yet complex enough to present a challenging situation. The situation chosen involved teaching the child a simple two- dimensional sorting task. With the mothers, the Toy Sorting Task, adapted from the Hess-Shipman procedures (Hess, et al., 1968), was used. This task involves sorting nine items on the basis of color and kind of object on a round divided board with marked divisions. The items are three plastic spoons (one red, one yellow, and one green); three doll- house size chairs (one red, one yellow, and one green); and three toy cars (one red, one yellow, and one green).1 The Nine Block Sorting Task (Cunningham and Boger, 1969) was used with the fathers. Conceptually analogous to the Toy Sorting Task, this procedure involves sorting nine blocks onto three poles on the basis of color and shape. The blocks are plastic and include three circles (one red, one yellow, and one blue); three squares (one red, one yellow, and one blue), and three triangular wedges (one red, one yellow, and one blue).2 Each sorting task involved a three-part session. First, the exam- iner explained the task to the parent (in the absence of the child), l . . . . . . Complete instructions for administration of the Toy Sorting Task are given in Appendix I-A. 2 . . . . . . Complete instructions for administration of the Nine Block Sort— ing Task are given in Appendix I—B. 66 using procedures specifically designed ng£_to provide a teaching model. During the second part of the session, the examiner left the room and the parent taught the child the designated task. This period was allowed to continue for no more than forty minutes for any parent-child pair; any session exceeding this time limit was terminated by the exam- iner. The third portion of the session began when the examiner returned to test the child's task performance. The parent, although instructed not to intervene during this time, remained physically present so the behavior of both the parent and the child could be observed and rated. Parent-Child Interaction Rating Procedure A systematic observational rating procedure was chosen as a means of getting the most complete and objective assessment of parent—child interaction. The Parent-Child Interaction Rating Procedure (P-CIRP) (Cunningham and Boger, 1970) was selected for this purpose. The P-CIRP includes three parts--a general information section, a rating form (Section 1) for the parent-child teaching period of the session, and a rating form (Section 2) for the parent-child—examiner testing period. No observational rating occurs during the time the examiner instructs the parent in the procedures. Only Section 1 of the P-CIRP was used for the present study. This section includes the following variables for the parent: (a) Communication process--initiation, response, response object, reinforcement, feedback, and reward; (b) Verbal communication--verbalization, verbal fantasy, voice tone, verbal specificity, time orientation, task orientation, and 1A complete description of the P-CIRP is given in Appendix I-C. 67 nature of interference; (c) Nonverbal communication--affective tone, anxiety, level of involvement, and physical behavior; (d) Teaching procedure--concept specification and teaching method. The P-CIRP uses a time-event sampling procedure with twenty-second observation intervals. Each interval is rated as an individual unit, with each variable rated each interval. Opportunity is provided for recording up to two behaviors for variables in the communication process, verbal communication, and teaching procedure sections. The single cate- gory best characterizing that time point is used for the other variables. Initial training of observers in the use of the P-CIRP was accom- plished with the aid of videotaped interaction segments. The training program included extensive practice in using the rating schedule, clari- fication of variable categories and rating procedures through group dis— cussion, and resolution of discrepancies among observers. Following a minimum of one week's practice with the P-CIRP, ob- server reliability was determined for each rater with the use of video- tapes not previously viewed by that observer. Reliability was computed by two methods, one based on total blanks and the other based on total recorded positions.1 All observers attained both inter- and intra—rater reliability at or above the minimum recommended levels for both indi- vidual scales and the entire instrument (ranging from .60 to .90). Re- liability checks were made periodically to assure maintenance of accept- able levels and decrease slippage in accuracy as new raters were added. 1Procedures for computing reliability and minimum recommended reliability rates are explained in detail in the instrument description in Appendix I-C. 68 Hypotheses The primary objective of this study was to determine whether any differences existed between the interactions of mothers and of fathers with their preschool children in the teaching of a two-dimensional sorting task as measured by the Parent-Child Interaction Rating Pro- cedure (P-CIRP). General areas of concern were identified from the literature, but the absence of any directly applicable previous re- search indicated the formulation of hypotheses in the null form. There- fore, the following null hypotheses were tested: Ho: Ho: HO: Ho: There is no difference between the communication process of mothers and that of fathers with their preschool children in the teaching of a two-dimensional sorting task as measured by the P-CIRP. There is no difference between the verbal communication of mothers and that of fathers with their preschool children in the teaching of a two-dimensional sorting task as measured by the P-CIRP. There is no difference between the nonverbal communication of mothers and that of fathers with their preschool children in the teaching of a two-dimensional sorting task as measured by the P-CIRP. There is no difference between the teaching procedure of mothers and that of fathers with their preschool children in the teaching of a two-dimensional sorting task as measured by the P—CIRP. Secondary hypotheses were also formulated in relation to selected demographic characteristics. The following null hypotheses were tested: Ho: Ho: There is no difference between the dyadic interaction of mothers with their preschool children and that of fathers with their preschool children in relation to socioeconomic group. There is no difference between the dyadic interaction of mothers with their preschool children and that of fathers with their preschool children in relation to race. 69 Ho: There is no difference between the dyadic interaction of mothers with their preschool children and that of fathers with their preschool children in relation to sex of the child. Ho: There is no difference between the dyadic interaction of mothers with their preschool children and that of fathers with their preschool children in relation to the interactions among socioeconomic group, race, and/or sex of the child. Each of these secondary hypotheses was tested for the four areas of interaction delineated by the primary hypotheses. This resulted in thirty-two hypotheses which were tested (four areas of interaction times eight sources of variation--i.e., main effects plus interactions of factors for each). Measures of all variables were with the use of the P-CIRP. Operational Definitions The dependent variables of interest were operationally defined as follows: (a) Initiation is the relative amount of emphasis on each type of initiation (i.e., unqualified power assertion, qualified power assertion, bargain, choice, neutral initiation, failure to continue interaction) during the teaching portion of the parent-child session as measured by the P-CIRP. (b) Response is the relative amount of emphasis on each type of response (acceptance, rejection, ignoral, ongoing interaction, failure to respond because of no opportunity from other person, failure to respond because of no immediately preceding communication) during the teaching portion of the parent-child session as measured by the P-CIRP. (c) Response object is the relative amount of emphasis on response to the child as an individual in contrast to his behavior or perform— ance during the teaching portion of the parent—child session as measured 70 by the P-CIRP. (d) Reinforcement is the relative amount of correct or appropriate responses to the child's behavior during the teaching portion of the parent-child session as measured by the P-CIRP. (e) Feedback is the relative amount of emphasis on overt feedback through responses during the teaching portion of the parent-child session as measured by the P-CIRP. (f) Reward is the relative amount of praise in contrast to criticism given as feedback during the teaching portion of the parent-child session as measured by the P-CIRP. (g) Verbalization contains both a quantitative and a qualitative dimension. It is (l) the total amount of verbalization which occurs during the teaching portion of the parent-child session as measured by the P-CIRP, and (2) the relative amounts of emphasis on types of ver- balization (i.e., orientation, evaluation, control over others, deci- sion, tension management, and integration) which occur during the teach- ing portion of the parent-child session as measured by the P-CIRP. (h) Verbal fantasy_is the relative amount of fantasy verbalization which is used in the teaching portion of the parent-child session as measured by the P-CIRP. (i) Voice tone is the average level of affect conveyed by voice tone during the teaching portion of the parent-child session as measured by the P-CIRP. (j) Verbal specificity is the relative amount of specific verbal- ization in contrast to global verbalization which is used in the teach- ing portion of the parent-child session as measured by the P-CIRP. (k) Time orientation is the average level of time orientation 71 (i.e., past, present, future) used during the teaching portion of the parent-child session as measured by the P-CIRP. (1) Task orientation is the relative amount of task-oriented verbalization which is used in the teaching portion of the parent-child session as measured by the P-CIRP. (m) Nature of interference is comprised of two aspects. These are (l) the relative amount of non—task verbalization which is initiated by the parent in contrast to the child or outside factors during the teaching portion of the parent-child session as measured by the P-CIRP, and (2) the average remoteness from the task (i.e., non-task use of task materials, immediate environment, nonsituational factors) of non- task verbalization during the teaching portion of the parent-child session as measured by the P-CIRP. (n) Affective tone is the average level of affective tone (i.e., negative, neutral, positive) shown by the parent during the teaching portion of the parent-child session as measured by the P-CIRP. (0) Anxiety is the average level of anxiety (i.e., low, moderate, high) demonstrated by the parent during the teaching portion of the parent-child session as measured by the P-CIRP. (p) Level of involvement is the average level of involvement (i.e., passive, moderate, intense) of the parent during the teaching portion of the parent-child session as measured by the P-CIRP. (q) Physical behavior contains both a quantitative and a qualita- tive dimension. It is (l) the total amount of physical interaction which occurs during the teaching portion of the parent-child session as measured by the P-CIRP, and (2) the relative amounts of positive and negative physical interactions which occur during the teaching 72 portion of the parent-child session as measured by the P-CIRP. (r) Concept specification is the relative amount of emphasis on each type of concept instruction (i.e., general orientation, general summary, new concept, different aspect, new approach, ongoing method) during the teaching portion of the parent-child session as measured by the P-CIRP. (5) Teaching method is the relative amount of emphasis on each type of teaching method (i.e., demonstration, explanation, illustration) during the teaching portion of the parent-child session as measured by the P-CIRP. Data Reduction and Analysis Data Processing All data were recorded on computer coding forms by a coder trained in these procedures. This coding was then quality checked in its en- tirety by a research assistant trained both in coding and in the P-CIRP rating procedure, thus permitting a check on the logical consistency of the rating as well as the maintenance of quality control in coding. All data were punched on computer cards and mechanically verified. A transformed data deck was generated with a computer program (M. Boger, 1972) which reduced the individual observations to category and variable frequencies for each subject for each raw variable. A second transformation program was then used to generate a new deck in which any category frequency which was "0" was changed to "1." This trans— formation was done because the frequencies were ultimately used for calculating log transformations, which would have been mathematically impossible to compute with the zero frequencies. It was assumed that 73 the relatively small changes in scores which resulted from this proce— dure were insignificant and were well within the range of errors of measurement. Using the second transformed data deck, scores were calculated within each subsequent analysis program for each subject on each variable as indicated by the operational definitions. Finally, logit transforma— tions were computed within each program for all proportion scores for the qualitative variables of interest. Use of the logits as input variables was chosen as a means of con- trolling the instability of variances which results from the use of standard proportions. Specification of the logits in the dichotomous case followed the form p logit = log — , q where p was the frequency of occurrence of an event and q was the fre- quency of nonoccurrence of that same event. For the polychotomous case, a series of logits was specified in the form a logit l = log‘——- ; n b logit 2 = log -——-; n m logit x = log-—- ; n where a, b, . . . n are the frequencies of occurrence of categories a, b, . . . n, and n is the category with the greatest expected rate of occurrence . 74 Using the above procedures, thirty-nine input variables were identified within the four categories of interaction as defined in this study. The nature of these variables is indicated in Table 3-8. Statistical Analysis A multivariate analysis of covariance (MANCOVA) Fixed Effects Model was used to test the hypotheses of interest. All analyses were carried out using the Michigan State University Control Data Corpora- tion 6500 and 3600 computers. Finn's (1967, 1971) basic multivariance program was used for the analysis, with adaptations for the MSU systems by Wright (1970) and Schmidt and Scheifley (1972). A three-way design over subjects model was used for the analysis. The independent variables in the design were race, socioeconomic group, and sex of the child, with two levels for each of these factors. The unit of analysis was the family of the sampled child, and the dependent variables were difference scores obtained by subtracting the father's score from the mother's score for each variable of interest. The MANCOVA model was chosen because there were multiple dependent variables which were seen as potentially interrelated, and the phenomena of interest constituted a global behavioral array rather than a uni- dimensional construct. In the absence of any evidence to the contrary, the assumptions of a normally distributed pOpulation, equal pOpulation group variances, and independence of errors were considered to have been satisfied. The parent difference scores were used as dependent variables in the analysis rather than including sex of parent as a design factor because of the lack of independence of observations involving the c m moH o m Ame c m 00H moccaHmEoo mo mUHO£o um fine 6 cm>am ma COmumm Hmtuo mcu soanz 0 CH COauom mo coflummmmSm mo aucmswwum n m cm .llll moH pumBmu v 6 mo ummwo Lua3 u0a>mpmb ace cofiummm m on u Imam no .Hmwmmm .CHMGMMQ mo >ucmsqmum n m w c ceauuwmmm umzom twawflamoq mo wocmsvmum u m m 00H :3 cofluummmm wm3om Uwamaamsqcs mo mocmsvmum u 9w 5w ADV 7 c seaboauacfi Hmuuomc wo xocmsqmwm u m cm Illl.moH coHuomumDCa mosaucoo ou musfiflmm mo >ucmoqmum H mm mm Ame coflumHuHcH meUOHnH COHUMUHCSEEOU Ammabmflum> meHOMmcmqu machexm mo cofluacmeo acaumoamaucmso webmuum> mflwxamc< ea poms mmahdaum> mmHoum mo coflumoawflucmso wlm mdmdfi 76 «osmEuowuwm Ho uofl>mnm£ m.pafl£o 0:» cu uncommon mo wocmswmum Hmsofl>flpcfl cm mm oHHno may 0» mmcommmu mo zoomswmum m mOH uomnbo mmcommmm cemumm Hmzuo map we coflumowGSEEoo one upmownm xamumflomEEa on ma mwmzu wmsmomh xUMQommm m>Hm o» mHSHHMM mo xocmsvmum coHuomumuca mcfiomco mo >ocwswmum Hmuocma mo xocmsqmum DH 30Ham uoc pep acmumm umsuo mg» mmsmomn ocommmu ou mHSHaMM mo xocwovmum COmwmm Hmnuo mo coHu IMUHCSEEOU m0 mocmumwoom mo >ocmovmum camuwm umzuo mo COADmUHCSEEoo mo cofluomflmu mo >ocmsvmum 00H m @0H Illl.moa mod Ame Ape Roe ADV Ame wmcommmm machesm uo coauucaumo Ammanmflum> pmeuoumcmuev coaumofiMHucmso manuaum> Aposcflucoov mlm mqmdfi 77 oumzmu no mam: mm mo mm>am no .msumum m.umnuo mwmwmu .muaumo m + m we IflHOm m30£m adambnm> poanSm msu now£3 m zd Am H a Ca mafia: use» omamEmm mo xocmsgmum u A m m + m A we COADMNHHmnuw> coHuMUHCSEEoo ampum> mo xumnpwmm uuw>oo mo wocmswmum n m xUMmemm pum>o mo wocmsvmum u mow mo mo 0 m + m Emaoflufluo mo mocmswwum u m m u m mmamum mo mocmsvmum u m m I m pumBmm macaumoacseeoo . no on mmmcommmu unm>oo mo >ocmsvmum u m mu m 00H macaumoacDEEoo . . m0 m0 ou mmmcommmu uum>o mo wocmowwum u m m xomnpmmm padre mo uoa>mnmb on mmcommmu + upwaumoummm no uomuuoo mo hocmsqmum u m + m mod paaco mo H0H>mnmn ou uncommon mutaMQOHQQMCa no uomuuoocfl mo xocmsvmum n um um unwEmUHOMGme AmeQMwum> omEMowmcmqu m o o cod a m a besm m .u.cau o coaumuauaucmso mannaum> prscwucoov mum mamda 78 coflumEuHmcou Ho .coHuflummmw .coHumEHowcfi .coflumucmauo MOM mxmm >Hamnnm> pomflnsm mnu coagz cfl muacs mEHu pmHQEMm mo wocmsqmum mfiuamcou Ho .mmflm lawmao .mummmwu .coflumEu0mcH no cofiumu Icmflmo mw>Hm maamnum> wothSm mzu £Ufl£3 cH muflcs mEHu omHmEMm mo >ocmsvmum Lmfl3 no mcHHmmm mmmmmum Ixm no xmflmxamcm Mo .cofiumsHm>m .mCOa Icflmo mm>Hm >HHMQH0> womthm on» noHLB CH muacs mEHu pmHmEMm wo >ocmsqmum uwcuo wow >Eocousm mcH>HmEH .mcofluomnflp Ho mQOHummw loom mm>flm >aa6hum> uomfln5m we» LUH£3 CH muaco mEHu pmHQEMm mo xocwswmum mmHHmEoo no .muso Icoo .mpcmumumpcs .mocmummoom m>Hmmmm mBOLm .mmmumm >Hamnuw> powhbom wnu noan3 CH muacs mEHu pmHmEMm mo xocmdwwwm COHuommmapmm mzonm no .mcmsma .mmen .coflmcmu mommmawu >Hawbnm> uommnsm mnu £0fi£3 ea muaco meflu pmHmEMm mo >ocm9qmum ma m0 m0 m9 m¢ m0 @0H m4 do me mo 00H m4 Dm md m0 00H mo 04 mm m0 90H Em m9 Aomscflucoov coaumnaambum> maObssm mo coHuecHqu AmmHQMaum> pmewowmcmuev coHDMOAMHucmsO webmaum> Aemsceucoov mum mamas 79 coHnmNaaanm> manna x z: 24 am mo nflaamucncs cm moanese uuwmnsm on» soars z: m + m + w + m + m a“ muwcs use» omHQEmm mo >ocmdenm m mm mm mm mo + mi. n+ m+ m+ mamm muummmm HO mflcmwmfl m0 Dm Ufl Aw no .moumum m.nm£uo mmumemo .Emficommu m + m + w + m :26 m3onm waamnnm> noanSm mnu £0fl53 ea mufics meHu omamEMm mo >ocmsvmnm zen 22m + z¢w + 9mm + mom md ma md m0 =6nmnn 66 + u + u + u + u + poo: msmnonufi3 no .mamc nom mxmm .con mo um um Am ucuu mzonm haamnnm> noonnsm 0:» songs m + m + m + m . . m mod cfl mafia: mafia omHQEMm mo wocmsvmnm a w x z: m + m + mam: modosnufi3 no am we m< .hufiamenou no coHuowflwn w>fimmmm mzonm m + m + m + .mmonommwo >Hamnnm> noanSm on» nownz mo mm mo mo cw wanes 08H» pmamemm mo >ocoswmnm m + m + m + m + Dm Dd m8 cofluom mo m>63 w + m + w I a oabwmmom no .cOAuomnfio .mcofiumwmmsm now mxmm xaamnnm> noonnzm may nown3 22m + ca mafia: man» omHmEmm mo aocmsvmnm mam Hm mo mt m + m + m + mosaamom mo scammonm 1x0 no .mnmhamcm .co«um9Hm>m .cowcnmo + mew + «Om + now mxmm waamnnm> nomflnsm on» soas3 mé ow 04 me Aboacwucoov aw muwcs man» omHmEum mo hucwswmnm m u + m + m “by scauuufiambnw> mm Ann meno mcmn maonexw mo cofiuaCauwo A Hammauwowwaucwso Bo manmwnm> Avmscwucoov mlm Manda 8O Amnmbmn onunummm mooaocH no: 0o ..m.mv Hmnon wnm £0H£3 mcoflumunHMQnm> ompnoomn mo >ucmsqmnm Amc0nuownnp unoaamxm n0\pcm mHQOH mosaocn ..m.mv camaowmm mnm scans mcoHumnnHmbnm> pmpnoomn mo >ocwsgmnm >unuamauwmm Hmnnm> meow mono> >3 om>m>coo ma uomwmm m>auamom m nonc3 new mcoHumnaam9n0> omonoomn mo xocmswmnm pom: ma .uomwmm oc mcH>m>coo .mcon woeo> Hmnusmc m £UH£3 n0m mcoflumuaamnnm> pmpnoomn mo wocmsqwnm mcou moao> >b pm>m>coo me powwow m>flumomc m L0H£3 n0m mcoflumuflamnnw> Umpnoomn mo xocmsvmnm + II m + ow + w i+uc m + rout a + i-uv n mcou mono> mcoflnmuflamnnm> >mmucmMcoc mnm £0fl£3 mCOaUMNHHmnnm> ompnoumn mo mocwsvwnm macaumuflambnm> xmmucmu mnm nuan3 mcoHumufiHMbnw> owpnoomn mo xocwsgmnm mz .IIIII men xmmucmm Hmbnm> coaumuaamnnm> >cm m>am no: mmoo noanSm man Lonn3 CH manna man» pmeEMm mo xocmsvmnm Apmscnucoov coHuMNfiHmnnm> mnobenm no conuncnumo Ammanmanm> omen0mmcmnav conumonnnncmso mnb6nnm> icoscnucooe mum mamas 81 mo mm: IHHMbnm> mamanmume xmmu on» meUIcoc m>Ho>cH noa£3 mcoHumu omucanOIxmmuncoc mo wocmsvwnm mocmnmm nnmucw woamuso Eonm mnm non£3 mcoaumn naamnnm> poucmHnOIxmmuucoc mo mocmoqmnm omucwwnonpaflzo onm gown: chaumu uaamnnm> pmucanOuxmmuncoc mo xocmsvmnm omumauflcauucmnmm mnm zoflnz mcoHumn Inamnnm> omucmanOIxmmuucoc mo aucmswmnm mocmnmmnwuca mo wndumz omucmanonxmmu mnm nuan3 mCOAUMNHHMQnm> pmpnoomn mo mocmsgmnm pwucwnnonxmmUIcoc mnm noa£3 mCOauMNHHmnnm> pmpnoomn mo mocmnqmnm coflumucmano xmme mnouom mcu on nmmmn nuanB mcoflumufiamnnm> pmpnoown mo uncongmnm ucmmmnm man on nmmmn nuan3 mcoflumuaH6bn0> pmpnoumn mo xocmsgmnm owed on» on nmmmn noa£3 mcoHDMNAHmnnm> pmonoomn mo >ocmsvmnm c0aumucmflno wEHB mnobenm no conunenuma E n o m m + m 0 m c w E Anvm+invm+xta 3V 0 O U m m + m mod m m m n not a m e m moH z 92 e m m one mm on mm m u + n + eat a 8 mm a u A no m + A no N + lemme n Ammanmflnm> omenommcmnev cannmonnnucmzo mnbmnnm> AcmSCHucoov mum mqmda 82 Hm>ma >u0flxcm noes mm: ucmnmm msu £0H33 mcfl undo mHm>nmucfl meHu pmHmEmm mo xocmsgmnm Hm>wa xumfixcm mumnwpoe mm: ucmnmm may noazz meanso mam>nwucfl wEwu omHQEMm mo wucmsgmnm Hm>ma mumflxcm 30H mm: ncmnmm wen soanz meansp mHm>nmucH mafia pmamewm mo wucmsgmnm m A Z w + m + m u u rate m + isms m + name n xumaxc¢ econ m>anowmmm m>aufimom nuns mam>nmucw can» pmHQEMm mo xocmswmnm meow m>nuummmm Hmnusmc suaz mam>nmucfi wan» pmHQEMm mo >ocmswmnm moon m>auommmm m>aummmc suaz mam>nmucw mEHu UwHQEMm mo xucmavmnm "M +M+OM+I m u u r+nc m + rent a + 1-6c n mcou m>fluummwd coflumoacsEEOO Hobnm>coz mnouomw HMGOnumsuamcoc co maven £0H£3 mcoHumN Iaamnnm> owucmfinonxmmulcoc wo xocwswmnm ucmEconH>cw gunfiomEEH map so m300m goats mcoHumu Iaamnnm> pmucmflnOIxmmuucoc mo mocmsgmnm Aconcaucoov mucmnwmnmuca mo mnoumz mnobenm no conuncnnma AmmHQManm> owenommcmnev connmonnnucmso mnbmnnm> Apmscflncouv mum mqmda 83 omonoomn ma uomucoo Hmoamwcm mcw>ao> ucH n0w>mnmn Havamwsm m>aummmc songs ca mam>nmucfi mafia omHmEMm mo hocmswmnm m2m 22 oz mo dd mm m + w + m + m + m pmonoomn ma uumucoo Hmoemmcm mcw>ao>cfl no: noa>mnmn Hmonmxnm m>flummmc nown3 22 oz mo md mm ca mam>nwucfl 08H» thmEMm mo mocmavmnm mow m + m + w + m + m moH x 22 omonoomn ma uomucoo Hmoflmwnm ocfl>ao>cfi m + m + was noH>mnmn Hmonmmnm m>fiufimom nuanz dz mm mm mm ca mam>nmucH menu omHmEMm mo wocmswmnm mew m + m + m + m H pwonoomn ma uomucoo Hmonmmnm mea>ao> 22m + In“ nofi>mnmn HMUflmhnm m>nuamom now£3 mm c« mam>nmuca 08H» omeEMm mo hocmsvmnm m AZH + mom + mdm + mmw Ame noa>mnmn Hmoammnm ucmEm>ao>cw mo Ho>mH m>flmmmm 6 mm: ucmnmm on» nonz3 ca mam>nmqu menu pmHmEMm mo >ocmsgmnm mm ucmEm>Ho>cw mo Hm>wa mumnmoos 0 mm: ucmnmm on» goers ca mam>nmuCH man» pmHmEMm mo wocmzvmnm mm acmEm>HO>CH mm + mm + an m0 Hmbma swan m mm: ucmnmm on» sown3 H ucmEm>Ho>cH ca mam>nmuca mafia pmamEmm mo xocmavmnm m Aamv m + Ammo N + Ammo A no Hm>mq m maOQEhm mo coauacwmmo A manmfinm> oufinommcdnav mHannm> cofiumUHMflucmso Rowenaucoov mum Manda 84 umwocou 08mm may meanommu m4 now tonnes meaomco mo mm: mo mocwswmnm 20m w moa mumwocoo no xmmu men mo mom Ape Icnmma mo xnmEEdm Hmnmcmm mo mocmsvwnm w mé m mummocoo no xmmu mcwcnmma moH on cowumucmfino Hmnmcmm mo mucmsvwnm COM COM on ammocoo 056m 0:» meacommu on zomonm . mm mm Ian 30: M NO canuooponucfl mo >ocmsvwnm w m mo a noa>mgmb mdm Abe mcfinommu an ammocou 056m we» mo nummmm ucmanMHp m :0 meanUOM mo >ocmsvwnm mm£mn mcflnomwu cfi ammo 2 02 H coHumoHMflowmm Icoo 30a 0 mo cenuosponucH mo >oc09wmnm O m m Ame ammocoo endomoonm mcflnomme owpnoomn ma n0n>mgmb Hmonm>nm 0c nonnz ca mHm>nmucfl mEflu pmeEMm mo >ocmsqmnm xm pmpnoown 22m + mzw + mam + adv + mmm ma noH>mnmb Hmonm>nd Hmnuomc bon£3 zz Apmncfiucoov ca mam>nmuca menu pmHmEmm mo wocmsgmnm m AQZM + home I Adam + ammo any now>mswn HMUfimanm mache>m mo cowuficwwmo AmmHQManm> omenommcmnav conumonunucmso onbmnnm> Avwncflucoov mlm mqmda 85 ponume mcflnomwp m m we ammocoo mo coflnmnumSHHfl mo >ocmsvmnm 00H H ”on“ 0:.— M Any monnommu m mm COHumcmamxw mo mocmswmnm m m ponume man mod Icommu 6 mm coHnmnumCOEmp mo >oc05vmnm 0% Ame ponumfi meanomma mfi w Apmocfiucoov OOH acnumowwwommm 20w Ame ummucou maone>m mo cOwufianwo Ammanmfinm> omenowmcmnev manmfinm> QOaumowmwucmso prncfiucoov mlm mamme 86 mothers and the fathers. Although the child was not a subject in the analysis, the same child was involved in the two interaction sessions through which the parent data were collected for each family. Recogni- tion of the reciprocal influences in the socialization process (e.g., Bell, 1968) indicated the importance of considering the potential interdependence of the two measures. In addition, use of the difference scores gave a more direct test of the secondary hypotheses for differ- ences between parents in relation to the demographic variables. Preliminary analyses were carried out using a repeated measures model in which the mother-child and father-child interaction observa— tion constituted the repeated measures dimension. A transformation matrix was applied to obtain the contrasts of interest, those between the mothers' and fathers' scores for each variable. Comparison of the results obtained using this analytical procedure with those obtained using the mother-father difference scores as dependent variables indi- cated that the two procedures produced consistent results. The latter procedure had the additional advantage of permitting inclusion of co— variates in the analysis procedures, so this model was selected for testing the hypotheses of interest for this study. An implication of using the difference scores which had to be con- sidered in interpreting the results was that the test of any effect was actually a test of the interaction between that effect and sex of the parent. That is, a test of any main effect was actually a test of the interaction between that factor and sex of the parent; a test of an indicated two-way interaction was actually a test of the three-way interaction between the two designated factors and sex of the parent; and a test of the three-way interaction was actually a test of a four—way 87 interaction between the three factors and sex of the parent. The test of primary interest, therefore, was the test of the grand mean, or the difference between the mother-father difference scores and zero. The primary and secondary hypotheses were tested using four three- way multivariate analyses of covariance. The independent variables were socioeconomic group, race, and sex of the child. Age of the child and length of the teaching session were the covariates. Dependent variables were the mother-father difference scores for the various transformed variables grouped into the areas of communication process, verbal commu- nication, nonverbal communication, and teaching procedure. Since this study was designed to be exploratory in nature, a liberal experiment-wise alpha level was chosen. Although this value was not computed, alpha was set at 0.05 for testing each multivariate hypothesis. In addition, an alpha level of 0.01 was selected for testing each indi- vidual variable, while variables with alpha levels through 0.10 were considered to be indicative of trends which could be given further con- sideration. Although arbitrarily selected, these alpha levels were thought to be stringent enough to exercise a reasonable control over random significance yet liberal enough to detect potentially discrimi- nating variables which could be isolated for further study. CHAPTER 4 RESULTS In the beginning was the 6500, and the 6500 was with Absolute Octal and the 6500 was Absolute Octal. All things were computed by it and without it was not any- thing computed that was computed. In it was Output and the Output was the answer of man. And the Output came to the helproom and the helproom understood it not. And there was a man sent forth from the 6500 whose name was Hollerith. The same came for a wit- ness, to bear witness of the Output, that all men through him might believe. He was not himself the Output, but was sent to bear witness to the Output, that was the true Output which enlightens every man which cometh into the Computer Temple. He was in the Computer Temple and the Computer Temple was made through him, and the Computer Temple knew him not. He came unto his own, and his own received him not. But as many as received him, to them he gave the power to become programmers of the 6500, even to them that believe on his name, which were born not of blood, nor of the will of the flesh, nor of the will of man, but of the 6500. And the Absolute Octal was made Scope and dwelt among us, and we beheld its glory, the glory as of the latest begotten of the 6500, full of relative addresses and of dump. Hollerith bore witness concerning Scope, crying, this it is of which I spake. It which is to come after me has been set above me, because it is beyond me. And of its diagnostics we have all received, and error for error. For the method was given by Absolute Octal, but rela- tive addresses and dump came through Scope. And no man has understood the 6500, the latest begotten Scope, which is in the memory banks of the 6500, it has declared it. And this is the record of Hollerith. --The Gospel According to Hollerith, from a scroll found at Michigan State University, of unknown authorship, as modified by the RCC O 88 89 Communication Process Overview The multivariate analysis of covariance indicated that there were no significant differences between mothers and fathers on the overall mea- sure of communication process as measured by the P-CIRP, although a trend was noted for differences in relation to socioeconomic group (Table 4-1). In addition, inspection of the univariate F ratios indicated several individual variables which reflected area of potential differences. The multivariate analysis of covariance for communication process was computed using the three design variables of race, socioeconomic group, and sex of child. Covariates included length of the teaching session and age of the child. Dimensions of communication process used as criterion variables included type of initiation, type of response, re- sponse object, reinforcement, feedback, and reward. The fourteen scores for each parent which were used as measures of these dimensions included thirteen logit transformations of category variable ratios and one adjusted mean score, as delineated in Table 3-8. The difference scores for each mother-father pair on these fourteen variables were the depen- dent variables used in the analysis. Test of Three-Way Interaction The highest order interaction to be tested was that for race by socioeconomic group by sex of child. A multivariate alpha level of 0.7246 led to the conclusion that the null hypothesis for this three- way interaction could not be rejected (Table 4-1). This conclusion of 0 l ' . . . Cell means for all demographic groups on all communication process variables and covariates are given in Appendices II—A and II-E, respectively. 90 TABLE 4-1 Summary of Multivariate Tests of Differences between Mothers and Fathers on Communication Process Dimensions of Parent-Child Interaction Multivariate Test Statistics Hypothesis F ”23:22:.“ :52: Grand Mean 1.7286 14 and 9 0.2062 Race 0.3871 14 and 9 0.9459 Sex 0.9403 14 and 9 0.5574 Socioeconomic Group 2.4247 14 and 9 0.0923 Race x Sex 0.6377 14 and 9 0.7828 Race x Socioeconomic Group 0.9242 14 and 9 0.5684 Sex x Socioeconomic Group 0.6681 14 and 9 0.7595 Race x Sex x Socioeconomic Group 0.7246 14 and 9 0.7160 91 no significance was further supported by inspection of the univariate alpha levels. None were significant at the chosen level, and only one reflected a potential trend (Table 4-2). That variable, relative num- ber of interaction units for which there was a failure to continue interaction, had a univariate alpha level of 0.0354. Test of Two-Way Interactions Since potential confounding by the three-way interaction was not judged to be significant, the three sets of two-way interactions were considered next. In each case, there were no significant differences on the overall tests (Table 4-1); therefore, none of the null hypotheses relating to these interactions were rejected. Multivariate alpha levels for the three hypotheses were 0.7595 for the sex of child by socioeco- nomic group interaction, 0.5684 for the race by socioeconomic group interaction, and 0.7828 for the race by sex of child interaction. The conclusions of no significance were further supported by in- spection of the univariate alpha levels. None of these values were significant for any interaction, although several trends were indicated for individual variables. For the sex of child by socioeconomic group interaction, two variables reflected trends which might be considered (Table 4-3). The first, relative number of responses which were rejections, had a uni- variate alpha level of 0.0680. In addition, the response variable for the absence of an immediately preceding communication had a univariate alpha level of 0.0927. Examination of the univariate alpha levels for the test of race by socioeconomic group interaction indicated that no individual variable 92 TABLE 4-2 Analyses of Covariance for Race x Sex x SES Interaction: Communication Process Variable Between F Alpha Mean Square Ratio Level Multivariate Analysis (14 and 9 degrees of freedom) Communication process 0.7246 < 0.7160 Univariate Analysis (1 and 22 degrees of freedom) Initiation Failure to continue 4.9871 5.0278 < 0.0354 Unqualified power use 0.3941 0.5618 < 0.4615 Qualified power use 0.3084 0.2281 < 0.6377 Bribe 0.0091 0.0255 < 0.8748 Choice 0.0893 0.3044 < 0.5867 Response Rejection 0.3050 0.4320 < 0.5179 No response allowed 0.0981 0.4985 < 0.4876 No preceding message 0.3282 0.3479 < 0.5614 Ignoral 0.0068 0.0077 < 0.9311 Ongoing behavior 0.2061 0.2243 < 0.6405 Response object 0.1073 0.2631 < 0.6131 Reinforcement 0.7369 0.4358 <:o.Sl6l Feedback 0.0319 0.0064 <:0.9372 Reward 0.0092 1.2845 <:0.2693 Note.--Covariates were length of session and age of child. 93 TABLE 4-3 Analyses of Covariance for Sex x SES Interaction: Communication Process Variable Between F Alpha Mean Square Ratio Level Multivariate Analysis (14 and 9 degrees of freedom) Communication process 0.6681 < 0.7595 Univariate Analysis (1 and 22 degrees of freedom) Initiation Failure to continue 0.0481 0.0485 < 0.8278 Unqualified power use 0.0599 0.0854 < 0.7729 Qualified power use 0.8667 0.6410 < 0.4320 Bribe 0.3395 0.9501 < 0.3403 Choice 0.0431 0.1469 < 0.7052 Response Rejection 2.6030 3.6866 < 0.0680 No response allowed 0.5595 2.8439 < 0.1059 No preceding message 2.9153 3.0904 < 0.0927 Ignoral 1.6134 1.8263 < 0.1904 Ongoing behavior 1.7085 1.8592 < 0.1866 Response object 0.0414 0.1016 < 0.7530 Reinforcement 0.0192 0.0114 < 0.9162 Feedback 9.7262 1.9364 < 0.1780 Reward 0.0151 2.1088 < 0.1606 Note.--Covariates were length of session and age of child. 94 had values even approaching significance (Table 4-4). This evidence further supported the conclusion of no significance for the overall test. The third two—way interaction, that for the race by sex of child interaction, included two variables reflecting trends for consideration in terms of potential differences (Table 4-5). Relative number of on— going interaction sequences had a univariate alpha level of 0.0449, while the variable of relative number of qualified power assertions had a univariate alpha level of 0.0904. Test of Main Effects Since none of the interactions were judged to be significant, con- sideration was given next to the main effects of the three design fac- tors. Again, there were no significant differences for the overall tests (Table 4-1), so none of the main effects null hypotheses were rejected. However, marginal significance was indicated for the socio- economic group test, which had a multivariate alpha level of 0.0923. Multivariate alpha levels for the other two hypotheses were 0.5574 for sex of child and 0.9459 for race. These conclusions were further supported by inspection of the uni- variate alpha levels. Only one variable was within the range which might be considered significant; this was in the test for socioeconomic group differences. In addition, several marginal tendencies for indi- vidual variables were noted on the tests for main effects. Consistent with the marginal significance on the multivariate test for socioeconomic group differences, several potential differences were noted (Table 4-6). The greatest contribution to the overall 95 TABLE 4-4 Analyses of Covariance for Race x SES Interaction: Communication Process Variable Between F. Alpha Mean Square Ratio Level Multivariate Analysis (14 and 9 degrees of freedom) Communication process 0.9242 < 0.5684 Univariate Analysis (1 and 22 degrees of freedom) Initiation Failure to continue 0.5742 0.5789 < 0.4549 Unqualified power use 0.0648 0.0924 < 0.7641 Qualified power use 0.2335 0.1727 < 0.6818 Bribe 0.0064 0.0178 < 0.8952 Choice 0.1838 0.6264 < 0.4372 Response Rejection 1.2852 1.8203 < 0.1911 No response allowed 0.0350 0.1777 < 0.6775 No preceding message 0.9756 1-0341 < 0.3203 Ignoral 0.0575 0.0651 < 0.8010 Ongoing behavior 1.2269 1-3351 < 0.2603 Response object 0.3083 0.7558 < 0.3941 Reinforcement 0.0026 0-0016 < 0.9689 Feedback 0.0188 0.0037 < 0.9518 Reward 0 0001 0.0078 < 0.9306 Note.-—Covariates were length of session and age of child. 96 TABLE 4-5 Analyses of Covariance for Race x Sex Interaction: Communication Process Variable Between F Alpha Mean Square Ratio Level Multivariate Analysis (14 and 9 degrees of freedom) Communication process 0.6377 < 0.7828 Univariate Analysis (1 and 22 degrees of freedom) Initiation Failure to continue 0.6087 0.6136 < 0.4418 Unqualified power use 0.1178 0.1679 < 0.6860 Qualified power use 4.2420 3.1374 < 0.0904 Bribe 0.1995 0.5584 < 0.4629 Choice 0.0222 0.0758 < 0.7856 Response Rejection 0.1317 0.1866 < 0.6700 No response allowed 0.0005 0.0027 < 0.9591 No preceding message 1.1974 1.2692 < 0.2721 Ignoral 0.0942 0.1067 < 0.7471 Ongoing behavior 4.1600 4.5270 < 0.0449 Response object 0.3538 0.8675 < 0.3618 Reinforcement 0.0425 0.0251 < 0.8755 Feedback 7.2485 1.4431 < 0.2425 Reward 0.0078 1.0834 < 0.3093 Note.--Covariates were length of session and age of child. 97 TABLE 4-6 Analyses of Covariance for SES Main Effect: Communication Process Variable Between F. Alpha Mean Square Ratio Level Multivariate Analysis (14 and 9 degrees of freedom) Communication process 2.4247 < 0.0923 Univariate Analysis (1 and 22 degrees of freedom) Initiation Failure to continue 1.0915 1.1004 < 0.3056 Unqualified power use 4.9847 7.1059 < 0.0142 Qualified power use 0.3423 0.2531 < 0.6199 Bribe 0.0410 0.1148 < 0.7380 Choice 0.6712 2.2880 < 0.1447 Response Rejection 2.7598 3.9087 < 0.0608 No response allowed 0.9571 4.8644 < 0.0382 No preceding message 2.1443 2.2730 < 0.1459 Ignoral 2.8616 3.2391 < 0.0857 Ongoing behavior 0.0339 0.0369 < 0.8495 Response object 1.7054 4.1813 < 0.0531 Reinforcement 4.4276 2.6182 < 0.1199 Feedback 7.2457 1.4425 < 0.2425 Reward 0.0001 0.0198 < 0.8893 Note.--Covariates were length of session and age of child. 98 variance was by the variable for unqualified power assertion; the univariate alpha level of 0.0142 for this variable was considered to be within the range of significance. In addition, three response cate— gory variables indicated marginal significance on the univariate tests. The first, no response allowed, had a univariate alpha level of 0.0382; relative number of rejection responses had a univariate alpha level of 0.0608; and relative number of ignoral responses had a univariate alpha level of 0.0857. Finally, the response object variable had a marginally significant univariate alpha level of 0.0531. Because of the marginal significance of the socioeconomic group differences, cell means were examined for those variables reflecting potential differences (Appendix II-A). As shown graphically in Figure 4-1, these values indicated that there was greater use of unqualified power assertion by mothers than by fathers in the lower socioeconomic group, while the reverse was true for middle socioeconomic group par- ents (means of 0.0323 for lower socioeconomic group mothers, 0.0114 for lower socioeconomic group fathers, 0.0080 for middle socioeconomic group mothers, and 0.0202 for middle socioeconomic group fathers). As shown graphically in Figure 4-2, both mothers and fathers in the lower socioeconomic group gave more responses which were rejections than did their middle socioeconomic group counterparts. For both socioeconomic groups, however, mothers used more rejections than did fathers (means of 0.3658 for lower socioeconomic group mothers, 0.2919 for lower socioeconomic group fathers, 0.2455 for middle socioeconomic group mothers, 0.2239 for middle socioeconomic group fathers). No instances occurred for which fathers of either socioeconomic group were not allowed to respond. However, as shown graphically in 99 c >Io .Q-S 3... U n m m N m will) a m u n g p 2 1 Fathers n o m m .c o o m m-a E m -r4-H “'4 1 1 m g h 0 v Mot ers c 44:: c 0'44 3 0 0 3 4+ m m , a 3 Lower Middle Socioeconomic Group FIGURE 4-1. Comparison between mothers and fathers in relation to socioeconomic group for use of unqualified power assertion. o m m m "2 -H n o 40'~ (”Q n m o o m S n m c 30.. L o 0:3 Mothers p u m -H-m . Fathers 0 m 20 . n w: oIn u 0 0 4b a m 1 m m o 3 8» 04.0 0 ii 4' Lower Middle Socioeconomic Group FIGURE 4-2. Comparison between mothers and fathers in relation to socioeconomic group for use of rejection responses. 100 Figure 4-3, mothers in the middle socioeconomic group had more occasions during which they were not allowed to respond than did lower socioeco- nomic group mothers (means of 0.0003 for lower socioeconomic group mothers, 0.0008 for middle socioeconomic group mothers, and 0.0000 for both lower and middle socioeconomic group fathers). As shown graphically in Figure 4-4, lower socioeconomic group mothers ignored initiations made to them more frequently than did fathers in the same group. The reverse pattern, however, was indicated for parents in the middle socioeconomic group (means of 0.0151 for lower socioeconomic group mothers, 0.0121 for lower socioeconomic group fathers, 0.0080 for middle socioeconomic group mothers, and 0.0144 for middle socioeconomic group fathers). Finally, as shown graphically in Figure 4-5, fathers in the lower socioeconomic group responded relatively more to the behavior or per- formance of the children rather than to their worth as individuals than did mothers in this group. In contrast, the reverse pattern was indi- cated for middle socioeconomic group parents (means of 0.9819 for lower socioeconomic group mothers, 0.9912 for lower socioeconomic group fathers, 0.9953 for middle socioeconomic group mothers, and 0.9856 for middle socioeconomic group fathers). When absolute rather than relative amounts of this behavior are compared, however, a somewhat different pattern emerges. As shown graphically in Figure 4—6, mothers in both socio- economic groups respond relatively more to their children's behavior than do fathers in their respective groups (means of 0.9092 for lower socioeconomic group mothers, 0.8796 for lower socioeconomic group fathers, 0.8291 for middle socioeconomic group mothers, and 0.7734 for middle socioeconomic group fathers). 101 o .10 r g ,F, .09 .- m 3 0'2 '08 4 n c & Mothers m m m '07 ' 632 W. q, .06 I» 23 33 3 "4 S .05 «I 4 u 3' 04 C£.d . do '4 35 03 mIg u ' ‘4 O 38 02 u n o ' ‘4 6:3 8' 01 g ' 4 Fathers g .00 445 v? Lower Middle Socioeconomic Group FIGURE 4-3. Comparison between mothers and fathers in relation to socioeconomic group for initiations to which there is no opportunity to respond. U) c o -r-( ‘6 -H 2.0 f 4.) -H c -H g 1'5 4 Fathers U) E n 1.0 ‘ 8 Mothers m -H m 0.5 1 o 4.) 5 a Lower Middle Socioeconomic Group FIGURE 4-4. Comparison between mothers and fathers in relation to socioeconomic group for ignoral responses. 100.0 «I n o -.-I > O g'g 99.5 q? Mothers m m I: .Q 0 QIm 8:0 9900 AP n.a -.-I ‘H .C o u p u 98.5 1' Fathers c m m U 'U .. 3 £3 0 98.0 r 6 Jr n -.-I T '0 1 0.0 v 71L Lower Middle Socioeconomic Group FIGURE 4-5. Comparison between mothers and fathers in relation to socioeconomic group for relative distribution of response object categories. 95 T 90-» 85 t Mothers child's behavior ' Fathers 75 r L l O r Percent of time during which responses are directed at 1- Lower Middle Socioeconomic Group FIGURE 4-6. Comparison between mothers and fathers in relation to socioeconomic group for absolute amount of response object categories. 103 The sex of child main effect was not significant on the overall test, but there were several individual variables which possibly re- flected differences between parents (Table 4-7). Relative number of unqualified power assertions was again the only initiation variable indicating marginal significance, as shown by a univariate alpha level of 0.0878. One response variable, relative number of ongoing interac- tion sequences, had a marginally significant univariate alpha level of 0.0413. In addition, the response object variable had a univariate alpha level of 0.0805. The third main effect tested, that of race, did not include any univariate alpha levels which were even approaching significance (Table 4-8). This finding further supported the conclusion of no significance for the overall test. Test of Grand Mean Failure to find significance for any of the main effects or inter- actions justified interpretation of the hypothesis of primary interest, the test of the grand mean. This test, which essentially compared the scores for mothers with those for fathers, produced a multivariate alpha level of 0.2062 (Table 4-1). This value indicated that the null hypothesis for differences between parents could not be rejected. Even though the overall did not indicate significant differences between mothers and fathers, however, examination of the univariate alpha levels suggested one individual variable which was possibly indicative of such differences (Table 4-9). This was the initiation variable for failure to continue interaction, which had a univariate alpha level of 0.0710. This variable did not appear to be confounded by the marginal level of Analyses of Covariance for Sex Main Effect: Communication Process 104 TABLE 4-7 Variable Between F. Alpha Mean Square Ratio Level Multivariate Analysis (14 and 9 degrees of freedom) Communication process 0.9403 < 0.5574 Univariate Analysis (1 and 22 degrees of freedom) Initiation Failure to continue 1.9947 2.0110 < 0.1702 Unqualified power use 2.2398 3.1929 < 0.0878 Qualified power use 1.9585 1.4485 < 0.2416 Bribe 0.0199 0.0556 < 0.8158 Choice 0.4573 1.5588 < 0.2250 Response Rejection 0.4505 0.6380 < 0.4330 No response allowed 0.4378 2.2250 < 0.1500 No preceding message 0.4739 0.5023 < 0.4860 Ignoral 0.8604 0.9739 < 0.3345 Ongoing behavior 4.3213 4.7026 < 0.0413 Response object 1.3694 3.3574 < 0.0805 Reinforcement 2.3033 1.3620 < 0.2557 Feedback 7.0807 1.4097 < 0.2478 Reward 0.0092 1.2866 < 0.2689 Note.--Covariates were length of session and age of child. 105 TABLE 4-8 Analyses of Covariance for Race Main Effect: Communication Process . Between F Alpha Variable Mean Square Ratio Level Multivariate Analysis (14 and 9 degrees of freedom) \ Communication process 0.3871 < 0.9459 ._‘___p Univariate Analysis (1 and 22 degrees of freedom) \ Initiation Failure to continue 0.0684 0.0690 < 0.7953 Unqualified power use 0.0618 0.0882 < 0.7694 Qualified power use 0.6117 0.4524 < 0.5082 Bribe 0.0360 0.1007 < 0.7540 Choice 0.0227 0.0773 < 0.7837 Response Rejection 0.8048 1.1399 < 0.2973 No response allowed 0.1391 0.7069 < 0.4096 No preceding message 0.0196 0.0208 < 0.8867 Ignoral 0.7297 0.8260 < 0.3733 Ongoing behavior 0.6286 0.6840 < 0.4171 Response object 0.1916 0.4699 < 0.5002 Reinforcement 0.0533 0.0315 < 0.8607 Feedback 2.3473 0.4673 < 0.5014 Reward 0.0035 0.4850 < 0.4935 Note.--Covariates were length of session and age of child. 106 TABLE 4-9 Analyses of Covariance for Grand Mean: Communication Process Variable Between F. Alpha Mean Square Ratio Level Multivariate Analysis (14 and 9 degrees of freedom) Communication process 1.7286 < 0.2062 Univariate Analysis (1 and 22 degrees of freedom) Initiation Failure to continue 3.5713 3.6004 < 0.0710 Unqualified power use 0.2230 0.3180 < 0.5786 Qualified power use 0.2658 0.1966 < 0.6618 Bribe 0.3256 0.9113 < 0.3502 Choice 0.5939 2.0245 < 0.1689 Response Rejection 0.0410 0.0581 < 0.8118 No response allowed 0.4511 2.2927 < 0.1443 No preceding message 2.3905 2.5340 < 0.1257 Ignoral 0.4926 0.5576 < 0.4632 Ongoing behavior 1.0871 1.1830 < 0.2886 Response object 0.0655 0.1607 < 0.6925 Reinforcement 2.5136 1.4863 < 0.2357 Feedback 5.0397 1.0034 < 0.3274 Reward 0.0003 0.0432 < 0.8373 Note.--Covariates were length of session and age of child. 107 s;ixgnificance indicated on the test for the socioeconomic group main eeiffect, since this individual variable contributed relatively little t;<> the significance of that main effect test. Cell means were examined for the single variable which reflected potential differences between mothers and fathers (Appendix II-A) . The fféaii:hers had a higher rate of failure to continue interaction than did ITLCZ>1ZhEIS (means of 0.1850 for mothers and of 0.2750 for fathers). This t::z:weend was demonstrated for all but one subgroup (parents of lower socio- ee«c::<:>nomic group white males), as well as for the total sample. Contribution of Covariates Tests for the significance of the covariates indicated that these V'Ei-Jr‘j_ables made significant contributions to the analysis (Table 4-10), Only one of the as indicated by a multivariate alpha level of 0.0001. c:<>\’«a:z:'iates, length of session, was individually significant, however. TTPI‘EE lanivariate alpha level for that variable was also 0.0001, but the Lu'1‘31-‘ié'c‘ariate alpha level for age of the child was 0.1991. Verbal Communication (3 km: The multivariate analysis of covariance indicated that there were TIC) =5S-J'Lgnificant differences between mothers and fathers on the overall uneasesi Sai‘ulre of verbal communication as measured by the P-CIRP (Table 4-11) PIC) ‘-arqE3 . . . . . . . . . ”\Ier, inspection of the univariate F ratios indicated several indi— ‘er- "-4 4.) 0 Q) l 1r (H w d O % ‘r Lower Middle Socioeconomic Group FIGURE 4-7. Comparison between mothers and fathers in relation to socioeconomic group for affective tone. 3. H m 2 > P m A 5‘ Fathers w .. 1 >< é ' Mothers <2 0 a; a Lower Middle Socioeconomic Group FIGURE 4-8. Comparison between mothers and fathers in relation to socioeconomic group for anxiety level. 134 None of these values were even approaching the chosen level of sig- nificance for the hypothesis dealing with the race effect (Table 4-27). However, one variable, that of level of involvement, had a univariate alpha value of 0.0663 for the hypothesis relating to the sex of child factor (Table 4-28). Test of Grand Mean Although marginal significance was noted for the main effect of socioeconomic group, the test of the grand mean was examined for any clues to other potential differences. This test, which essentially compared the mother-father difference scores to zero, produced a multi- variate alpha level of 0.7769 (Table 4-21). Although this value indi- cated that the null hypothesis for differences between parents could not be rejected, this value was confounded by the significance of the previously noted test for socioeconomic group differences. The con- clusion of no significance for the test of the grand mean, however, was supported by inspection of the univariate alpha levels, none of which were even approaching significance (Table 4-29). Contribution of Covariates Tests for the significance of the covariates indicated that these variables did not significantly contribute to the analysis either indi- vidually or in combination (Table 4-30). The multivariate alpha level was 0.3977, while the univariate alpha levels were 0.5802 for length of session and 0.2493 for age of the child. As might be expected from these large alpha levels, no meaningful differences in the analyses were indicated when they were repeated without adjusting for these covariates. 135 TABLE 4-27 Analyses of Covariance for Race Main Effect: Nonverbal Communication . Between F Alpha Variable . Mean Square Ratio Level Multivariate Analysis (5 and 18 degrees of freedom) Nonverbal communication 0.7416 0.6024 Univariate Analysis (1 and 22 degrees of freedom) Affective tone 0.0224 0.3896 < 0.5390 Anxiety 0.0154 1.0359 < 0.3199 Level of involvement 0.0881 0.5234 < 0.4771 Physical behavior Quantity 0.9765 0.5498 < 0.4663 Affective direction 0.0108 0.6941 < 0.4138 Note.--Covariates were length of session and age of child. Analyses of Covariance for Sex Main Effect: 136 TABLE 4-28 Nonverbal Communication Va 'able Between F Alpha r1 Mean Square Ratio Level Multivariate Analysis (5 and 18 degrees of freedom) Nonverbal communication 1.2720 0.3186 Univariate Analysis (1 and 22 degrees of freedom) Affective tone 0.0550 0.9543 0.3393 Anxiety 0.0007 0.0483 < 0.8282 Level of involvement 0.6283 3.7341 < 0.0663 Physical behavior Quantity 1.5108 0.8507 < 0.3664 Affective direction 0.0191 1.2250 < 0.2804 Note.--Covariates we re length of session and age of child. 137 TABLE 4-29 Analyses of Covariance for Grand Mean: Nonverbal Communication Variable Between F. Alpha Mean Square Ratio Level Multivariate Analysis (5 and 18 degrees of freedom) Nonverbal communication 0.4935 0.7769 Univariate Analysis (1 and 22 degrees of freedom) Affective tone 0.0269 0.4664 < 0.5018 Anxiety 0.0003 0.0171 < 0.8973 Level of involvement 0.2989 1.7763 < 0.1963 Physical behavior Quantity 1.9695 1.1089 0.3038 Affective direction 0.0045 0.2882 < 0.5968 Note.--Covariates were length of session and age of child. 138 TABLE 4-30 Tests for Significance of Covariates in Multivariate Tests of Differences between Mothers and Fathers on Nonverbal Communication Dimensions of Parent-Child Interaction Test Statistics Covariates . Degrees of Alpha Chl Square Freedom Level Multivariate test 10.5016 10 < 0.3977 Univariate tests Age of child 6.6348 5 < 0.2493 Length of session 3.7890 5 < 0.5802 139 TeachingfiProcedure Overview The multivariate analysis of covariance indicated that there were significant differences between mothers and fathers in relation to the interaction between race and sex of the child on the overall measure of teaching procedure as measured by the P-CIRP (Table 4-31). No other sig- nificant differences were noted on the overall test, although inspection of the univariate F ratios indicated several individual variables which reflected areas of potential differences. The multivariate analysis of covariance for teaching procedure was computed using the three design variables of race, socioeconomic group, and sex of child. Covariates included length of the teaching session and age of the child. Dimensions of teaching procedure used as criterion variables included concept specification and teaching method. The seven scores for each parent which were used as measures of these dimensions were all logit transformations of category variable ratios, as delineated in Table 3-8. The difference scores for each mother-father pair on these seven variables were the dependent variables used in the analysis. Test of Three-Way Interaction The highest order interaction to be tested was that for race by socioeconomic group by sex of child. A multivariate alpha level of 0.7322 led to the conclusion that the null hypothesis for this three- way interaction could not be rejected (Table 4-31). This conclusion of no significance was further supported by inspection of the univariate l . . Cell means for all demographic groups on all teaching procedure variables and covariates are given in Appendices II-D and II—E, re- spectively. 140 TABLE 4-31 Summary of Multivariate Tests of Differences between Mothers and Fathers on Teaching Procedure Dimensions of Parent-Child Interaction Multivariate Test Statistics Hypothesis . Degrees of Alpha F Ratio Freedom Level Grand Mean 1.8575 7 and 16 < 0.1445 Race 0.3572 7 and 16 < 0.9141 Sex 0.6713 7 and 16 < 0.6937 Socioeconomic Group 2.0520 7 and 16 < 0.1108 Race x Sex 4.3070 7 and 16 < 0.0075 Race x Socioeconomic Group 0.7163 7 and 16 < 0.6600 Sex x Socioeconomic Group 1.6600 7 and 16 < 0.1896 Race x Sex x Socioeconomic Group 0.6202 7 and 16 < 0.7322 141 alpha levels, none of which were even approaching significance (Table 4-32). Test of Two-Way Interactions Since the potential confounding by the three-way interaction was judged to be minimal, the three sets of two—way interactions were con- sidered next. There were no significant differences for two of these, but the third did reflect significant between-group differences. Multi- variate alpha levels of 0.1896 for the sex of child by socioeconomic group interaction and of 0.6600 for the race by socioeconomic group interaction indicated that the null hypotheses pertaining to these two interactions could not be rejected. However, the null hypotheses for the race by sex of child interaction was rejected, based on a multi- variate alpha level of 0.0075. The conclusions of no significance for two of the interactions were further supported by inspection of the univariate alpha levels for their respective hypotheses. For the sex of child by socioeconomic group interaction, one variable was marginally significant when con- sidered individually (Table 4-33). This variable, relative amount of general summarizing by the parent during the teaching activity, had a univariate alpha level of 0.0328. The test of the race by socioeconomic group interaction revealed no significant univariate alpha levels for any individual variables (Table 4-34). Also consistent with the conclusion of significance for the race by sex of child interaction were indications of significance or marginal significance for several individual variables for that hypothesis (Table 4-35). One variable, relative amount of general summarizing by the 142 TABLE 4-32 Analyses of Covariance for Race x Sex x SES Interaction: Teaching Procedure . Between F Alpha V ariable Mean Square Ratio Level Multivariate Analysis (7 and 16 degrees of freedom) Teaching procedure 0.6202 < 0.7322 Univariate Analysis (1 and 22 degrees of freedom) Concept specification New concept 0.0015 0.0020 < 0.9650 New approach 2.7746 1.4114 < 0-2475 General orientation 1.2820 1.8352 < 0-1893 General summary 0.0285 0.0450 < 0.8340 Ongoing concept 0.1181 0.0957 < 0.7600 Teaching method Demonstration 0.0041 0.0134 < 0.9091 Illustration 0.0006 0.0007 < 0.9786 Note.--Covariates were length of session and age of child. 143 TABLE 4-33 Analyses of Covariance for Sex x SES Interaction: Teaching Procedure Variable Between F, Alpha Mean Square Ratio Level Multivariate Analysis (7 and 16 degrees of freedom) Teaching procedure 1.6600 < 0.1896 Univariate Analysis (1 and 22 degrees of freedom) Concept specification New concept 0.3968 0.5377 < 0.4712 New approach 0.0257 0.0131 < 0.9100 General orientation 0.2912 0.4168 < 0.5253 General summary 3.2881 5.1939 < 0.0328 Ongoing concept 0.1214 0.0984 < 0.7568 Teaching method Demonstration 0.1031 0.3357 < 0.5682 Illustration 0.0001 0.0001 < 0.9984 Note.--Covariates were length of session and age of child. 144 TABLE 4-34 Analyses of Covariance for Race x SES Interaction: Teaching Procedure Variable Between F. Alpha Mean Square Ratio Level Multivariate Analysis (7 and 16 degrees of freedom) Teaching procedure 0.7163 < 0.6600 Univariate Analysis (1 and 22 degrees of freedom) Concept specification New concept 0.0186 0.0251 < 0.8755 New approach 2.9354 1.4932 < 0.2347 General orientation 0.5317 0.7611 < 0.3925 General summary 0.1001 0.1582 < 0.6947 Ongoing concept 0.1233 0.1000 < 0.7549 Teaching method Demonstration 0.0010 0.0034 < 0.9543 Illustration 1.0751 1.2409 < 0.2774 Note.--Covariates were length of session and age of child. 145 TABLE 4-35 Analyses of Covariance for Race x Sex Interaction: Teaching Procedure Va 'able Between F Alpha r1 Mean Square Ratio Level Multivariate Analysis (7 and 16 degrees of freedom) Teaching procedure 4.3070 < 0.0075 Univariate Analysis (1 and 22 degrees of freedom) Concept specification New concept 3.6210 4.9068 < 0.0375 New approach 0.9806 0.4988 < 0.4875 General orientation 0.1411 0.2020 < 0.6575 General summary 4.5959 7.2597 < 0.0133 Ongoing concept 1.5690 1.2719 < 0.2716 Teaching method Demonstration 0.0034 0.0110 < 0.9174 Illustration 2.9607 3.4173 < 0.0781 Note.--Covariates were length of session and age of child. 146 parent during the teaching activity, had a univariate alpha level of 0.0133, which could be considered within the chosen level of signifi— cance. The variable reflecting relative amount of time spent in intro— duction of new concepts had a univariate alpha level of 0.0375, while a third variable, relative amount of illustration used by the parent in comparison with other teaching methods, had a univariate alpha level of 0.0781. Cell means were examined for the three significant and marginally significant variables which contributed to the greatest extent to the overall variance for the significant race by sex of child interaction (Appendix II-D). As shown graphically in Figure 4-9, these values indi- cated that mothers of black males and of white females tended to spend relatively more time in general summarizing during their teaching ac- tivities than did mothers of white males and black females (means of 0.0282 for mothers of black females, 0.0479 for mothers of black males, 0.0540 for mothers of white females, and 0.0385 for mothers of white males). This pattern was reversed for fathers, however, who tended to engage in relatively more general summarizing activity with black females and white males than with black males and white females (mean of 0.0454 for fathers of black females, 0.0541 for fathers of white males, 0.0392 for fathers of black males, and 0.0368 for fathers of white females). Mothers also tended to do more summarizing than fathers with black males and white females, while the reverse was true for the other groups. As shown in Figure 4-10, black fathers devoted more of their total teaching activity to introduction of new concepts than did white fathers when the child was a female, but the reverse was true with male chil— dren (means of 0.0866 for fathers of black males, 0.1019 for fathers of Mothers m >,c “'3 .H . >.,.4 White Hi: 5? g m «E l 6.8 4 1 C o mic-4 6‘8. rag 3 b 3 w Black m “-4 2 1h 0 0 .LJ 126 m m l > o u u o 838 A c O ‘ 4r Male Female Sex of Child Fathers q x c u.a -H N >144 "-4 u m 5 82 m a Black c 4 “ an o H White «30) l w c 3 u o c 1450 r u 21 4.) CT) 83 . HO 1 QJ> m 0 Male Female Sex of Child FIGURE 4-9. Comparison between mothers and fathers in relation to race and sex of child in relative amount of teaching activity devoted to general summarization. 148 Mothers White L T H Ot—‘qubUlO‘kaDO Black Percent of teaching activity devoted to introduction of new concepts 4_ 4 f T Male Female Sex of Child Fathers ‘—“—‘_' Black 9 8 1 White O\\l Percent of teaching activity devoted to introduction of new concepts OHNwt—m Male Female Sex of Child FIGURE 4-10. Comparison between mothers and fathers in relation to race and sex of child in relative amount of teaching activity devoted to the introduction of new concepts. 149 black females, 0.0978 for fathers of white males, and 0.0790 for fathers of white females). With mothers, however, the trend was reversed (means of 0.0934 for mothers of black males, 0.1065 for mothers of black fe- males, 0.0756 for mothers of white males, and 0.1026 for mothers of white females). As indicated graphically in Figure 4—11, white mothers used illus- tration as a teaching method relatively more for female than for male children, while the reverse was true for black mothers (means of 0.0772 for mothers of black males, 0.0673 for mothers of black females, 0.0499 for mothers of white males, and 0.1260 for mothers of white females). With fathers, however, the trend was reversed (means of 0.0832 for fathers of black males, 0.0768 for fathers of black females, 0.0988 for fathers of white males, and 0.0575 for fathers of white females). This contrast was particularly pronounced for mothers. Test of Main Effects The tests of main effects for race and sex of child must be con— sidered only in relation to the potential confounding by the signifi- cance of the race by sex of child interaction. Failure to find sig- nificance in any interaction involving socioeconomic group, however, coupled with the orthogonality of the design, made it possible to con— sider the test of the main effect for that factor independently of the significant two-way interaction. No significant difference was indi- cated for the overall test (Table 4-31), as indicated by a multivariate alpha level of 0.1108. Therefore the null hypothesis for socioeco— nomic group differences could not be rejected. The conclusion of no significant differences for socioeconomic groups was further supported by inspection of the univariate alpha Mothers ll ' -—————— >‘ r 4) U) 'U 10 -r~I-H O > .C 9 r «4 £2 +J J-J O Q) U'H E 8 > ('6 U (U 0‘ as: 7 + ”.4 a) J: BlaCk .C :3 U 6 > U H (U (U H Q) Q) "'4 u 5 ' 4J .C'. (U '44 U 4 L O-H U) .C “3 "J 3 3 P C '0 G) H G) 0 O U) 2 4» H '-H .‘3 (D Di 1 b O r ' Male Female Sex of Child H N Fathers H H OH 4 . Black White Percent of teaching activity for which illustration is used as a teaching method OHNWQWONNCDQ Male Female Sex of Child FIGURE 4-11, Comparison between mothers and fathers in relation to race and sex of child in relative amount of teaching activity devoted to use of illustration as a teaching method. 151 levels for that hypothesis. Only one variable, relative amount of teaching activity using demonstration as a method, was marginally sig— nificant when considered individually (Table 4-36). A univariate alpha level of 0.0667 was obtained for this variable. The two main effects which were confounded by the two-way inter- action both had nonsignificant overall tests. The multivariate alpha level for sex of child was 0.6937 and that for race was 0.9141. In— spection of the univariate alpha levels indicated that no individual variable was significant for either sex of child (Table 4-37) or race (Table 4-38). Although these results would indicate that the null hypotheses relating to these two factors could not be rejected, the previously discussed two-way interaction made it impossible to reach any conclusions concerning either variable in isolation. Examination of the previously reported results indicated that differences did in fact occur between the racial groups and also between parents of male and female children; these differences occurred, however, in relation to each other rather than in isolation, thus cancelling each other out in the tests for main effects. Test of Grand Mean The significance of the race by sex of child interaction had a potentially confounding effect on the test of the grand mean. However, consideration was given to this test with a caution directed toward interpretation as a means of identifying additional areas of potential difference between mothers and fathers. The overall test produced a multivariate alpha level of 0.1445 (Table 4-31). Taken in isolation, this value would not lead to rejection of the null hypothesis but to 152 TABLE 4-36 Analyses of Covariance for SES Main Effect: Teaching Procedure Variable Between F Alpha Mean Square Ratio Level Multivariate Analysis (7 and 16 degrees of freedom) Teaching procedure 2.0520 < 0.1108 Univariate Analysis (1 and 22 degrees of freedom) Concept specification New concept 1.6700 2.2631 < 0.1468 New approach 1.8349 0.9334 < 0.3445 General orientation 0.2183 0.3125 < 0.5818 General summary 0.0992 0.1566 < 0.6961 Ongoing concept 0.0001 0.0001 < 0.9943 Teaching method Demonstration 1.1441 3.7240 < 0.0667 Illustration 1.1985 1.3833 < 0.2522 Note.--Covariates were length of session and age of child. 153 TABLE 4-37 Analyses of Covariance for Sex Main Effect: Teaching Procedure Variable Between F. Alpha Mean Square Ratio Level Multivariate Analysis (7 and 16 degrees of freedom) Teaching procedure 0.6713 < 0.6937 Univariate Analysis (1 and 22 degrees of freedom) Concept specification New concept 0.2092 0.2834 < 0.5998 New approach 0.0139 0.0071 < 0.9339 General orientation 0.2320 0.3322 < 0.5703 General summary 0.0001 0.0002 < 0.9889 Ongoing concept 0.0001 0.0001 < 0.9987 Teaching method Demonstration 0.0211 0.0688 < 0.7956 Illustration 2.3160 2.6732 < 0.1163 Note.--Covariates were length of session and age of child. 154 TABLE 4-38 Analyses of Covariance for Race Main Effect: Teaching Procedure Variable BEtween F Alpha Mean Square Ratio Level Multivariate Analysis (7 and 16 degrees of freedom) Teaching procedure 0.3572 < 0.9141 Univariate Analysis (1 and 22 degrees of freedom) Concept specification New concept 0.1089 0.1476 < 0.7045 New approach 0.1643 0.0836 < 0.7753 General orientation 0.3011 0.4310 < 0.5183 General summary 0.1007 0.1591 < 0.6939 Ongoing concept 0.1949 0.1580 < 0.6949 Teaching method Demonstration 0.0783 0.2549 < 0.6187 Illustration 0.0144 0.0167 < 0.8985 Note.--Covariates were length of session and age of child. 155 conclusion of no differences between mothers and fathers. However, differences were noted in relation to the design variables involved in the interaction, as previously discussed, and univariate alpha levels indicated several other areas of potential differences (Table 4-39). Examination of tests of the grand mean for the individual vari- ables indicated trends which might be considered for six of the seven variables. The only one not included was relative amount of teaching activity which included illustration as a teaching method. This vari- able, however, was one which was involved in the race by sex of child interaction and for which differences were in fact noted (Figure 4-11). The individual variable making the greatest contribution to the overall variance was relative amount of teaching activity which em— ployed demonstration as a teaching method. A univariate alpha level of 0.0026 was obtained for this variable. The five concept specification variables had individual levels which reflected potential differences between mothers and fathers. Rela- tive amount of teaching activity devoted to general orientation had a univariate alpha level of 0.0196. Presentation of new concepts and on- going presentation of the same concept had univariate alpha levels of 0.0345 and 0.0314, respectively, while the univariate alpha level for use of a new approach was 0.0542. Finally, the general summary variable had a univariate alpha level of 0.0947; it should be noted also that this was the greatest contributor to the race by sex of child interaction. Cell means were examined for all those variables included in this analysis (Appendix II-4). As previously reported, fathers tended to devote a relatively greater amount of teaching activity to general sum- marization than did mothers (means of 0.0421 for mothers and 0.0439 156 TABLE 4-39 Analyses of Covariance for Grand Mean: Teaching Procedure . Between F Alpha Variable Mean Square Ratio Level Multivariate Analysis (7 and 16 degrees of freedom) Teaching procedure 1.8575 < 0.1445 Univariate Analysis (1 and 22 degrees of freedom) Concept specification New concept 3.7519 5.0842 < 0.0345 New approach 8.1357 4.1385 < 0.0542 General orientation 4,4314 6.3435 < 0.0196 General summary 1.9312 3.0505 < 0.0947 Ongoing concept 6.5244 5.2890 < 0.0314 Teaching method Demonstration 3.5480 11.5489 < 0.0026 Illustration 0.8016 0.9252 < 0.3466 Note.--Covariates were length of session and age of child. 157 for fathers), although this difference was primarily in relation to the race by sex of child interaction. They also spent somewhat less time in the introduction of new concepts (means of 0.0913 for fathers and 0.0945 for mothers), although again, these differences were primarily in relation to the race by sex of child interaction. Although differences for the other three concept specification codes must be interpreted in relation to the significance of the multi- variate test for the race by sex of child interaction, no significant tests of a higher order were indicated for these specific variables. The variable for use of new approaches indicated that fathers devoted relatively more of their teaching activity to this procedure than did mothers (means of 0.1360 for mothers and 0.1427 for fathers). Fathers also devoted relatively more teaching activity than did mothers to general orientation (means of 0.0234 for mothers and 0.0265 for fathers) and to ongoing teaching of the same concept (means of 0.2967 for mothers and 0.3242 for fathers). As reported previously (Figure 4-11), there was little difference between black parents in the relative amount of teaching activity for which illustration was used as a method. For white parents, however, a difference was noted; mothers used relatively more illustrations with their daughters than with their sons, while the reverse was true for fathers. The contrast was particularly pronounced for mothers. In contrast, relative amount of teaching activity employing demon- stration as a teaching method varied in relation to socioeconomic group as well as sex of parent. As shown in Figure 4-12, use of this tech- nique was greater for mothers than for fathers in the lower socioeco- nomic group, while the reverse was true for middle socioeconomic group 158 3 4r 2 .. Mothers l o Fathers 0 G ; Lower Middle Socioeconomic Group FIGURE 4-12. Comparison between mothers and fathers in relation to socioeconomic group in relative amount of teaching activity devoted to use of demonstration as a teaching method. 159 parents (means of 0.0171 for lower socioeconomic group mothers, 0.0134 for lower socioeconomic group fathers, 0.0038 for middle socioeconomic group mothers, and 0.0197 for middle socioeconomic group fathers). Contribution of Covariates Tests for the significance of the covariates indicated that these variables made significant contributions to the analysis (Table 4-40). The multivariate alpha level was 0.0002, while the univariate alpha levels were 0.0002 for length of session and 0.0708 for age of the child. Summary Overall Comparisons The results indicated that there were both similarities and differ- ences between mothers and fathers in their interactions with their pre- school children in a dyadic teaching situation. Differences were con- centrated in the areas of nonverbal communication and teaching proce- dure, with marginal differences in communication process. Similarities occurred primarily in the area of verbal communication. Multivariate tests of the general hypotheses for this study led to the following conclusions: (8) There was a trend toward a difference in relation to socio- economic group between the communication process of mothers and that of fathers with their preschool children in the teaching of a two— dimensional sorting task as measured by the P-CIRP. (b) There was no difference between the verbal communication of mothers and that of fathers with their preschool children in the teach- ing of a two-dimensional sorting task as measured by the P-CIRP. 160 TABLE 4-40 Tests for Significance of Covariates in Multivariate Tests of Differences between Mothers and Fathers on Teaching Procedure Dimensions of Parent-Child Interaction Test Statistics Covariates . Degrees of Alpha Chl Square Freedom Level Multivariate test 41.1455 14 < 0.0002 Univariate tests Age of child 13.0549 7 < 0.0708 Length of session 28.4676 7 < 0.0002 161 (c) There was a difference in relation to socioeconomic group be- tween the nonverbal communication of mothers and that of fathers with their preschool children in the teaching of a two-dimensional sorting task as measured by the P-CIRP. (d) There was a difference in relation to an interaction between race and sex of child between the teaching procedure of mothers and that of fathers with their preschool children in the teaching of a two— dimensional sorting task as measured by the P-CIRP. Since the study was exploratory in nature, tests for individual variables were examined in order to identify areas of potential differ- ences. Trends indicated that although there were basic similarities between parents on some of the interaction dimensions, there were also some potential differences of a more specific nature. These results are summarized in Table 4-41. Comparisons on Individual Variables Communication process. Nine of the fourteen variables considered as dimensions of communication process reflected at least a marginal significance on one of the eight hypotheses. The covariate of length of session was significantly related to the communication process dimen- sion, but age of the child was not. The individual criterion variables and the significance attributed to each for the six dimensions of commu- nication process were as follows: (a) Initiation--Fathers tended to terminate interaction sequences more frequently than did mothers; a marginally significant race by sex of child by socioeconomic group interaction was also noted for this variable. There was a tendency for lower socioeconomic group mothers 162 TABLE 4-41 Summary of Significance and Marginal Significance for Individual Variables Univariate Alpha Variable < 0.01 0.01 - 0.05 0.06 — 0.10 Communication Process (Multivariate alpha for SES = 0.09) Initiation Failure to continue interaction - R x S x SES [P] Unqualified power assertion [*1 SES - S Qualified power assertion - - R x S Bribe - - — Choice — - - Response S x SES [*1 SES No response allowed — [t] SES _ Rejection - - Ignoral - - [*] SES Ongoing interaction - - No immediately pre- ceding message - - S x SES Response object - [*] SES S Reinforcement - - — Feedback - - - Reward - - - Verbal Communication Verbalization Quantity - P — 163 TABLE 4-41 (continued) Variable Univariate Alpha < 0.01 0.01 - 0.05 0.05 - 0.10 Type Integration Tension reduction Decision Control Evaluation Fantasy verbalization Voice tone Verbal specificity Time orientation Task orientation Nature of interference Initiator Remoteness from task R x S x SES SES R x S X SES Nonverbal Communication (Multivariate alpha for SES = 0.05) Affective tone Anxiety Level of involvement Physical behavior Quantity Affective direction * SES * SES R x S x SES (Multivariate alpha for R x S < 0.01) Teaching Procedure Concept specification New concept * R x S (P) 164 TABLE 4-41 (continued) Univariate Alpha Variable < 0.01 0.01 - 0.05 0.05 - 0.10 New approach - (P) - General orientation - (P) - * General summary - SRxXSES (P) Ongoing teaching - (P) - Teaching method Demonstration (P) - SES Illustration - - * R x S Note.--( ) indicates potential confounding by significance of other factors or interactions. [ ] indicates potential confounding by marginal significance of other factors or interactions. * indicates significance within significant multivariate test. P indicates difference between parents. R indicates race difference. S indicates sex difference. SES indicates socioeconomic group difference. R x S indicates race by sex interaction. S x SES indicates sex by socioeconomic group interaction. R x SES indicates race by socioeconomic group interaction. R x S x SES indicates race by sex by socioeconomic group interaction. 165 to use relatively more unqualified power assertion than did fathers, while the reverse was true for middle socioeconomic group parents; mothers also tended to use more unqualified power assertion with their sons than did fathers, while the reverse was true for parents with fe— male children. Qualified power assertion, however, was used relatively more by black mothers with their sons and by white fathers with their daughters than by other parents. No differences were noted between parents in the extent to which they used bribes or gave their children choices. (b) Response-—There was a tendency for lower socioeconomic group mothers to have higher proportions of responses which were rejections or ignorals or for which there was no opportunity to respond than did fathers in the same group, while the reverse tendency was indicated for middle socioeconomic group parents. Although both mothers and fathers used more rejections with their sons than with their daughters, this contrast was greatest for middle socioeconomic group mothers and least for lower socioeconomic group mothers, with fewer differences between fathers from the two socioeconomic groups. There was a tendency for fathers in general, but particularly those of middle socioeconomic group females, to have a rather high proportion (compared to that for mothers) of interaction sequences in which there was no immediately preceding communication. Finally, the occurrence of ongoing interac- tion sequences was somewhat greater for mothers with males and for fathers with females. These contrasts tended to be greater for the black families with males and for the white families with females. (c) Response object--There was a tendency for lower socioeconomic group mothers to respond relatively more to the child as a person 166 instead of to his behavior than was true for fathers in this group, while the reverse situation was indicated for middle socioeconomic group families. In addition, mothers tended to respond more to the individual worth of their sons than of their daughters, while the reverse was true of fathers. (d) Reinforcement-~No differences between mothers and fathers were indicated. (e) Feedback--There were no differences that were indicated between mothers and fathers. (f) Reward--There were no differences that were indicated between mothers and fathers. Verbal communication. Nine of the thirteen variables considered as dimensions of verbal communication reflected at least a marginal significance on one of the eight hypotheses. Neither age of the child nor length of the teaching session was significantly related to the verbal communication dimension. The individual criterion variables and the significance attributed to each for the seven dimensions of verbal communication were as follows: (a) Verbalization--Consistent with popular rumor, mothers tended to talk somewhat more than did fathers. Fathers, however, tended to use relatively more verbalization concerned with both tension management and with evaluation than did mothers. They also engaged in signifi- cantly more decision-oriented verbalization than did mothers, a trend which was especially true for those from lower socioeconomic black and middle socioeconomic white families. There was a tendency for mothers to engage in more verbalization concerned with integration (either 167 supportive or antagonistic) with male than with female children, while the reverse was true for fathers. No differences were indicated be- tween mothers and fathers in the relative amount of verbalization which dealt with control over others. (b) Fantasy verbalization--There was a tendency for black mothers, particularly those from the middle socioeconomic group, to engage in more fantasy verbalization than did other parents. (c) Voice tone--No significant differences between mothers and fathers were indicated for voice tone. (d) Verbal specificity--Mothers used significantly more speci- ficity in verbalization in contrast to the predominantly global verbal- ization of fathers. (e) Time orientation--No significant differences between mothers and fathers were indicated for verbal time orientation. (f) Task orientation--Lower socioeconomic group mothers tended to engage in relatively more non-task verbalization than did fathers, while the reverse was true for middle socioeconomic group parents. (9) Nature of interference-~No differences between parents were indicated for initiation of non-task verbalization. However, fathers' non-task verbalizations were focused on factors significantly more re— moved from the immediate situation than were the non-task verbalizations of mothers. Nonverbal communication. Three of the five variables considered as dimensions of nonverbal communication reflected at least a marginal significance on one of the eight hypotheses. Neigher age of the child nor length of the teaching session was significantly related to the 168 nonverbal communication dimension. The individual criterion variables and the significance attributed to each for the four dimensions of non- verbal communication were as follows: (a) Affective tone--Middle socioeconomic group parents tended to have a more positive affective tone than did lower socioeconomic group parents. Within the middle socioeconomic group, mothers were more positive than were fathers, while the reverse was true for lower socio- economic group parents. (b) Anxiety--Significant differences between parents in relation to socioeconomic group were indicated. Greater anxiety was shown by lower socioeconomic group mothers than by fathers, while the reverse was true for middle socioeconomic group parents; this contrast was par- ticularly pronounced for mothers in comparison with fathers. A mar- ginally significant race by sex of child by socioeconomic group inter- action was also indicated. (c) Level of involvement--Mothers tended to be more involved with their sons and fathers with their daughters. (d) Physical behavior--No differences between parents were indi- cated for either quantity of physical behavior or for the affective direction of that behavior. Teaching procedure. All seven of the variables considered as di- mensions of teaching procedures reflected at least a marginal signifi- cance for one of the eight hypotheses. In addition, length of the teaching session was highly related to the teaching procedures used by the parents, while the age of the child was somewhat related. The individual criterion variables and the significance attributed to each 169 for the two dimensions of teaching procedure used in this analysis were as follows: (a) Concept specification-~Significant differences between parents in relation to race and sex of the child were indicated for relative amount of teaching activity devoted to general summarizing. Differ- ences in relative amount of time devoted to introduction of new con- cepts were marginally significant and followed the same general pat- tern. Parents of black females and of white males tended to devote more time to both these activities than did other parents. The con- trasts were particularly great for fathers, whose scores were also higher overall for both variables. Mothers tended to spend more time with their daughters than with their sons in general summarizing in particular, while the reverse was true for fathers. Contrasts between low and middle socioeconomic group parents were greater for mothers than for fathers of females, while the reverse was true for parents of males. Amount of time spent in presenting different approaches was greater for black fathers and white mothers with their sons than with their daughters, while the reverse was true for white fathers and for black mothers. Relative amount of teaching activity devoted to general orientation was greatest for mothers in black families and for fathers in white families. Finally, fathers in general tended to spend more time in ongoing activity than did mothers. (b) Teaching method--There was little difference between black parents in the extent to which illustration was used as a teaching method. In white families, however, mothers used this technique more with their daughters than with their sons, while the reverse was true for fathers. The use of demonstration as a teaching method was greater 170 for mothers than for fathers in the lower socioeconomic group, but higher for fathers than for mothers in the middle socioeconomic group. CHAPTER 5 CONCLUSION A view of Mother by those who keep her under closest surveillance and probably know her best . . . (And) the observations of those who may see (Father) least but love him best . . . --(Their) children: "A mother is someone to LOVE. And someone to play with. You will enjoy a MOTHER." "(A father) is like a guidence counslor and guides you through the early stages of life. He leads us from right and wrong." "A mother spanks you, but only for a reason, and that reason your supposed to know." "What is a father -- He keeps saying when I was your age." "A mommy is a wife. A mommy looks after children and she yells." "A Father is a man who picks you up and holds you when your little and afraid." "Mothers say no, no, before the child can ever tell her what she wants but I think mothers should let them finnish." "What is a father -- For most children he is an idle." "Just what is a mother? A mother is undescribable." "Fathers are parents, but are male." ——From Lee Parr McGrath and Joan Scobey (Eds.). What is a Mother. New York: Simon & Schuster, 1968; and Lee Parr McGrath and Joan Scobey (Eds.). What is a Father. New York: Simon & Schuster, 1969. 171 172 Discussion Summary of Results Investigation of four major areas of interaction indicated that there were no overall differences between mothers and fathers in the area of verbal communication, but such differences did occur in the areas of nonverbal communication and teaching procedure; trends were also indicated for the communication process dimension. The differences in teaching procedure were a function of an interaction between race and sex of the child, while the differences in nonverbal communication and the trends in communication process were a function of socioeconomic group. In addition, a number of individual variables reflected poten— tial differences between parents. Overall Comparison A general comparison between mothers and fathers indicated that, in contrast to fathers, mothers tended to be relatively more verbal. In addition, the nature of this verbalization differed in several ways. Mothers used greater specificity than did fathers in their verbaliza- tion, and they used relatively more verbalization dealing with tension management. Fathers, on the other hand, were more decision-oriented and more evaluation-oriented. In addition, their non-task verbaliza- tion dealt with matters less closely related to the immediate situation. There was a tendency for mothers' teaching interaction sequences to be somewhat shorter than was the case with fathers; however, mothers tended to continue interaction by initiation of a new sequence more fre- quently than did fathers. The interactions of the mothers also had a relatively low frequency of cases in which there was no immediately 173 preceding communication, which indicated that the child less frequently terminated the interaction sequence than with fathers. Teaching pro- cedure used by parents differed both in kinds of concept specification used and also in teaching method. In general, mothers tended to intro- duce new concepts more often than did fathers, while fathers' teaching behavior was characterized by relatively more use of general orienta- tion, use of new approaches, and use of demonstration as a teaching method. The composite picture of the mothers' interactions indicated that they tended to focus to a relatively great extent on the specifics of the task presented them. Their behavioral involvement was high, par- ticularly with their daughters, and apparently this involvement carried over to the children as well. Relatively frequent changes of teaching strategies occurred. The nurturant role which has been used to charac- terize mothers' behavior (e.g., Emmerich, 1962) was indicated only to the extent that this may be reflected in the greater quantity of verbal- ization, particularly that dealing with tension management, and the more continuous interaction in which the mothers engaged. By comparison, fathers appeared to have a more process-oriented approach, but with less focused content. Their interaction sequences were of longer duration, but with greater lapses between. Whether this pattern reflected less rapport or communication skill or less bombard- ing of a child in a marathon setting was impossible to determine. Socioeconomic Group Comparison The greatest differences between parents were those which were indi- cated in relation to socioeconomic group. Lower socioeconomic group 174 mothers seemed to present a particularly striking-—and bleak--picture. Their interactions with their children reflected a relatively negative affective tone, a high level of anxiety, and relatively high rates of unqualified power assertions, rejection, ignorals, and failures to re- spond because of no opportunity to do so. In addition, these mothers responded more often to their children as individuals rather than to the behavior displayed--and in a critical manner more often than did any other group of parents. More verbalizations of the lower socio- economic group mothers were also focused on non-task factors. Lower socioeconomic group fathers also showed a relatively nega- tive affective tone (though not as much as mothers). However, they seemed to demonstrate less anxiety and relatively lower rates of un- qualified power assertions, rejections, ignorals, and failures to re- spond because of no Opportunity to do so than did mothers. They re- sponded relatively more to the children's behavior than to their indi- vidualities, and these responses were less often critical in nature than were the responses of the mothers. These fathers also gave less non-task verbalization than did the mothers. Both mothers and fathers in the middle socioeconomic group re- flected a more positive affective tone than did their lower socioeco— nomic group counterparts, with a particularly high rating on this dimension for mothers. On the other aspects, however, the behavior for the mothers who were in the middle socioeconomic group was similar to the behavior of the fathers who were in the lower socioeconomic group, while the middle socioeconomic group fathers showed behaviors which were more like those of the mothers who were in the lower socio- economic group. 175 This interesting pattern of behavior was somewhat inconsistent with the results of other studies, which typically indicated that middle socioeconomic group parents were both more supportive and_more con- trolling with their children than were lower socioeconomic group parents (e.g., Straus, 1967). Only the mothers in the present study fit this model, a finding which was particularly interesting in view of the ten- dency of previous studies to deal 921y_with mothers' behavior or to ob- tain reports of fathers' behavior by asking mothers. It was not surprising that the lower socioeconomic group parents generally demonstrated more negative affect than did middle socioeconomic group parents. This tendency has been supported by previous research (e.g., Hess, 1970) and is furthermore quite understandable in view of the unfortunate circumstances under which these families are often forced to live. By definition, they live in housing conditions which are not only less desirable and in poor condition but also more crowded; they work at jobs with lower status which are typically more physically exhausting for lower pay; they have limited financial means and therefore less control over their own circumstances. The fact that it was the mother in the lower socioeconomic group who reflected the most negatively-oriented behavior may be explained by the fact that she is the one who typically has to cope on a day-to-day, hour- to-hour basis with the circumstances of this existence--including care of the children over extended periods of time. It is little wonder that the behavioral pattern characterizing these mothers approximates White's (1969) descriptions of the "zookeeper" or "overwhelmed" mother. Lower socioeconomic group fathers, on the other hand, have some op- portunities to escape the pressures of the home through their employment 176 or other contacts. They may not have as much opportunity to establish higher thresholds for rejecting behavior as do the mothers. Middle socioeconomic group fathers, in contrast to mothers, play a more traditional authoritarian role. They probably have less oppor- tunity for interacting with their children at home--which may mean they are less at ease in the situation and also that control over the chil- dren is maintained less overtly and more by the aura of expectation. Middle socioeconomic group mothers represent the group that is probably most experienced in interacting with their preschool children in situations of the type used in this study; at the same time, they have the financial means of managing their family circumstances so they are not forced to become overwhelmed by their circumstances. Middle socioeconomic group mothers are also members of a group which places a high value on education, are themselves better educated, and probably are the most accustomed of all to interacting with their preschool chil- dren in teaching situations similar to the one used in this study. Sex of Child Comparison As with previous research, several differences were noted in rela- tion to the interaction between sex of parent and sex of child. Mothers tended to use more unqualified power assertion, more integration—oriented verbalization, and more ongoing interaction sequences with their sons than with their daughters, while the reverse was true for fathers. This pattern was the reverse of Emmerich's (1962) finding that parents tended to exert more power toward their same-sex children than toward their opposite-sex children. Somewhat more consistent with Emmerich's (1962) study was the fact that mothers also tended to be less critical and more 177 involved with their sons than with their daughters, however, while fathers were less critical and more involved with their daughters than with their sons. Fathers also tended to respond more to the behavior rather than to the individuality of their sons than they did with their daughters, while the reverse was true for mothers. A smattering of differences were also noted for parents in relation to demographic group interactions which included sex of the child. These were primarily in relation to race and were concentrated in the dimensions of teaching procedure; additional differences were noted in the area of communication process. In summary, mothers of black males and of white females used more illustration in their teaching and devoted more teaching time to general summary and less time to introduction of new concepts than did mothers of white males and black females; they also used relatively more quali- fied power assertion than did mothers of white males and black females. These patterns were reversed for fathers from the same groups. Both mothers and fathers used more rejections with their sons than with their daughters, but this contrast was greatest for middle socioeconomic group mothers and least for lower socioeconomic group mothers. Middle socio- economic group fathers also had a particularly high proportion of inter— action sequences in which there was no immediately preceding communica- tion. Finally, the occurrence of ongoing interaction sequences was somewhat greater for mothers with males and for fathers with females, contrasts which were greater for black males than for white females. It is interesting to note that while differences did in fact exist in relation to an interaction between sex of parent and sex of child, the direction was sometimes the reverse of what might be expected. 178 Examination of the data, however, indicated that the patterns typically described have been those which characterized white families, particu- larly those from the middle socioeconomic group. For all families, though, it is evident that differential socialization through parental interaction does in fact occur for boys and girls; this difference is mediated by differences in interaction with same-sex and opposite-sex parents. Race Comparison Few differences were found between parents based on race, but an interaction between race and sex of child was indicated for teaching procedure, and several individual variables were marginally significant. In particular, fathers of white females tended to use less general summarization, less introduction of new concepts, and less illustration in their teaching than did other parents. Black parents, particularly mothers, were especially high on these dimensions. Although neither theoretical bases nor previous research were available to support these differences, they were consistent with the subjective observation that these two groups of parents tended to be somewhat more permissive with their children and that middle socioeconomic group black mothers tended toward White's (1969) "super-mother" category. While rather different interaction patterns do appear to exist be- tween black and white families on some dimensions of interaction, these appear to be related primarily to sex. A greater equalitarian orienta- tion appeared to characterize black families, particularly those from the middle socioeconomic group. This observation was consistent with the response often given by black women to Women's Liberation advocates 179 that they have been liberated for a long time. It is also lOgical in consideration of the more recent upward mobility of the middle socio- economic group black families and the social emphasis on black conscious- ness which was at its peak during the period these data were collected. Covariates It was logical that no differences were found on the basis of either age of the child or length of the teaching session for the dimen- sions of either verbal communication or nonverbal communication. Both of these areas of interaction reflect more basic methods of relating in the interaction situation and are probably less situation-specific than other dimensions. Age of the child was marginally significant for teaching procedure and nonsignificant for communication process. It was logical that the procedure used in teaching a child would be related to his age. This would particularly be the case in teaching a task of the type used in this study--i.e., one on which ability to learn was highly related to developmental level. Finally, length of the teaching session was highly related to both teaching procedure and to communica- tion process. Again, this was logical, in that length of the session could in some respects be considered a dimension of the communication process; the patterns of teaching probably also reflected in part the child's performance in the teaching-learning situation, as did length of the session. The teaching sessions were longer for fathers than for mothers for all but two groups. This was particularly interesting since the session with the father was in each case held after the session with the mother. Therefore, if the child did in fact learn anything in the session with 180 the mother, less time would be expected to be necessary for the session with the father. The fact that the father-child session took a longer time may have reflected less skill in teaching by the fathers or less security by them in the teaching situation. Limitations It was undoubtedly a researcher who noted that "to err is human, but to really foul up things requires a computer." Having involved both humans an§_computers, this study was not without its limitations. Among those which should be considered in interpreting the results are the following: (a) Although possessing certain critical advantages, the laboratory setting is an unavoidably artificial situation for assessment of parent- child interaction. (b) Only one aspect (i.e., the parent's behavior) of a reciprocal process was considered in these analyses. (c) Variables other than or in addition to those used in this study may be critical for consideration in the assessment of parent-child interaction. (d) The sample size was small, and the subjects were not randomly selected from the general population. (e) The parent-child interaction measures for the mother and the father, though conceptually comparable, have not been empirically shown to be psychometrically equivalent. (f) The task situation may not have been equally appropriate for all subjects due to individual differences in the developmental levels of the children, experience of the parents and/or children with the 181 type of activity called for, and/or previous familiarity with the task materials themselves. (9) The order of testing was not systematically varied between mothers and fathers, so there is the possibility that the child's ex- perience in the mother-child interaction situation and/or the mother's possible discussion of the task with the father may have altered the nature of the subsequent father-child interaction situation. (h) The time lapse between the mother-child and father-child measures varied within a range that was greater than desirable, thus producing possibly confounding effects that were not adequately adjusted for by the covariance procedures. (i) Previous validation of the P-CIRP as a measure of parent—child interaction was minimal. (j) The rating procedure used for this study is a very complex instrument, and although every effort was made to maintain reliability, errors of measurement are inevitable and undoubtedly occurred in the collection of these data. (k) Statistical models used in this study had not been previously applied to data in the manner used here and may have limitations which are not yet evident. (1) The four areas of interaction which were tested separately were potentially inter-related, so a multivariate test across these four areas would probably have been more appropriate if there had been enough degrees of freedom to permit such an analysis. (m) The relatively large number of dependent variables included within the study was a limiting factor in tests of statistical infer- ence. The possibility of random significance was thus quite high. 182 Implications Since this study was by design exploratory in nature, it raises more questions than it answers. Its primary importance, therefore, lies in the basis which is provided for further investigations. A sidelight emphasized by these data is that most of the current litera- ture and research has been focused on a highly selective group--mothers, white families, and the middle socioeconomic group. Since different patterns appear to exist for these different groups, it is important to turn attention to groups more representative of the total population to avoid misrepresentation of parental interaction, as well as other behaviors and attitudes. Even the preliminary results of this study suggest several important considerations for family and child scientists and educators. Since there do appear to be some differences between the ways in which mothers and fathers interact with their preschool children, there are obvious implications for the child's socialization within the family context. The effects of the absence of a parent from the home-~either permanently or temporarily--should certainly be considered in relation to the unique socializing influences of the mother and the father. Information concerning the way in which children are socialized, particularly the differential patterns among various groups, is important to educators and social scientists in planning programs for individuals and groups. Enrichment of the environment of the lower socioeconomic group appears to be critical, particularly since the affective mileau in which children live plays a key role in the development of their own atti— tudes and behavior. This cycle of despair is perpetuated through parent— child interaction; attention to its remediation would appear to be of a 183 top priority for society. Parent education programs in particular might well be planned on the basis of the information from this study, including such factors as teaching techniques and communication process dimensions, as well as information about general developmental considerations. The pre- liminary results of this study suggest that two groups toward which programs should be focused are lower socioeconomic parents and fathers. This is particularly relevant in view of the fact that most parent edu— cation programs currently tend to be directed toward middle socioeco- nomic group mothers. An interesting kind of pattern for parental behavior reflected by these data was the tendency for fathers to employ more decision-oriented and evaluation-oriented verbalization and also to engage in more global verbalization, a profile which is consistent with current societal pat- terns. Men are often thought to deal more with abstractions and with process-oriented activities, while women's activities are typically thought to be of a more mechanical nature. Since society is built around the idea of choice, it would seem to be particularly important for the socialization of all children to be oriented toward dealing with process and decision-making. Since the family is the key social- izing agent, and since the mother typically spends the greatest amount of time with young children, these differences in parental interaction patterns bear important implications for the development of children's behavioral patterns. Sex-role identification of girls with their mothers suggests greater chances of perpetuating the societal patterns reflecting these somewhat discrepant orientations. The potential dif— ferences between societal values and expectations, apparent sex-role 184 differences, and differential socialization of children need to be given careful consideration. Only by bringing these to the level of awareness can their resolution be accomplished. As John Kennedy pointed out, "Children are the world's most valuable resource and its best hope for the future"--and an important tool in the mining of this resource is obviously through the parents of those children. Suggestions for Further Investigation Based on the findings of this exploratory study, a number of topics for further investigation are suggested. In particular, the following are recommended: (a) The present study should be replicated with a larger and statistically representative sample. (b) A more detailed investigation of those variables identified by this study as areas of potential differences between mothers and fathers should be undertaken. (c) Attention should be given to the process dimension within the interaction situation by considering possible changes across time. (d) Other dimensions of parent-child interaction than those in- cluded here (e.g., level of standards for performance, expectations for behavior, sensitivity to child's behavior, facial expressions) should be investigated. (e) Further validation of the P-CIRP and other parent-child inter- action measures should be undertaken. (f) The reciprocal influences of the parent and the child in the interaction sequence should be investigated. (9) A transactional analysis framework for investigation of the 185 parent-child system should be applied. (h) The interaction patterns of the parents should be studied in relation to the children's learning as indicated by their performance during the post-task testing session. (1) The interaction behavior of parents in different kinds of situations should be compared with that observed in the didactic setting of the present study. (j) The parental interaction behavior should be studied in rela- tion to parental attitudes about child-rearing. (k) The interaction patterns of the parents should be considered in relation to such characteristics as educational level, number of children, age, and amount of time typically spent with the child each week. (1) Interaction patterns of parents with children both younger and older than the preschoolers in the present study should be investigated. (m) Comparative studies should be done with children and adults of both sexes other than their parents (e.g., teachers, grandparents, older siblings). (n) Comparative studies should be done with parents and children from other kinds of family situations (e.g., variant family forms) and from groups with special needs. (0) The parent-child interaction patterns should be compared both to the interaction of the parents in other settings and of the children in other settings. (p) Parents' interactions should be studied with more than one child from the same family, particularly with children of both sexes from the same family. 186 (q) The interaction should be compared with different types of tasks, particularly those which might be sex-role related. (r) The interaction situation involving the mother-father-child triad should be investigated. ,1 c: .. 99 g; \7 / \ . o 0.4 _' e. 1 1 ."‘-,.~."-.; ', : a E: p Q , ‘ r ‘1 .5139 o . J“ L: 1.49 (Ii/211113 BIBLIOGRAPHY BIBLIOGRAPHY Aberle, D. F. Culture and socialization. In F. L. K. Hsu (Ed.), Psycho- logical anthropology: Approaches to culture and personality. Home- wood, I11.: Dorsey Press, 1961. Anderson, H. H., and Brewer, J. E. Studies of teachers' classroom per- sonalities. II. Effects of teachers' dominative and integrative contacts on children's classroom behavior. Applied Psychology Monographs, 1946, No. 8. Baldwin, A. L. Socialization and the parent-child relationship. Child Development, 1948, 19, 127-136. Baldwin, A. L., Kalhorn, J., and Breese, F. H. The appraisal of parent behavior. Psychological Monographs, 1949, 63 (4, Whole No. 299). Baldwin, A. L., Kalhorn, J., and Breese, F. H. Patterns of parent be- havior. Psychological Monographs, 1945, 58 (3, Whole No. 268). Bales, R. F. Interaction process analysis. Cambridge: Addison-Wesley, 1951. Banta, T. J. Tests for the evaluation of early childhood education: The Cincinnati Autonomy Test Battery (CATB). In J. Helmuth (Ed.), Cognitive studies, Vol. I. New York: Brunner/Mazel, 1970. Baumrind, D. Child care practices anteceding three patterns of pre- school behavior. Genetic Psychology Monographs, 1967, 75, 43-88. Baumrind, D., and Black, A. E. Socialization practices associated with dimensions of competence in pre-school boys and girls. Child Development, 1967, 38, 291-327. Bean, N. Communications. Unpublished paper, Michigan State University, 1967. Beberman, M. An emerging program of secondary school mathematics. Cambridge, Mass.: Harvard University Press, 1958. Becker, W. C. Consequences of different kinds of parental discipline. In M. L. Hoffman and L. W. Hoffman (Eds.), Review of child develop— ment research, Vol. I. New York: Russell Sage, 1964. 187 188 Becker, W. C., Peterson, D. R., Luria, Z., Shoemaker, D. J., and Hellman, L. A. Relations of factors derived from parent interview ratings to behavior problems of five-year olds. Child Development, 1962, 33, 509-535. Bee, H. L. Parent-child interaction and distractibility in 9 year old children. Merrill-Palmer Quarterly, 1967, 13, 175-190. Bee, H. L., Van Egeren, L. F., Streissguth, A. P., Nyman, B. A., and Leckie, M. S. Social class differences in maternal teaching strate- gies and speech patterns. Developmental Psychology, 1969, l, 726- 734. Beecher, C. Data-gathering devices employed in the Wisconsin studies. Journal of Experimental Education, 1961, 30, 30-47. Bell, R. Q. The American family: Myths and realities. Panel discussion at the Family and Society Conference, Merrill-Palmer Institute, Detroit, November, 1969. Bell, R. Q. A reinterpretation of the direction of effects in studies of socialization. Psyphological Review, 1968, 75 (2), 81-95. Bell, R. Q. Retrospective attitude studies of parent-child relations. Child Development, 1958, 29, 323-337. Bell, R. Q. Structuring parent-child interaction situations for direct observations. Child Development, 1964, 35, 1009-1020. Bell, R. R. The significance of the father: Some theoretical and em— pirical considerations. Unpublished paper, Temple University, n.d. Bereiter, C., and Engelmann, 8. Teaching disadvantaged children in the preschool. Englewood Cliffs, N. J.: Prentice-Hall, 1966. Berlo, D. K. The process of communication. New York: Holt, Rinehart, and Winston, 1960. Bernard, J. Marriage and family among Negroes. Englewood Cliffs, N. J.: Prentice-Hall, 1966. Bernstein, B. Elaborated and restricted codes: Their social origins and some consequences. In J. J. Gumperz and D. Mymes (Eds.), Thg_ ethnography of communication. American Anthropologist: Special Publication, 1964, 66 (6, Part 2), 55-69. Bernstein, B. Social class and linguistic development. In A. H. Halsey, J. Floud, and C. A. Anderson (Eds.), Education, economy, and so- ciety. New York: Free Press, 1961. Biller, H. B. Father, child, and sex role: Paternal determinants of personality development. Lexington, Mass.: Heath Lexington Books, 1971. 189 Bing, B. Effect of child rearing practices on development of differen- tial cognitive abilities. Child Development, 1963, 34, 631-648. Bishop, B. M. Mother-child interaction and the social behavior of chil- dren. Psychological Monographs, 1951, 65 (11, Whole No. 328). Bloom, B. Stability and change in human characteristics. New York: Wiley, 1964. Boger, M. Program Datacount. Unpublished computer program, Michigan State University, 1972. Boger, R. P., and Cunningham, J. L. Differential socialization patterns of preschool children. Interim report #2. Michigan State Univer- sity, Early Childhood Research Center, 1970. Boger, R. P., Kuipers, J., and Beery, M. Parents as primary change agents. Experimental program report, Michigan State University, Head Start Research Center, 1969. Bonner, Hubert. On being mindful of man. Boston: Houghton, 1965. Borgatta, E. F., Couch, A. S., and Bales, R. F. Some findings relevant to the great man theory of leadership. American Sociological Re- view, 1954, 19, 755-759. Brody, G. F. Maternal child-rearing attitudes and child behaviors. Developmental Psychology, 1969, 1, 66. Brody, S. Patterns of mothering. New York: International Universities Press, 1956. Bronfenbrenner, U. Toward a theoretical model for the analysis of parent- child relationships in a social context. In J. Glidewell (Ed.), Parental attitudes and child behavior, Conference on Community Mental Health Research, Washington University, St. Louis, 1961. Brophy, J. E. Maternal teaching style in a structured mother-child interaction. Unpublished doctoral dissertation, University of Chicago, 1967. Brophy, J. E. Mothers as teachers of their own preschool children: The influence of socioeconomic status and task structure on teach- ing specificity. Child Development, 1970, 41, 79—94. Brunner, C. Deprivation--its effects, its remedies. Educational Leader- ship, 1965, 23, 103-107. Bugelski, B. R. The psychology of learning. New York: Holt, 1956. Caldwell, B. M., and Hersher, L. Mother-infant interaction during the first year. Merrill-Palmer Quarterly, 1964, 10, 119-128. 190 Carmichael, L., Roberts, 5., and Wessell, N. A study of the judgment of manual expressions as presented in still and motion pictures. Journal of Social Psychology, 1937, 8, 115-142. Carroll, J. B. Words, meanings, and concepts. Harvard Educational Re- view, 1964, 34, 178-202. Casey, J. E., and Weaver, B. E. An evaluation of lecture method and small group method of teaching in terms of knowledge of content, teacher attitude, and social status. Journal of Colorado—Wyoming Academy of Science, 1956, 4, 54. Chamberlin, R. W. A study of an interview method for identifying family authority pattern. Genetic Psychology Monographs, 1969, 80, 129- 148. Chapman, J. E. Early childhood research and development needs, gaps, and imbalances: overview. Interagency Panel on Early Childhood Research and Development, Washington, D. C., 1972. Christensen, H. T. Development of the family field of study. In H. T. Christensen (Ed.), Handbook of marriage and the family. Chicago: Rand McNally, 1964. Child, I. L. Socialization. In G. Lindzey (Ed.), Handbook of social psychology. Cambridge, Mass.: Addison-Wesley, 1954. Clausen, J. A. (Ed.). Socialization and society. Boston: Little, Brown, and Company, 1968. Cloyd, J. S. Patterns of role behavior in informal interaction. Socio- metry, 1964, 27, 161-173. Crandall, V. J., Preston, A., and Rabson, A. Maternal reactions and the development of independence and achievement behavior in young chil- dren. Child Development, 1960, 31, 243-251. Cunningham, J. L., and Boger, R. P. Father-Child Interaction: Nine Block Sorting Task. Unpublished instrument description, Michigan State University, Head Start Research Center, 1969. Cunningham, J. L., and Boger, R. P. Parent-Child Interaction Rating Procedure. Unpublished instrument description, Michigan State University, Early Childhood Research Center, 1970. Dager, E. Z. Socialization and personality development in the child. In H. T. Christensen (Ed.), Handbook of marriage and the family. Chicago: Rand McNally, 1964. Davis, K. Human society. New York: Macmillan, 1949. Davitz, J. R., and Davitz, L. J. The communication of feelings by content-free speech. Journal of Communication, 1959, 9, 6—13. 191 Delaney, E. J. Parental antecedents of social aggression in young chil— dren. Dissertation Abstracts, 1965, 26, 1763. Deutsch, M. The disadvantaged child and the learning process. In A. H. Passow (Ed.), Education in depressed areas. New York: Teachers College Press, Columbia University, 1963. Douglas, J. W. B., Lawson, A., and Cooper, J. B. Family interaction and the activities of young children. Journal of Child Psychology and Psychiatry, 1968, 9, 157-171. Dwyer, F. M., Jr. Adapting visual illustrations for effective learning. Harvard Educational Review, 1967, 37, 250-263. Eisenberg, A. M., and Smith, R. R., Jr. Nonverbal communication. Indianapolis: Hobbs-Merrill, 1971. Ekman, P., and Friesen, W. The repertoire of nonverbal behavior: Cate- gories, origins, usage, and coding. Semiotics, 1969, 1, 63-92. Ekman, P., Friesen, W., and Ellsworth, P. Emotion in the human face: Guidelines for research and a review of findings. Elmsford, N. Y.: Pergamon Press, 1971. Elder, G. H., Jr. Parental power legitimation and its effect on the adolescent. Sociometry, 1963, 26, 50-65. Elkin, F. The child and society: The process of socialization. New York: Random House, 1960. Emmerich, W. Variations in the parent role as a function of the parent‘s sex and the child's sex and age. Merrill-Palmer Quarterly, 1962, 8, 3-11. Emmerich, W., and Smoller, F. The role patterning of parental norms. Sociometry, 1964, 27, 382-390. Eron, L. D., Banta, T. J., Walder, L. 0., and Laulicht, J. H. Compari- son of data obtained from mothers and fathers on childrearing prac- tices and their relation to child aggression. Child Development, 1961, 32, 457-472. Eysenck, H. J. Handbook of abnormal psychology. New York: Basic Books, 1961. Fields, J. E. Experiences of efficacy within the family, and adaptive ego functioning in the child. Dissertation Abstracts, 1969, 29, 2632. Finn, J. D. Multivariance--univariate and multivariate analysis of variance and covariance: A Fortran IV program. State University of New York at Buffalo, School of Education, 1967. 192 Finn, J. D. Multivariance--univariate and multivariate analysis of variance, covariance, and regression: A Fortran IV program. Ver— sion 4-1. State University of New York at Buffalo, School of Edu- cation, 1971. Gage, N. L. (Ed.) Handbook of research on teaching, Chicago: Rand McNally, 1963. Getzels, J. W., and Jackson, P. W. The teacher's personality and char- acteristics. In N. L. Gage (Ed.), Handbook of research on teaching, Chicago: Rand McNally, 1963. Gildea, M. C. L., Glidewell, J. C., and Kantor, M. D. Maternal atti- tudes and general adjustment. In J. Glidewell (Ed.), Parental atti- tudes and child behavior. Conference on Community Mental Health Research, Washington University, St. Louis, 1961. Glidewell, J. (Ed.) Parental attitudes and child behavior. Conference on Community Mental Health Research, Washington University, St. Louis, 1961. Gordon, E. W. Characteristics of socially disadvantaged children. 527 view of Educational Research, 1965, 35, 377-388. Goslin, D. A. (Ed.) Handbook of socialization theory and research. Chicago: Rand McNally, 1969. Haggard, E. A., Brekstad, A., and Skard, A. G. On the reliability of the anamnestic interview. Journal of Abnormal and Social Psychology, 1960, 61, 311-318. Hall, C. S., and Lindzey, G. Psychoanalytic theory and its applications in the social sciences. In G. Lindzey (Ed.), Handbook of social psychology. Vol. 1. Cambridge, Mass.: Addison-Wesley, 1954. Hall, E. Silent language. Garden City, N. Y.: Doubleday, 1959. Hart, I. Maternal child—rearing practices and authoritarian ideology. Journal of Abnormal and Social Psychology, 1957, 55, 232-237. Hatfield, J. S., Ferguson, L. R., and Alpert, R., Mother-child inter- action and the socialization process. Child Development, 1967, 38, 365-414. Henderson, K. 8. Research on teaching secondary school mathematics. In N. L. Gage (Ed.), Handbook of research on teaching. Chicago: Rand McNally, 1963. Hess, R. D. Education and rehabilitation: The future of the welfare class. Marriage and Family Living, 1964, 26, 422-429. HeSS, R. D. Social class and ethnic influences on socialization. In P. H. Mussen (Ed.), Manual of child psychology. Vol. II. New York: Wiley, 1970. 193 Hess, R. D., and Bear, R. M. Early education. Chicago: Aldine, 1968. Hess, R. D., Olim, E. G., and Shipman, V. C. Relationship between moth- er's language styles and cognitive styles of urban preschool chil- dren. Paper presented at biennial meeting of the Society for Re- search in Child Development, Minneapolis, Minn., March, 1965. Hess, R. D., Shipman, V. C., Brophy, J. E., and Bear, R. M. The cogni- tive environments of urban preschool children. Final report on Re- search Grant §R34, University of Chicago, Graduate School of Edu- cation, 1968. Hess, R. D., Shipman, V. C., Brophy, J. E., and Bear, R. M. The cogni- tive environments of urban preschool children: Follow-up phase. Research report, University of Chicago, Graduate School of Educa— tion, 1969. Heyns, R. W., and Lippitt, R. Systematic observational techniques. In G. Lindzey (Ed.), Handbook of socialgpsychology. Vol. I. Reading, Mass.: Addison-Wesley, 1954. Hildebrand, V. Introduction to early childhood education. New York: Macmillan, 1971. Hoffman, L. W., and Lippitt, R. The measurement of family life vari- ables. In P. H. Mussen (Ed.), Handbook of research methods in child development. New York: Wiley, 1960. Hoffman, M. L. An interview method for obtaining descriptions of parent- child interaction. Merrill—Palmer Quarterly, 1957, 4, 76—83. Hoffman, M. L. Personality, family structure, and social class as ante- cedents of parental power assertion. Child Development, 1963, 34, 869-884. Hoffman, M. L. Power assertion by the parent and its impact on the child. Child DevelOpment, 1960, 31, 129-143. Homans, G. C. Social behavior as exchange. American Journal of Soci— ology, 1958, 63, 597-606. Hughes, M. M., and associates. The model of good teaching. In J. Raths, J. R. Pancella, and J. S. Van Ness (Eds.), Studying teach— ing. Englewood Cliffs, N. J.: Prentice-Hall, 1967. Hurley, J. Parental acceptance-rejection and childrens' intelligence. Merrill-Palmer Quarterly, 1965, 11, 19-31. Hutt, S. J., and Hutt, C. Direct observation and measurement of behav- ior. Springfield, I11.: Thomas, 1970. Jackson, P. W. Verbal solutions to parent-child problems. Child Development, 1956, 27, 339-351. 194 Kagan, J. Acquisition and significance of sex typing and sex role iden- tity. In M. L. Hoffman and L. W. Hoffman (Eds.), Review of child development research. Vol. I. New York: Russell Sage, 1964. Kamii, C. K., and Radin, N. L. Class differences in the socialization practices of Negro mothers. Journal of Marriage and the Family, 1967, 29, 302-310. Keller, S. The American lower-class family. Albany, N. Y.: New York State Division for Youth, 1966. Klineberg, O. Negro-white differences in intelligence test performance: A new look at an old problem. American Psychologist, 1963, 18, 198-203. Kluckhohn, F. Variations in value orientations. Evanston, 111.: Row, Peterson, and Co., 1961. Kogan, K., and Wimberger, H. C. An approach to defining mother-child interaction styles. Perceptual and Motor Skills, 1966, 23, 1171- 1177. Kohn, M. L. Social class and parent-child relationships: An interpre- tation. American Journal of Sociology, 1963, 68, 471-480. Kohn, M. L., and Carroll, E. E. Social class and the allocation of parental responsibilities. Sociometry, 1960, 23, 372-392. Krech, D., and Crutchfield, R. S. Elements of psychology. New York: Alfred A. Knopf, 1958. Lafore, G. G. Practices of_parents in dealing with preschool children. New York: Teachers College Press, Columbia University, 1945. Leler, H. Mother-child interaction and language performance in young disadvantaged Negro children. Unpublished doctoral dissertation, Stanford University, 1970. LeMasters, E. E. Parents in modern America. Homewood, 111.: Dorsey, 1970. Lesser, G. S. (Ed.) Psychology and educational practice. Glenview, 111.: Scott, Foresman, and Co., 1971. Levin, H., Hilton, T. L., and Leiderman, G. F. Studies in teacher be— havior. Journal of Experimental Education, 1957, 26, 81-91. Levine, J., Fishman, C., and Kagan, J. Social class and sex as determi- nants of maternal behavior. American Journal of Orthopsychiatry, 1967, 37, 397. Livson, N. Parental behavior and children's involvement with their parents. Journal of Genetic Psychology, 1966, 109, 173-194. 195 Longabaugh, R. Category system for coding interpersonal behavior as social exchange. Sociometry, 1963, 26, 319-344. Loomis, C. P. Social systems. New York: D. Van Nostrand, 1960. Lumsdaine, A. A. Instruments and media of instruction. In N. L. Gage (Ed.), Handbook of research on teaching, Chicago: Rand McNally, 1963. Lynn, D. B. The process of learning parental and sex-role identification. Journal of Marriage and the Family, 1966, 28, 466-470. Lynn, D. B. Sex differences in identification development. Sociometry, 1961, 24, 372-383. Lynn, D. B. Sex-role and parental identification. Child Development, 1962, 33, 555-564. Lytton, H. Observation studies of parent-child interaction: A methodo- logical review. Child Development, 1971, 42, 651—684. Magrabi, F. M., Paolucci, B., and Heifner, M. E. Framework for studying family activity patterns. Journal of Home Economics, 1967, 59, 714-719. Marshak, M. A method for evaluating child-parent interaction under con- trolled conditions. Journal of Genetic Psychology, 1960, 97, 3-22. May, R. Psycholoqy and the human dilemma. New York: D. Van Nostrand, 1967 O McCall, R. B. The use of multivariate procedures in developmental psychology. In P. H. Mussen (Ed.), Manual of child psychology. Vol. II. New York: Wiley, 1970. McCandless, B. R. Childhood socialization. In D. A. Goslin (Ed.), Handbook of socialization theory and research. Chicago: Rand McNally, 1969. McGee, H. M. Measurement of authoritarianism and its relation to teachers' classroom behavior. Genetic Psychology Monographs, 1955, 52, 89-146. McGrath, L. P., and Scobey, J. (Eds.) What is a father. New York: Simon and Schuster, 1969. McGrath, L. P., and Scobey, J. (Eds.) What is a mother. New York: Simon and Schuster, 1968. McKeachie, W. J. Research on methods of teaching. In N. Sanford (Ed.), The American college. New York: Wiley, 1962. McKinley, D. G. Social class and family life. Glencoe, I11.: Free Press, 1964. 196 Merrill, B. A measurement of mother-child interaction. Journal of Abnormal and Social Psychology, 1946, 41, 37-49. Michael, R. E. The relative effectiveness of two methods of teaching certain topics in ninth grade algebra. Mathematics Teacher, 1949, 42, 83-87. Miller, G. A. Language and communication. New York: McGraw-Hill, 1951. Milmoe, S., Novey, N. S., Kagan, J., and Rosenthal, R. The mother's voice: Postdictor of aspects of her baby's behavior. Proceedings of the 76th annual convention of the American Psychological Asso- ciation, 1968, 3, 463-464. Mischel, W. Preference for delayed reward and social responsibility. Journal of Abnormal and Social Psychology, 1961, 62, 1-7. Mischel, W. Sex typing and socialization. In P. H. Mussen (Ed.), Manual of child psychology. Vol. II. New York: Wiley, 1970. Moreno, J. L. The sociometric reader. New York: Free Press, 1960. Moss, H. A. Methodological issues in studying mother-infant interac- tions. American Journal of Orthopsychiatry, 1965, 35, 482-486. Moss, H. A. Sex, age and state as determinants of mother-infant inter- action. Merrill—Palmer Quarterly, 1967, 39, 401-408. Moustakas, C. E., Sigel, I. E., and Schalock, M. D. An objective method for the measurement and analysis of child-adult interaction. Child Development, 1956, 27, 109-134. Mussen, P. H. (Ed.) Handbook of research methods in child development. New York: Wiley, 1960. Newcomb, M. An approach to the study of communicative acts. Psycho- logical Review, 1953, 60, 393-404. Nolte, D. L. Children learn what they live. Campbell, Ca1if.: John Philip Co., 1963. O'Rourke, J. F. Field and laboratory: The decision-making behavior of family groups in two experimental conditions. Sociometry, 1963, 26, 422-435. Osofsky, J. D. The shaping of mother's behavior by children. Journal of Marriage and the Family, 1970, 32, 400-405. Ostrovsky, E. S. Father to the child. New York: G. P. Putnam's Sons, 1959. Pavenstedt, E. A comparison of the childrearing environment of upper- 1ower and lower-lower class families. American Journal of Ortho- psychiatry, 1965, 35, 89-98. 197 Radke, M. J. The relation of parental authority to children's behavior and attitudes. University_of Minnesota Child Welfare Monographs, 1946, 22. Read, K. H. The nursery school. Philadelphia: W. B. Saunders Co., 1971. Report to the President: White House Conference on Children. Washing- ton, D. C.: U. S. Government Printing Office, 1970. Reusch, J. Synopsis of the theory of human communication. Psychiatry, 1953, 16, 215-243. Reynolds, J. H., and Glaser, R. Effects of repetition and spaced review upon retention of a complex learning task. Journal of Educational Psychology, 1964, 55, 297-308. Rheingold, H. The measurement of maternal care. Child Development, 1960, 31, 565-575. Rosen, B. C. Social class and the child's perception of the parent. Child Development, 1964, 35, 1147-1153. Rothbart, M. K., and Maccoby, E. E. Parents' differential reactions to sons and daughters. Journal of Personality and Social Psychology, 1966, 4, 237-243. Ryans, D. G. Characteristics of teachers. Washington, D. C.: American Council on Education, 1960. Sanders, L. W. Issues in early mother-child interaction. Journal of the American Academy of Child Psychiatry, 1962, 1, 141-166. Santostefano, S. Miniature situations and methodological problems in parent-child interaction research. Merrill-Palmer Quarterly, 1968, 14, 285-312. Sapir, E. The unconscious patterning of behavior in society. In D. Mandelbaun (Ed.), Selected writings of Edward Sapir in language, culture, and personality. Berkeley: University of California Press, 1949. Schaefer, E. A circumplex model for maternal behavior. Journal of Abnormal and Social Psychology: 1959, 59, 226-235. Schaefer, E. S. Converging conceptual models for maternal behavior and for child behavior. In J. Glidewell (Ed.), Parental attitudes and child behavior. Conference on Community Mental Health Research, Washington University, St. Louis, 1963. Schmidt, W. H., and Scheifley, V. Jeremy D. Finn's multivariate and univariate analysis of variance, covariance, and regression: Adaptation to the MSU CDC 6500 computer. Unpublished paper, Michigan State University, 1972. 198 Schulman, F. R., Shoemaker, D. J., and Moelis, I. Laboratory measure- ments of parental behavior. Journal of Consulting Psychology, 1962, 26, 109-114. Sears, R. R., Maccoby, E. E., and Levin, H. Patterns of child rearing. Evanston, 111.: Row, Peterson, 1957. Sears, R. R., Rau, L., and Alpert, R. Identification and child rearing. Stanford, Calif.: Stanford University Press, 1965. Sears, R. R., Whiting, J. W. M., Nowlis, V., and Sears, P. S. Some childrearing antecedents of aggression and dependency in young children. Genetic Psychology Monographs, 1953, 47, 135-234. Sewell, W. H. Some recent developments in socialization theory and research. Annals of American Academy of Political and Social Science, 1963, 349, 163-181. Shakow, D. Research in child development: A case illustration of the psychologist's dilemma. American Journal of Orthopsychiatry, 1959, 29, 45-59. Shibutani, T. Societyyand personality. Englewood Cliffs, N. J.: Prentice-Hall, 1961. Sigel, I. E. Influence techniques: A concept used to study parental behaviors. Child Development, 1960, 31, 799-806. Sigel, I. E., Hoffman, M. L., Dreyer, A. S., and Torgoff, I. Influence techniques used by parents to modify the behavior of children: A case presentation. American Journal of Orthopsychiatry, 1957, 27, 356-364. Smilansky, S. The effects of sociodramatic play on disadvantaged pre- school children. New York: Wiley, 1968. Smith, D. E. P., Wood, R. L., Downer, J. W., and Raygor, A. L. Reading improvement as a function of student personality and teaching method. Journal of Educational Psychology, 1956, 47, 47-58. Smith, H. T. A comparison of interview and observation methods of maternal behavior. Journal of Abnormal and Social Psychology, 1958, 57, 278-282. Solomon, R. L., and Wynne, L. C. Traumatic avoidance learning: Acqui- sition in normal dogs. Psychological Monographs, 1953, 67 (4, Whole No. 354). Spodek, B. Teaching in the early years. Englewood Cliffs, N. J.: Prentice-Hall, 1972. Stern, G. G., Caldwell, B. M., Hersher, L., Lipton, E. L., and Richmond, J. B. A factor-analytic study of the mother-infant dyad. Child Development, 1969, 40, 163-182. 199 Stone, C. Three-generation influences on teenagers' conceptions of family culture and parent-child relationships. Marriage and Family» Living, 1962, 24, 287-288. Struas, M. A. Conjugal power structure and adolescent personality. Marriage and Family Living, 1962, 24, 17-25. Straus, M. A. The influence of sex of children and social class on in- strumental and expressive family roles in a laboratory setting. Sociology and Social Research, 1967, 52, 7-21. Straus, M. A. Measuring families. In H. T. Christensen (Ed.), Handbook of marriage and the family. Chicago: Rand McNally, 1964. Straus, M. A. Power and support structure of the family in relation to socialization. Journal of Marriage and the Family, 1964, 26, 318- 326. Strodtbeck, F. L. Family interaction, values and achievement. In D. C. McClelland, et al. (Eds.), Talent and society. Princeton, N. J.: D. Van Nostrand, 1958. Symonds, P. M. The psychology of parent-child relationships. New York: Appleton-Century-Crofts, 1939. Theodorson, G. The relationship between leadership and popularity roles in small groups. American Sociological Review, 1957, 22, 58-67. Thomas, D. S., and associates. Some new techniques for studying social behavior. Child Development Monographs, 1929, 1. Wallen, N. E., and Travers, R. M. W. Analysis and investigation of teaching methods. In N. L. Gage (Ed.), Handbook of research on teaching. Chicago: Rand McNally, 1963. Walter, J., Connor, R., and Zunich, M. Interaction of mothers and chil- dren from lower-class families. Child Development, 1964, 35, 433- 440. Walters, J., Pearce, D., and Dahms, L. Affectional and aggressive be- havior of preschool children. Child Development, 1957, 44, 227-249. Walters, J., and Stinnett, N. Parent-child relationships: A decade review of research. Journal of Marriage and the Family, 1971, 33, 70-111. Waters, E., and Crandall, V. J. Social class and observed maternal be- havior from 1940 to 1960. Child Development, 1964, 35, 1021-1032. Weick, K. E. Systematic observational methods. In G. Lindzey and F. Aronson (Eds.), Handbook of social psychology, (2nd ed.) Vol. 2. Reading, Mass.: Addison-Wesley, 1968. 200 Weiss, J. Fathers, the missing link: Some methodological implications for environmental research. Division Generator, American Educa- tional Research Association Division G, 1970, l (2), 6-9. White, B. L. The Harvard Preschool Project: An ethoecological study of the development of competence. Symposium presented at biennial meeting of the Society for Research in Child Development, Santa Monica, Calif., March 27, 1969. White, R. W. The abnormal personaliEy. New York: Ronald Press, 1956. Williams, F., and Sundene, B. Dimensions of recognition: Visual vs. vocal expression of emotion. Communication Review, 1965, 13, 44-52. Wilson, J. A. R., Robeck, M. C., and Michael, W. B. Psychological foundations of learning and teaching. New York: McGraw-Hill, 1969. Winch, R. F. Identification and its familial determinants. New York: Hobbs-Merrill, 1962. Witkin, H. A., Dyk, R. B., Paterson, H. F., Goodenough, D. R., and Karp, S. A. Psychological differentiation: Studies of development. New York: Wiley, 1962. Wright, D. J. Jeremy D. Finn's multivariate and univariate analysis of variance and covariance: A Fortran IV program--Adaptation for the MSU CDC 3600. Occasional paper No. 8, Michigan State University, College of Education, Office of Research Consultation, 1970. Wright, H. F. Observational child study. In P. H. Mussen (Ed.), Hand- book of research methods in child development. New York: Wiley, 1960. Yarrow, L. J. Separation from parents during early childhood. In M. L. Hoffman and L. W. Hoffman (Eds.), Review of child development re- search. Vol. 1. New York: Russell Sage, 1964. Yarrow, M. R. Problems of methods in parent-child research. Child Development, 1963, 34, 215-226. Yarrow, M. R., Campbell, J. D., and Burton, R. V. Child rearing: An inguiry into research and methods. San Francisco: Jossey-Bass, 1968. Yarrow, M. R., Campbell, J. D., and Burton R. V. Reliability of maternal retrospection: A preliminary report. Family Process, 1964, 3, 207- 218. Zigler, E. Social class and the socialization process. Review of Edu- cational Research, 1970, 40, 87-110. Zigler, E., and Child, I. L. Socialization. In G. Lindzey and E. Aronson (Eds.), The handbook of social psychology, (2nd ed.) Vol. 3. Reading, Mass.: Addison-Wesley, 1969. 201 Zunich, M. A study of relationships between child rearing attitudes and maternal behavior. Journal of Experimental Education, 1961, 30, 231-241. APPENDICES APPENDIX I INSTRUMENTS APPENDIX I -A MOTHER-CHILD INTERACTION: TOY SORTING TASK The Toy Sorting Task was developed at The University of Chicago in the project entitled "Cognitive Environments of Urban Preschool Chil- dren."1 Robert D. Hess and Virginia C. Shipman directed the major pro- ject, and Mildred Levine assumed primary responsibility for the design of the tasks used for the research instrument. Specific materials and procedures are described in the sections which follow. Materials Nine toys and a partitioned board are used for this task. The board is a circle 12 3/4" in diameter and made of brown tempered mason- ite. The board is divided into three equal sections by white lines (®). The toys which are used for this task are as follows: (a) chairs-- three small plastic or wooden chairs (dollhouse furniture), (b) spoons-- three plastic picnic spoons, and (c) cars--three small metal toy cars. The toys are of three colors, with one toy of each type in each of the following colors: (a) red, (b) green, and (c) yellow. All are bright basic shades of approximately the same hue, value, and intensity. 1R. D. Hess, V. C. Shipman, J. E. Brophy, and R. M. Bear. The cognitive environments of urban preschool children. Final report on Research Grant #R34, University of Chicago Graduate School of Education, 1968. 202 203 Procedure The two sorting methods are taught to the mother while the child is out of the room. The task begins with the board empty and the toys in random order on the table. The §_says, HERE ARE SOME TOYS. THERE ARE DIFFERENT WAYS THEY CAN BE PUT TOGETHER ON THE BOARD. §_sorts by object, randomly varying the colors in each section. After sorting, §_says, THESE GO TOGETHER BECAUSE THEY'RE ALL . . . (pause), and points to each group to elicit the answer (spoons, chairs, cars). E then says, THE TOYS CAN BE PUT TOGETHER IN ANOTHER WAY, TOO. :E takes the toys off the board and sorts by color, with random placement within each section. E again says, THESE GO TOGETHER BECAUSE THEY'RE ALL . . . (pause), and points to each group to elicit the answer (reg, yellow, green). After demonstrating both sorts, §_says, I'D LIKE YOU TO TEACH WHAT I HAVE SHOWN YOU--TO PUT THE TOYS TOGETHER IN THESE TWO WAYS. I'LL BE OUTSIDE WHILE YOU TEACH HIM. AFTER YOU'RE SURE HE UNDERSTANDS H9W_TO PUT THE TOYS TOGETHER IN THESE TWO WAYS, AND KNOWS WHY THE TOYS IN EACH GROUP BELONG TOGETHER, CALL ME BACK INTO THE ROOM (TRAILER). I'LL ASK HIM TO PUT THE TOYS ON THE BOARD IN THE TWO WAYS YOU HAVE TAUGHT HIM-- AND TO DO IT WITHOUT ANY HELP FROM ME OR FROM YOU. TAKE AS MUCH TIME AS YOU NEED TO TEACH HIM. WHEN YOU'VE FINISHED, BE SURE TO CALL ME BACK INTO THE ROOM (TRAILER). The §_removes the toys from the board and then brings in the child and leaves the room. The mother is allowed complete freedom of method. The teaching task is terminated when the mother summons the examiner or at the end of forty minutes (from the time the §_1eaves the room), 204 whichever occurs first.1 When E returns, the child is asked to repeat the sorts. §_first takes the toys off the board and randomizes them. §_then says, SHOW ME ONE OF THE WAYS TO PUT THE TOYS ON THE BOARD THAT YOUR MOTHER TAUGHT YOU. After the child sorts the toys, §_says, THESE GO TOGETHER BECAUSE THEY'RE ALL . . . (pause), and points to each group to elicit the answer (spoons, chairs, cars; or red, yellow, green). Finally, §_says, NOW SHOW ME THE OTHER WAY TO PUT THE TOYS ON THE BOARD THAT YOUR MOTHER TAUGHT YOU. After the child has sorted the toys, E says, THESE GO TOGETHER BECAUSE THEY'RE ALL . . . (pause), and points to each group to elicit the answer (spoons, chairs, cars; or red, yellow, green). If the child does not sort correctly by color apd_by object on the first two sorts, a third trial is administered. Probing by the examiner is restricted to repetition and/or rephrasing of the questions and at- tempts to clarify the child's intent when it is unclear whether he has finished an intended sort or which toys are intended to be in particular groups. Probing continues until the examiner ascertains which toys are considered as members of a group and the total number of groups intended by the child. No attempt is made to stop prompting by the mother.2 The standard testing procedures must be followed for three trials. At the conclusion of the third trial, the examiner may prompt the child to enable him to complete the task correctly so the session is ended on a positive note for both the mother and the child. 1This is a deviation from the Hess and Shipman procedure, which allowed unlimited time for the teaching session. 2This is a deviation from the Hess and Shipman procedure, which did not permit prompting by the mother except for support or encouragement. 205 Scoring A record is made of the child's performance on the post-task test. Credit is given if the child meets the criteria without help. Credit is given in cases where he spontaneously corrects errors and also cases where the child first responds correctly but then becomes confused under continued questioning. Responses following probing by the examiner are allowed to raise the child's score (when they involve passing additional criteria), but not to lower it (since probing may have induced confusion or inhibition). Credit is not given for verbalization unless the child has previously sorted correctly (three groups, clearly differentiated). A note is made of responses prompted by the mother, but no credit is given for them. A copy of the scoring form is shown in Figure I-l. The entire session is videotaped for subsequent analysis. Anec- dotal notes of the session are made by the examiner. Both the mother and the child are later rated from the tape with the Parent-Child Interaction Rating Procedure. 206 MOTHER-CHILD INTERACTION: TOY SORTING TASK Name Date Child ID Examiner . Partial Full Sort Grou in p g Explanation Explanation Notes Color First Sort Object Color Second Sort Object Color Third Sort Object Comments FIGURE I-l. Scoring form for Mother-Child Interaction: Toy Sorting Task. APPENDIX 1‘8 FATHER-CHILD INTERACTION: NINE BLOCK SORTING TASK The Nine Block Sorting Task was developed by Jo Lynn Cunningham and Robert P. Boger, Head Start Research Center, Michigan State Univer- I sity. It was developed through the project entitled "Differential Socialization Patterns of Preschool Children," directed by the same authors.3 Aspects of the task were adapted from the Toy Sorting Task, developed by Robert D. Hess and Virginia C. Shipman, The University of . 4 . . . . . Chicago. SpeCific materials and procedures are described in the sec- tions which follow. Materials Nine blocks and a board with three dowels attached are used in this task. The board is of tempered masonite. It is a 5 1/2" x 20" rectangle 1J. L. Cunningham and R. P. Boger. Father-Child Interaction: Nine Block Sorting Task. Unpublished instrument description, Michigan State University, Head Start Research Center, 1969. 2Sharon Evans and Eileen Earhart assisted with the development of this instrument. Helpful suggestions by Chris Walter, Carol Culler, Marilyn Beery, Marilyn Jenks, and Bill Brown are also acknowledged. 3R. P. Boger and J. L. Cunningham. Differential socialization patterns of preschool children. Interim report #2, Michigan State University, Early Childhood Research Center, 1970. 4R. D. Hess, V. C. Shipman, J. E. BrOphy, and R. M. Bear. The cognitive environments of urban preschool children. Final report on Research Grant #R34, University of Chicago Graduate School of Education, 1968. 207 208 with three natural-colored dowels (6" x 1/2") evenly spaced along the length of the board ( L o o 0*] ). The nine blocks are standard Fisher-Price Creative Blocks. All are polyethylene and are 1 1/4" thick with a diameter of approximately 2 3/4" and a 5/6" hole in the center. They differ on two dimensions: (a) color--red, blue, and yellow; and (b) shape--square, circle, and triangle (pie-shaped). There is one shape of each color and one color of each shape. Procedure The two sorting methods are taught to the father while the child is out of the room. The task begins with the board empty and the blocks in random order on the table. E says, HERE ARE SOME BLOCKS. THERE ARE DIFFERENT WAYS THEY CAN BE PUT TOGETHER ON THE BOARD. §_sorts by £2i2£1 randomly varying the shapes. After sorting, E_says, THESE GO TOGETHER BECAUSE THEY'RE ALL . . . (pause), and points to each group to elicit the answer (red, blue, yellow). §_then says, THE BLOCKS CAN BE PUT TOGETHER IN ANOTHER WAY, TOO. E takes the blocks off the poles and sorts by shape, with random place- ment with respect to color on each pole. E again says, THESE GO TO- GETHER BECAUSE THEY'RE ALL . . . (pause), and points to each group to elicit the answer (circles, squares, triangles). After demonstrating both sorts, §_says, I'D LIKE YOU TO TEACH WHAT I HAVE SHOWN YOU--TO PUT THE BLOCKS TOGETHER IN THESE TWO WAYS. I'LL BE OUTSIDE WHILE YOU TEACH HIM. AFTER YOU'RE SURE HE UNDERSTANDS HOW TO PUT THE BLOCKS TOGETHER IN THESE TWO WAYS, AND KNOWS WHY THE BLOCKS IN EACH GROUP BELONG TOGETHER, CALL ME BACK INTO THE ROOM 209 (TRAILER). I'LL ASK HIM TO PUT THE BLOCKS ON THE BOARD IN THE TWO WAYS YOU HAVE TAUGHT HIM--AND TO DO IT WITHOUT ANY HELP FROM ME OR FROM YOU. TAKE AS MUCH TIME AS YOU NEED TO TEACH HIM. WHEN YOU'VE FINISHED, BE SURE TO CALL ME BACK INTO THE ROOM (TRAILER). E removes the blocks from the pole and then brings in the child and leaves the room. The father teaches the sorting task to the child. He is allowed complete freedom of method. The teaching task is termi- nated when the father summons the examiner or at the end of forty min- utes (from the time B leaves the room), whichever occurs first. When E returns, the child is asked to repeat the sorts. E_first takes the blocks off the poles and randomizes them. §_then says, SHOW ME ONE OF THE WAYS TO PUT THE BLOCKS ON THE BOARD THAT YOUR DADDY TAUGHT YOU. After the child has sorted the blocks, §_says, THESE GO TOGETHER BECAUSE THEY'RE ALL . . . (pause), and points to each group to elicit the answer (circles, sqgares, triangles; or red, ygllow, blue). Finally, E says, NOW SHOW ME THE OTHER WAY TO PUT THE BLOCKS ON THE BOARD THAT YOUR DADDY TAUGHT YOU. After the Child has sorted the blocks, §_says, THESE GO TOGETHER BECAUSE THEY'RE ALL . . . (pause), and points to each group to elicit the answer (circles, sqgares, tri- angles; or red, yellow, blue). If the child does not correctly sort by color EBS.bY object on the first two sorts, a third trial is administered. Probing by the examiner is restricted to repetition and/or rephrasing of the questions and at- tempts to clarify the child's intent when it is unclear whether he has finished an intended sort or which toys are intended to be in particular groups. No attempt is made to stop prompting by the father. 210 The standard testing procedures must be followed for three trials. At the conclusion of the third trial, the examiner may prompt the child to enable him to complete the task correctly so the session is ended on a positive note for both the father and the child. m A record is made of the child's performance on the post-task test. Credit is given if the child meets the criteria without help, including cases where he spontaneously corrects errors and also cases where the child first responds correctly but then becomes confused under continued questioning. Responses following probing by the examiner are allowed to raise the child's score (when they involve passing additional cri- teria), but not to lower it, since probing may have induced confusion or inhibition. Credit is not given for verbalization unless the child has previously sorted correctly (three groups, clearly differentiated). A note is made of responses prompted by the father, but no credit is given for them. A copy of the scoring form is shown in Figure I-2. The entire session is videotaped for subsequent analysis. Anec- dotal notes of the session are made by the examiner. Both the father and the child are later rated from the tape with the Parent-Child Interaction Rating Procedure. 211 FATHER-CHILD INTERACTION: NINE BLOCK SORTING TASK Name Date Child ID Examiner . Partial Full So t Grou n . . Notes r pl 9 Explanation Explanation Color First Sort Shape Color Second Sort Shape Color Third Sort Shape Comments FIGURE I-2. Scoring form for Father-Child Interaction: Nine Block Sorting Task. APPENDIX I-C PARENT-CHILD INTERACTION RATING PROCEDURE The Parent-Child Interaction Rating Procedure (P-CIRP) is an ob- servational rating system designed for use with videotaped dyadic inter- action situations, particularly those involving a parent-child struc- tured teaching task. It was developed by Jo Lynn Cunningham and Robert P. Boger, Early Childhood Research Center, Michigan State University.l' 2 The instrument was developed through the project entitled "Differential Socialization Patterns of Preschool Children," directed by the same 3 . . . . . . authors. SpeCific procedures are described in the sections which follow. General Procedures This rating system contains two sections, each designed for a dif- ferent portion of the parent-child teaching situation. In addition, a general information section is included which pertains to the entire 1J. L. Cunningham and R. P. Boger. Parent-Child Interaction Rating Procedure. Unpublished instrument description, Michigan State University, Early Childhood Research Center, 1970. 2Special thanks are given to Sharon Evans and Marilyn Olson for their assistance with the development of this instrument. Jere Brophy and Ellen Strommen were also helpful in reacting to an early draft of the instrument description. 3 . . . . . . R. P. Boger and J. L. Cunningham. Differential soc1alization patterns of preschool children. Interim report #2, Michigan State University, Early Childhood Research Center, 1970. 212 213 session. Concurrent behavioral ratings for the parent and the child are made each twenty seconds during the observation. Each frame (represent- ing twenty seconds) for each person is rated as an individual unit. Therefore, impressions from any previous behavior should not influence the ratings made for any subsequent interval, except insofar as the con- text of a preceding interval must be considered for adequate interpre— tation of a unit of behavior (primarily for verbalization, inferred motivation, or reinforcement codes). Rating is facilitated if the videotape unit has an automatic signal tone attachment for recording purposes. Such an attachment may be used to provide an audio signal at the designated twenty-second intervals. Several viewings of each tape are necessary for completing the rat- ings. It is suggested that one rater be used to rate an entire tape. This is important for gaining the most accurate and complete information with maximum efficiency in the complex situation presented. Forms The instrument consists of two observational rating sections and a general information section. The first section of the rating procedure is used for the dyadic interaction situation in which the parent teaches the child a designated task. The second section is used for a post-task test, in which the examiner interacts with the child with the parent present. No observational rating is done for the portion of the session during which the examiner instructs the parent in the task procedures. General Information General information is recorded by the observer concerning the total session. Information included in this section is as follows: 214 (a) length of demonstration section of session, (b) length of teaching section of session, (c) length of testing section of session, (d) length of orientation, (e) use of orienting statement, (f) decision for termi- nation, and (g) reason for termination. In addition to identifying in- formation about the subjects and the session, a place is also provided for additional information or notes which the observer may wish to record. An example of the general information form is given in Figure I-3. One copy of this form is used with each set of rating forms for a session (Sections 1 and 2). Section 1 Section 1 is used for rating the parent and the child during the time that the parent teaches the child the designated task. Each of the twenty-second frames includes the following information for the parent: (a) verbalization, (b) verbal fantasy, (c) voice tone, (d) Verbal specificity, (e) time orientation, (f) task orientation, (g) l’lature of interference, (h) feedback, (i) reward, (j) response, (k) initiation, (1) response object, (m) reinforcement, (n) affective tone, (0) level of involvement, (p) anxiety, (q) physical behavior, (r) con- <3ept specification, and (5) teaching method. Information included for the child in each twenty-second frame is as follows: (a) verbalization, (b) verbal fantasy, (c) voice tone, (d) verbal specificity, (e) time cDrientation, (f) task orientation, (g) nature of interference, (h) feed- back, (i) reward, (j) response, (k) initiation, (1) response object, (m) reinforcement, (n) affective tone, (0) level of involvement, (p) anxiety, ((1) physical behavior, (r) dependency, and (s) inferred motivation. 215 PARENT-CHILD INTERACTION RATING PROCEDURE Child Task Child ID Date Chi 1d Peer Code Race Examiner Sex Observer SES Child's Age Years Months Days Child's Previous Preschool Experience Months Adult Re lation to Child General Information Length of demonstration section Length of teaching section Length of testing section Length of orientation Use of orienting statement C] Yes D No DeCision for termination [:3 Parent D Child Re ason for termination Examiner Notes I"IGURE I-3. General information form: Parent-Child Interaction Rating p rOctedure . 216 An example of the rating frame used for recording an observation segment for the parent is shown in Figure I-4; the format for this frame is shown in Figure I-5. An example of the rating frame used for record- ing an observation segment for the child is shown in Figure I-6, and the format for this frame is shown in Figure I-7. An example of the rating form used for recording the observations in Section 1 (four minutes for both parent and child) is shown in Figure I-8. Section 2 Section 2 is used for rating the parent and the child during the time that the examiner tests the child on the task following the parent- child teaching session. Each twenty-second frame for the parent in- cludes the following information: (a) verbalization, (b) verbal fan- tasy, (c) voice tone, (d) task orientation, (e) verbal receiver, (f) anxiety, (g) physical behavior, (h) level of involvement, (i) cue, (j) cue directiveness, (k) cue type, (1) defensiveness, (m) defensiveness target, and (n) defensiveness object. Information included for the child in each twenty-second frame is as follows: (a) verbalization, (b) verbal fantasy, (c) voice tone, (d) task orientation, (e) verbal receiver, (f) anxiety, (g) physical behavior, (h) level of involvement, (i) dependency, and (j) inferred motivation. An example of the rating frame used for recording an observation segment for the parent is shown in Figure 1‘9; the format for this frame is shown in Figure I-lO. An example of the rating frame used for recording an observation segment for the child is shown in Figure I-11, and the format for this frame is shown in Figure I-l2. An example of the rating form used for recording the observations in Section 2 (four minutes for both parent and child) is shown in Figure I-l3. 217 F NF NF + O - + O - G S S PA PR FU PA PR FU T NT NT 1 I v V FIGURE I-4. Rating frame for Section l--Parent observation: Parent-Child Interaction Rating Procedure. 218 .muspmooum mcflumm cofluomumucH UHHnunucoumm "cofium>uwmno ucmummlua cofluomm Mom umeuom .mIH mmeHm N: 0600 Ht 0600 mt mpoo mocmuwmumucH mocmummumucH conumz ocwnomme mt mpou ammocou mo musumz mo mudumz mt mpou ax mpou Ha moou coflumucmfiuo cofiumucmfluo vogue: mcflcomwe an mpou ummocou xmme xmme ; a--1:: .1 mt wpou at mooo coflumucmfluo cofiumucwwuo moou uofl>Mme mpou mEHB mfiwe Hmoamssa sumfixca ms mcoo as mvoo Spfioflmaummm sufioamaommm Hmnum> Hmnum> mpoo ucwEm>ao>cH mpou once 0 w>w m>wuom u H a Lee as «coo He «coo (I) (lit wcoe mofio> mace woflo> mt opou we mpou N: @600 we mpou uumnbo cofiumHuHcH oum3mm mu moou a: 0000 ucwEwUHOMCHmm mmcommom \Nt mpou \Nu mpoo ammucmm ammucmm mmcommmm xomnpmmm Hmnum> Hmnum> an opou a: ovou COMamwpmw mewpou unwEwUMOMCAmm uomflno .u .u. H 0 mm «t opoo at 0000 . wmcommmm \Ht mpoo \Ht @000 Hmnu0> Hmnum> mmcommmm xomnpmmmlfi 219 NF NF + O - + 0 - S G S PA PR FU PA PR FU NT T NT ¢ +— FIGURE I-6. Rating frame for Section l--Child observation: Parent-Child Interaction Rating Procedure. 220 .musomooum mcflumm coHuomumucH oaflnouucmumm “cofium>ummno cannonufl cofluomm new umsuom .kuH mmoon me mooo He moou Nt wvaU UUGOHOMHOUCH GUCMHGMHMUCH coHum>fluoz pmuummcH mo musumz mo musumz mocmocmmmo N: 0000 at ovou at 0000 coflumucmfiuo newumucmwuo coHum>wuoz pwuummcH xmme xmme Na mpou at woou cofiumucmflwo coHumucmwuo mpou u0w>m£wm wpou mafia mEHB Hmowmhnm >umfixc¢ mt mpou at mpou >ufioflufiommm aufloflwflommm mpou acmEm>H0>cH mpou mace Hmbum> Hmnum> mo Hm>mq 0>Huommu< at mpoo 2 «.600 mcoe mowo> mace mowo> mt mpoo uownno coflumfiuHcH pumzmm ucwEwouOMCAwm wmcommmm \Nu @000 \N* onou Mt mpou Mt mpoo mmcommmm xomnpwmm mmucmm mmucmm Hobum> ambum> ovou an mpou a: mpoo at mpou Ht coflumfiuficH cumzmm uowflno mt mpou Ht mpou ucmEmou0wcwmm mmcommm \Ht opoo \H¢ mpou a mm m um m mmcommmm xomnpmmm H b > H b > 221 PARENT-CHILD INTERACTION RATING PROCEDL'RE--Section 1 Adult Child Tuk Child 1!) Relation to Child Date Child Peer Cod. Baa-iner Observer .. FIGURE I-8. Rating form--Section l: Parent-Child Interaction Rating Procedure. 222 L M H F NF NF I C + o - - NV NT T NT I C C E G G NV FIGURE I-9. Rating frame for Section 2--Parent observation: Parent-Child Interaction Rating Procedure. 223 .musvmooum mcflumm cofluomumucH paflconucmumm "cofium>ummbo ucwummuum cofiuomm now umeuom .OHIH mmson mu mooo we wpou Ht mpou poombo ma mpou um>wwomm um>flmomm . m> mmmcm>flmcmwma Na mpou m B moo Na Hmnuw> Habum> mmMMWMMm mpoo mt mpou m» moon @90 ma 0000 at mcou ummwme mmwcm>fluowuflo coflumucmfluo coflumucmfluo mmmcm>fimcmmwo woo xmme xmme an oooo nomflno He mwoo me 6606 Ht mooo mmmcm>amcmmoo wm>e moo wcoe moHo> wcoe moflo> . Ht opou a: mmmcm>am ovou an mooo ucwmmo Ht @600 mso ummpme mmmcw>fluomuflo mt mpou at mpou mmmcm>flmcomoo moo >mmucwm wmmucwm ambum> annuw> o o w o p O p O wpou ucoEm>Ho>cH uofl>mcmm >umaxc¢ mt wpou Ha mnou mo Hw>mq Hm0flm>£m . ambuw> Hmnum> (e 224 L M F NF F NF H + O - + 0 - T NT T NT P E G P E G FIGURE I-ll. Rating frame for Section 2--Child observation: Parent-Child Interaction Rating Procedure. 225 mcflumm cofluomnmucH pHH£UIucmumm "coaum>umwbo paflbOIIN cofluomm How . OHSUQUOHQ umEhom .NHIH mmDOHh N# ovou an wpoo me $000 um>Hmomm um>flmowm coaum>fluoz Hubum> amnum> pmnummcH wpoo xocmpcmmmo Nu mpoo an mpoo at mpou cofiumucmfluo coflumucmauo coflum>fluoz xmme xmma UmuuwmcH mu meow Ht meow wcoe muflo> mcoe woflo> mu wpou at mpou xmmucmm xmmucmm Hmnum> Hmnum> wpoo mpou mpou ucmEm>Ho>cH uofl>mbmm xuwaxcc mo Hw>wq Hm0flm>cm . NamRO HamRO Hmnuw> Hmnuw> PARENT-CHILD INTERACTION RATING PROCEDURE--Section 2 Adult Ta Child a): Child ID Relation to Child Date Child Peer Code Examiner Observer FIGURE I-13. Rating form--Section 2: Parent-Child Interaction Rating Procedure. 227 RecordingfiObservations Both sections of the rating form must contain a code for each per- son for each interval in each code position. If nothing is applicable, an "X" is coded in that position or a line drawn through it. Affective tone, level of involvement, anxiety, dependency, re- sponse code #1, and initiation code #1 must each be a code other than "X" for each interval. All other code positions may contain either a specific alpha-numeric code or an "X" ("not applicable"), depending on the particular situation occurring during that interval. The only ex- ception to this rule is the rare case in which the person being observed leaves the scene (is out of camera range). In such cases, "X" is coded for each section of the frame. Coding of each category is done by writing in the appropriate code (for verbalization, nature of interference, feedback, reward, response, initiation, level of involvement, physical behavior, concept specifica- tion, dependency, inferred motivation, and defensiveness object) or by circling the appropriate code letters (for verbal fantasy, voice tone, verbal specificity, time orientation, task orientation, response object, reinforcement, affective tone, anxiety, teaching method, verbal re- ceiver, cue, cue directiveness, cue type, defensiveness, and defensive- ness target). For those series of categories for which the applica- bility depends on the coding of another category (fantasy, voice tone, specificity, time orientation, task orientation, nature of interfer- ence, verbal receiver, response object, reinforcement, teaching method, cue directiveness, cue type, defensiveness target, and defensiveness object), a line should be drawn through those code positions to indicate their lack of applicability, as shown in Figure I-l4. F NF NF P + + - + 0 - O .. G S S M H PA FU PA PR FU D E ’r T \t NT 1 D E I Section l--Parent observation L M H F NF F NF D I C \ \/ D \ + 0 - + o - A V V T NT T NT D I C D C C E G V NV Section 2--Parent observation FIGURE I-l4. Indication of nonapplicable P-CIRP categories. 229 Reliability Interobserver reliability is established by two independent ob- servers simultaneously recording the behaviors of the same person in the same intervals on their respective recording forms. Intraobserver reliability is established by a single observer rerating a previously observed tape. Two methods of computing reliability are used, one based on total blanks and the other based on total recorded positions. Each type of reliability should be computed for both parent and child observations for each section of the instrument and also for each separate scale. Minimum suggested reliability indices are given in Table I-l. Points for figuring section reliability are assigned as shown in Figure I-lS. Procedures for computation of interobserver reliability and intraobserver reliability are the same using either method. Total Blanks Count and evaluate the total number of possible codes, regardless of whether anything was recorded within that area for that time interval or not. This method credits the observers with agreements for those in- stances on which there is agreement that no recordable behavior occurred-- i.e., both observations contain an "X" for that category of that interval. Formulas used for figuring reliability by this method are as follows: (a) Section l--Parent observation: Agreements (Number of points) . Number of frames x 34 ' % reliability = (b) Section l--Chi1d observation: Agreements (Number of points) Number of frames x 33 % reliability = o I 230 TABLE I-l Minimum Suggested Observer Reliability Indices for Parent-Child Interaction Rating Procedure Type of Reliability Method Interobserver Intraobserver Entire Section Total Blanks .85 .90 Total Recorded Positions .65 .75 Individual Scales Total Blanks .80 .85 Total Recorded Positions .60 .70 231 CD @@ C9 C9 (9 Section l--Chi1d observation CD @ <9 oooo <9 @ ®@@@ 9 @ oooo @ @ oeoo (D '@ (9 o o o o <9 -@ o o o o t observation -Paren Section 2--Chi1d observation t observation Section 2--Paren liability. FIGURE I-15. Assignment of points for P-CIRP observer re 232 (c) Section 2--Parent observation: - - - _ Agreements (Number of points) ; % reliability — Number of frames x 21 (d) Section 2--Child observation: Agreements (Number of points) . % IEIlablllty = Number of frames x 16 Total Recorded Positions Count and evaluate only those positions in which one or both ob- servations contain a code other than "X." The formula used for figuring reliability by this method is as follows: Agreements (Number of points) Agreements plus disagreements (Number of points possible for positions in which either observation contains any code) % reliability = Codes Description of information included in the general information section is given in Table I-2. The categories for each code used in Sections 1 and 2 of the rating procedure and descriptions of them are given in Table I-3. 233 TABLE I -2 Category Descriptions for General Information Section of Parent-Child Interaction Rating Procedure Category Description Length of demonstration section Length of teaching section Length of testing section Length of orientation Use of orienting statement Decision for termination Reason for termination Number of minutes in first section of session, during which the examiner teaches the task to the parent Number of minutes in second section of session, during which the parent teaches the task to the child Number of minutes in third section of session, during which the examiner tests the child on the task in the presence of the parent Number of minutes in the teaching section (beginning at the time the child enters the room and rating of Section 2 begins) until the parent calls for the first block placement by the child Whether or not the parent gives a statement to the child explaining the general nature of the situation and the task to be learned Who makes the decision to terminate the teaching section of the session (parent, child, or examiner) Reason given for ending the teaching section of the session (e.g., time limit, child learns task, etc.) 234 TABLE I-3 Category Descriptions for Sections 1 and 2 of Parent-Child Interaction Rating Procedure Category Description Verbalization SL Shows solidarity; raises other's status; gives help or reward TR Releases tension; jokes; laughs; shows satisfaction AG Agrees; shows passive acceptance; understands; concurs; complies SU Gives suggestions or direction, implying autonomy for others OP Gives opinion, evaluation, or analysis; expresses feeling or wish OR Gives orientation or information; repeats; clarifies; confirms AR Asks for orientation, information, repetition, confirmation AP Asks for opinion, evaluation, analysis, expression of feelings As Asks for suggestion, direction, possible ways of action DS Disagrees; shows passive rejection or formality; withholds help ST Shows tension; asks for help; withdraws "out of field" AN Shows antagonism; deflates other's status; defends or asserts self MM Mumbles X No verbalization Verbal Fantasy F Fantasy verbalization NF Nonfantasy verbalization 235 TABLE I-3 (continued) Category Description Voice Tone + Positive affect conveyed by voice tone 0 Neutral voice tone; no affect conveyed - Negative affect conveyed by voice tone Verbal Specificity G Global verbalization; no specific labels S Specific verbalization; includes labels and/or explicit directions Time Orientation PA Past reference; refers to anything which occurred in the past (even in present situation) PR Present reference; reference to immediate situation or task FU Future reference; refers to anything to occur in the future (even in present situation) Task Orientation T Task-oriented verbalization; refers to performance of task presented in situation NT Non-task-oriented verbalization; refers to situation or experience other than the task presented or to specific avoidance of that task Nature of Interference 0 Outside interference; not initiated by either parent or child Pm Parent-initiated diversion involving non-task use of materials 236 TABLE I-3 (continued) Category Description Pe Parent-initiated diversion which focuses on immediate environment Pn Parent-initiated diversion which focuses on nonsituational factors Cm Child-initiated diversion which involves non-task use of materials Ce Child-initiated diversion which focuses on immediate environment Cn Child-initiated diversion which focuses on nonsituational factors X No interference; task-related verbalization Verbal Receiver C Child as intended receiver of verbalization P Parent as intended receiver of verbalization E Examiner as intended receiver of verbalization G Group-directed verbalization (undifferentiated receiver) Feedback 0 Overt acceptance or rejection of immediately preceding communication of other person C Covert acceptance or rejection of immediately preceding communication of other person X No immediately preceding communication by other person or no response to such a communication (used with Response codes D, I, N, and O) Reward P Praise (used only with Feedback codes 0 and C) C Criticism (used only with Feedback codes 0 and C) 237 TABLE I-3 (continued) Category Description x No praise or criticism given (implied with Feedback code X) Response D Failure to respond because the other did not allow it I Ignoral; failure to respond to communication by other person, even though there is opportunity to do so N No feedback given because there is no immediately preceding communication by the other person A Acceptance of communication of other person R Rejection of communication of other person 0 Ongoing interaction which is continued from previous interval Initiation u Unqualified power assertion; attempt to make use of juris- diction by such means as physical punishment, isolation, or insistence upon a specified behavior q Qualified power assertion; attempt to make use of jurisdic- tion by punishment, isolation, direction, or insistence, but with the addition of a reason or a cushion or both to this exertion of authority r Bargain, appeal, or suggestion for behavior with offer of a reward (which may be tangible or intangible) for com- pliance; bribe c Suggestion of action in which the other person is given a choice of compliance n Simple command or request for response which does not assert power or offer a choice or reward for compliance; neutral initiation (used with 0 if continuation of communication from previous interval) f Failure to continue interaction by initiation of new idea or direction (implied with Response codes D and with 0 if on- going to next interval; cannot be used with Response code N) 238 TABLE I-3 (continued) Category Description Response Object P Acceptance or rejection of the person as an individual B Acceptance or rejection of the behavior or performance of the other person X No response given (used with Response codes D, I, N, and O) Reinforcement + Correct or appropriate response to behavior - Incorrect or inappropriate response to behavior X No response given (used with Response codes D, I, N, and O) Affective Tone + Positive social-emotional tone; conveys affection for the other person 0 Neutral tone; no indicated affect - Negative social-emotional tone; conveys lack of affection or annoyance with the other person Level of Involvement 1 Extremely involved in situation or task 2 Involved with situation or task at an average or moderate level 3 Passively involved with situation or task; "serving time" in the situation Anxiety L Low anxiety level; no apparent anxiety shown M Moderate anxiety level; some indication of anxiety expressed 239 TABLE I-3 (continued) Category Description High anxiety level; obvious tension or discomfort in situation Physical Behavior NP SP DF NM Negative physical behavior; behavior which is not socially acceptable and which involves physical contact between the subject and some other person or object (e.g., hitting, pushing) Social physical behavior; behavior which is socially accept- able and/or conveys affection and which involves physical contact between the subject and some other person; inten- tional physical contact conveying affection for an object (e.g., patting, holding hands) Approach gesture; behavior which is socially acceptable or positive in connotation and which does not involve actual physical contact between the subject and another person or object (e.g., beckoning) Defensive posture; behavior which is not socially acceptable or is negative in connotation and which does not involve actual physical contact between the subject and another per- son or object (e.g., shaking fist) Neutral motion; physical behavior which does not convey either positive or negative connotation but which is an in- tentional communicative act (e.g., head nod) No nonverbal (physical) behavior Concept Specification NC Introduction of new concept in teaching (task-oriented) behavior; may be verbal and/or nonverbal Introduction of a new approach to teaching the same concept; may be verbal and/or nonverbal but must apply to teaching (task-oriented) behavior Focus on a different aspect of the same concept; may be verbal and/or nonverbal but must apply to teaching (task- oriented) behavior 240 TABLE I-3 (continued) Category Description ON Ongoing method for teaching the same concept; no change in focus GO General orientation to learning task or concept(s) to be presented GS General summary of learning task or concept(s) presented X No task-oriented (teaching) behavior present (used only with Task Orientation code NT) Teaching Method D Demonstration only; nonverbal approach used E Explanation only; approach used is verbal only I Illustration of concept; both verbal and nonverbal teaching methods used X No task-oriented (teaching behavior or_task-oriented (teach- ing) behavior with ON by parent because child is doing task and parent does nothing during interval Dependency 1 High level of psychological dependency shown 2 Some psychological dependency shown 3 No psychological dependency shown Inferred Motivation NO Innovativeness ND Independence AG Aggression IM Imitation AT Attention-seeking 241 TABLE I-3 (continued) Category Description BO Boasting FE Fear X No apparent inferred motivation Cue C Cuing occurs by parent to child concerning performance of task X No cuing by parent to child concerning performance of task Cue Directiveness D Direct cuing to child by parent I Indirect cuing to child by parent X No cuing occurs (covered by Cue code X) Cue Type V Verbal cue given (may be used alone or in combination with NV) NV Nonverbal cue given (may be used alone or in combination with V) X No cuing occurs (covered by Cue code X) Defensiveness D Defensive statement made by parent concerning situation X No defensive behavior shown by parent Defensiveness Target C Child as target of parent's defensive statement E Examiner as target of parent's defensive statement 242 TABLE I-3 (continued) Category Description X No defensive statement given (covered by Defensiveness code X) Defensiveness Object Ca Child designated as being at fault, with blame directed to his abilities CO Child designated as being at fault, with blame directed to reasons other than his abilities Pa Parent designated as being at fault, with blame directed to his abilities (e.g., teaching effectiveness) PO Parent designated as being at fault, with blame directed to reasons other than his abilities Se Situation designated as being at fault, with emphasis on environment or immediate circumstances St Situation designated as being at fault, with emphasis on the task per se X No defensive statement given (covered by Defensiveness code X) APPENDIX II CELL MEANS OF VARIABLES FOR ALL DEMOGRAPHIC GROUPS APPENDIX II-A CELL MEANS OF COMMUNICATION PROCESS VARIABLES Cell means and standard deviations are given for those variables included within the dimensions of communication process used in the analyses. To facilitate interpretation, raw proportions are given for those variables which are qualitative in nature rather than the logit transformations of the dichotomous and polychotomous ratios which were used in the statistical analyses. Dimensions of communication process described are as follows: initiation (Table II-l), response (Table II-2), response object (Table II-3), reinforcement (Table II-4), feedback (Table II-S), and reward (Table II-6). 243 244 TABLE II-l Demographic Group Means for Initiation Categories on the Parent-Child Interaction Rating Procedure Mothers Fathers Category Mean Standard Mean Standard Proportion Deviation Proportion Deviation Lower Socioeconomic Group Black Males Unqualified power use 0.037001 0.031695 0.009868 0.019737 Qualified power use 0.025496 0.021692 0.000000 0.000000 Bribe 0.002941 0.005882 0.001250 0.002500 Choice 0.002941 0.005882 0.000000 0.000000 Neutral initiation 0.720661 0.089208 0.714695 0.175341 Failure to continue 0.192120 0.049442 0.250878 0.188170 Middle Socioeconomic Group Black Males Unqualified power use 0.018519 0.037037 0.022032 0.037925 Qualified power use 0.054226 0.066457 0.002976 0.005952 Bribe 0.012731 0.025463 0.014423 0.028846 Choice 0.004360 0.008721 0.000000 0.000000 Neutral initiation 0.640818 0.125708 0.675672 0.080763 Failure to continue 0.220882 0.025837 0.276990 0.084457 Lower Socioeconomic Group Black Females Unqualified power use 0.043103 0.065275 0.018385 0.022670 Qualified power use 0.004310 0.008621 0.019799 0.032314 Bribe 0.013744 0.017941 0.000000 0.000000 Choice 0.004310 0.008621 0.007143 0.014286 Neutral initiation 0.790743 0.066833 0.578092 0.056971 Failure to continue 0.102231 0.056791 0.370768 0.060021 Middle Socioeconomic Group Black Females Unqualified power use 0.000000 0.000000 0.016360 0.018928 Qualified power use 0.009259 0.018519 0.069311 0.052818 Bribe 0.000000 0.000000 0.015625 0.031250 Choice 0.000000 0.000000 0.058961 0.106830 Neutral initiation 0.733125 0.127911 0.510368 0.201925 Failure to continue 0.190949 0.131051 0.194916 0.064534 245 TABLE II-l (continued) Mothers Fathers Category Mean Standard Mean Standard Proportion Deviation Proportion Deviation Lower Socioeconomic Group White Males Unqualified power use 0.020527 0.024099 0.000000 0.000000 Qualified power use 0.004683 0.003498 0-004990 0.005917 Bribe 0.000000 0.000000 0.006829 0.004906 Choice 0.011585 0.019914 0.000000 0.000000 Neutral initiation 0.752946 0.114736 0.629090 0.132840 Failure to continue 0.185815 0.109071 0.346985 0.135135 Middle Socioeconomic Group White Males Unqualified power use 0.002358 0.004717 0.005693 0.007110 Qualified power use 0.034755 0.064963 0.052024 0.063038 Bribe 0.002358 0.004717 0.003459 0.004637 Choice 0.001736 0.003472 0.012176 0.010078 Neutral initiation 0.709998 0.197075 0.693418 0.183430 Failure to continue 0.184008 0.111960 0.186819 0.092098 Lower Socioeconomic Group White Females Unqualified power use 0.028589 0.021083 0.017183 0.021190 Qualified power use 0.003178 0.006356 0.008147 0.007210 Bribe 0.006737 0.008281 0.000000 0.000000 Choice 0.001059 0.002119 0.008333 0.016667 Neutral initiation 0.710357 0.123249 0.654485 0.174934 Failure to continue 0.229134 0.131003 0.287259 0.169118 Middle Socioeconomic Group White Females Unqualified power use 0.011241 0.017856 0.036885 0.073770 Qualified power use 0.037671 0.043392 0.041165 0.036027 Bribe 0.013021 0.015625 0.004098 0.008197 Choice 0.000000 0.000000 0.000000 0.000000 Neutral initiation 0.769434 0.139140 0.659579 0.187397 Failure to continue 0.114218 0.069906 0.218550 0.067532 Note.--N = 4 for each cell. Each proportion indicates prOportion of observations designated by that category. Since not every observation was characterized by one of these behaviors, totals do not necessarily equal 1.00. 246 TABLE II-2 Demographic Group Means for Response Categories on the Parent-Child Interaction Rating Procedure Mothers Fathers Category Mean Standard Mean Standard Proportion Deviation Proportion Deviation Lower Socioeconomic Group Black Males Acceptance 0.519813 0.087754 0.590942 0.144175 Rejection 0.393715 0.049334 0.306502 0.188001 Ignoral 0.018840 0.024946 0.000000 0.000000 No response allowed 0.000000 0.000000 0.000000 0.000000 No preceding message 0.013731 0.012773 0.032538 0.015788 Ongoing behavior 0.035061 0.008802 0.046710 0.043113 Middle Socioeconomic Group Black Males Acceptance 0.504648 0.110036 0.637866 0.149771 Rejection 0.285574 0.127771 0.272654 0.147855 Ignoral 0.013515 0.009714 0.008940 0.010382 No response allowed 0.000000 0.000000 0.000000 0.000000 No preceding message 0.016458 0.012597 0.014226 0.011809 Ongoing behavior 0.131341 0.028160 0.058408 0.052336 Lower Socioeconomic Group Black Females Acceptance 0.555412 0.057217 0.647353 0.117468 Rejection 0.356618 0.049394 0.197521 0.094909 Ignoral 0.004717 0.009434 0.006757 0.013514 No response allowed 0.000000 0.000000 0.000000 0.000000 No preceding message 0.018579 0.011072 0.056046 0.080296 Ongoing behavior 0.031737 0.024998 0.086509 0.035358 Middle Socioeconomic Group Black Females Acceptance 0.688403 0.036231 0.404845 0.192183 Rejection 0.160694 0.143353 0.173832 0.123526 Ignoral 0.000000 0.000000 0.016396 0.012789 No response allowed 0.000000 0.000000 0.000000 0.000000 No preceding message 0.026065 0.014323 0.105019 0.100298 Ongoing behavior 0.058171 0.068197 0.165449 0.139961 247 TABLE II-2 (continued) Mothers Fathers Category Mean Standard Mean Standard Proportion Deviation Proportion Deviation Lower Socieoconomic Group White Males Acceptance 0.533859 0.117116 0.555505 0.056999 Rejection 0.354179 0.039378 0.340658 0.040515 Ignoral 0.036990 0.073980 0.022635 0.026421 No response allowed 0.000000 0.000000 0.000000 0.000000 No preceding message 0.008614 0.007332 0.017893 0.018753 Ongoing behavior 0.041914 0.049171 0.051202 0.017181 Middle Socioeconomic Group White Males Acceptance 0.570431 0.038024 0.610497 0.170971 Rejection 0.264063 0.077689 0.192131 0.041689 Ignoral 0.010357 0.016414 0.005040 0.010081 No response allowed 0.000000 0.000000 0.000000 0.000000 No preceding message 0.026131 0.024167 0.056248 0.059925 Ongoing behavior 0.059516 0.053730 0.089673 0.049348 Lower Socioeconomic Group White Females Acceptance 0.573429 0.134004 0.580223 0.093489 Rejection 0.358656 0.140931 0.322891 0.137488 Ignoral 0.000000 0.000000 0.018951 0.022749 No response allowed 0.001059 0.002119 0.000000 0.000000 No preceding message 0.016835 0.011343 0.009163 0.005834 Ongoing behavior 0.029076 0.023594 0.044179 0.024587 Middle Socioeconomic Group White Females Acceptance 0.583419 0.107631 0.524627 0.216674 Rejection 0.271774 0.090723 0.253360 0.088340 Ignoral 0.008116 0.011793 0.027349 0.042847 No response allowed 0.003125 0.006250 0.000000 0.000000 No preceding message 0.027254 0.008045 0.068616 0.061517 Ongoing behavior 0.051897 0.066693 0.086326 0.024932 Note.--N = 4 for each cell. Each proportion indicates proportion of observations designated by that category. Since not every observation was characterized by one of these behaviors, totals do not necessarily equal 1.00. 248 TABLE II-3 Demographic Group Means for Response Object on the Parent-Child Interaction Rating Procedure Mothers Fathers Demographic Group Mean Standard Mean Standard Proportion Deviation Proportion Deviation Lower socioeconomic group black males 0.030869 0.035650 0.000000 0.000000 Middle socioeconomic group black males 0.009804 0.019608 0.014280 0.022997 Lower socioeconomic group black females 0.010101 0.020202 0.004658 0.009317 Middle socioeconomic group black females 0.000000 0.000000 0.014706 0.029412 Lower socioeconomic group white males 0.015596 0.018419 0.009572 0.006689 Middle socioeconomic group white males 0.008979 0.011234 0.007003 0.006339 Lower socioeconomic group white females 0.015713 0.013969 0.021015 0.024374 Middle socioeconomic group white females 0.000000 0.000000 0.021429 0.042857 Note.--N = 4 for each cell. Each proportion indicates proportion of responses which were di- rected at the individual worth of the child rather than his behavior or performance. 249 TABLE II -4 Demographic Group Means for Reinforcement on the Parent-Child Interaction Rating Procedure Mothers Fathers Demographic Group Mean Standard Mean Standard Proportion Deviation Proportion Deviation Lower socioeconomic group black males 0.918128 0.062780 0.952777 0.087388 Middle socioeconomic group black males 0.956208 0.054448 0.936675 0.099792 Lower socioeconomic group black females 0.956245 0.050524 0.915667 0.104486 Middle socioeconomic group black females 0.984369 0.020622 0.978448 0.043103 Lower socioeconomic group white males 0.972449 0.018878 0.980865 0.024292 Middle socioeconomic group white females 0.950001 0.055897 0.904265 0.145810 Lower socioeconomic group white females 0.922904 0.060808 0.949026 0.033037 Middle socioeconomic group white females 0.968434 0.053509 0.936842 0.049067 Note.--N = 4 for each cell. Each proportion indicates proportion of responses which employed correct or appropriate reinforcement of child's behavior or performance. 250 TABLE II-S Demographic Group Means for Feedback on the Parent-Child Interaction Rating Procedure Mothers Fathers Demographic Group Mean Standard Mean Standard Proportion Deviation Proportion Deviation Lower socioeconomic group black males 0.656186 0.132934 0.535568 0.190497 Middle socioeconomic group black males 0.391514 0.303594 0.659140 0.126213 Lower socioeconomic group black females 0.786508 0.116386 0.613627 0.433601 Middle socioeconomic group black females 0.428718 0.281128 0.117578 0.142561 Lower socioeconomic group white males 0.762984 0.135811 0.571880 0.229166 Middle socioeconomic group white males 0.387574 0.442842 0.443027 0.387311 Lower socioeconomic group white females 0.632936 0.168706 0.566169 0.260664 Middle socioeconomic group white females 0.460638 0.461617 0.272016 0.267488 Note.--N = 4 for each cell. Each proportion indicates proportion of responses which gave covert rather than overt feedback. 251 TABLE II-6 Demographic Group Means for Reward Categories on the Parent-Child Interaction Rating Procedure Mothers Fathers Category Mean Standard Mean Standard Proportion Deviation PrOportion Deviation Lower Socioeconomic Group Black Males Praise 0.025391 0.050781 0.019868 0.022943 Criticism 0.019017 0.024985 0.009868 0.019737 Middle Socioeconomic Group Black Males Praise 0.090280 0.045427 0.042299 0.036882 Criticism 0.001157 0.002315 0.003099 0.006198 Lower Socioeconomic Group Black Females Praise 0.073956 0.085786 0.051968 0.045498 Criticism 0.008621 0.017241 0.002841 0.005682 Middle Socioeconomic Group Black Females Praise 0.046875 0.080463 0.106154 0.019055 Criticism 0.000000 0.000000 0.014223 0.016582 Lower Socioeconomic Group White Males Praise 0.011572 0.017034 0.025912 0.024114 Criticism 0.013782 0.019439 0.007353 0.014706 Middle Socioeconomic Group White Males Praise 0.057238 0.051390 0.079973 0.060286 Criticism 0.003247 0.006494 0.004464 0.008929 Lower Socioeconomic Group White Females Praise 0.018559 0.034352 0.037818 0.044832 Criticism 0.005000 0.010000 0.002525 0.005051 252 TABLE II-6 (continued) Mothers Fathers t Ca egory Mean Standard Mean Standard Proportion Deviation Proportion Deviation Middle Socioeconomic Group White Females Praise 0.028078 0.013725 0.085902 0.093593 Criticism 0.000000 0.000000 0.014344 0.028689 Note.--N = 4 for each cell. Each proportion indicates proportion of time during which the designated reward (either praise or criticism) was given to the child. APPENDIX II-B CELL MEANS OF VERBAL COMMUNICATION VARIABLES Cell means and standard deviations are given for those variables included within the dimensions of verbal communication used in the analyses. To facilitate interpretation, raw proportions are given for those variables which are qualitative in nature rather than the lOgit transformations of the dichotomous and polychotomous ratios which were used in the statistical analyses. Dimensions of verbal communi- cation described are as follows: verbalization (Table II-7), fantasy verbalization (Table II-8), voice tone (Table II-9), verbal specificity (Table II-lO), time orientation (Table II-ll), task orientation (Table II-12), initiator of non-task verbalization (Table II-l3), and task remoteness of non-task verbalization (Table II-14). 253 254 TABLE II-7 Demographic Group Means for Verbalization Categories on the Parent-Child Interaction Rating Procedure Mothers Fathers Category Mean Standard Mean Standard Proportion Deviation Proportion Deviation Lower Socioeconomic Group Black Males Communication 0.484163 0.136686 0.386526 0.081668 Evaluation 0.014988 0.013467 0.043860 0.050771 Control 0.248187 0.102901 0.267169 0.107036 Decision 0.142926 0.063440 0.133545 0.092096 Tension management 0.008050 0.012489 0.022178 0.023749 Integration 0.054019 0.044351 0.019868 0.022943 Mumbling 0.000000 0.000000 0.007829 0.012548 Middle Socioeconomic Group Black Males Communication 0.486541 0.053557 0.434396 0.082587 Evaluation 0.072677 0.047638 0.039420 0.046097 Control 0.237251 0.068988 0.271316 0.049140 Decision 0.069187 0.027684 0.129365 0.126029 Tension management 0.014107 0.013988 0.008817 0.008518 Integration 0.067911 0.026000 0.033605 0.032072 Mumbling 0.001157 0.002315 0.002404 0.004808 Lower Socioeconomic Group Black Females Communication 0.513834 0.143990 0.360312 0.161318 Evaluation 0.015290 0.014609 0.014024 0.016215 Control 0.268558 0.075092 0.298672 0.108281 Decision 0.051833 0.023502 0.173240 0.136487 Tension management 0.008621 0.009954 0.011051 0.011617 Integration 0.066447 0.071826 0.049094 0.046626 Mumbling 0.000000 0.000000 0.004992 0.006750 Middle Socioeconomic Group Black Females Communication 0.526921 0.097377 0.459609 0.081908 Evaluation 0.046991 0.041016 0.126460 0.047106 Control 0.192824 0.094718 0.180843 0.053383 Decision 0.100694 0.116410 0.037764 0.045051 Tension management 0.005208 0.010417 0.018273 0.024470 25 5 TABLE II-7 (continued) Mothers Fathers Category Mean Standard Mean Standard Proportion Deviation Proportion Deviation Integration 0.065046 0.091570 0.082279 0.013331 Mumbling 0.000000 0.000000 0.000000 0.000000 Lower Socioeconomic Group White Males Communication 0.494571 0.124439 0.474008 0.151835 Evaluation 0.032599 0-024794 0.038127 0.030435 Control 0.274968 0.076743 0.207213 0.041195 Decision 0.088824 0-036813 0.126422 0.099279 Tension management 0.011139 0.017102 0.009130 0.013800 Integration 0.018805 0.018335 0.031332 0.020566 Mumbling 0.000000 0-000000 0.007843 0.009092 Middle Socioeconomic Group White Males Communication 0.500900 0.193516 0.441018 0.076589 Evaluation 0.089939 0.037118 0.078333 0.053213 Control 0.225811 0.051776 0.236268 0.060364 Decision 0.080241 0.079738 0.106858 0.039440 Tension management 0.008699 0.013422 0.004464 0.008929 Integration 0.050574 0.039334 0.058922 0.049233 Mumbling 0.006340 0.008895 0.006048 0.012097 Lower Socioeconomic Group White Females Communication 0.568916 0.147011 0.483704 0.146895 Evaluation 0.052616 0.045395 0.043381 0.018450 Control 0.177834 0.086535 0.186469 0.109623 Decision 0.136063 0.080687 0.105904 0.077050 Tension management 0.008523 0.017045 0.000000 0.000000 Integration 0.020435 0.028298 0.027976 0.028088 Mumbling 0.000000 0.000000 0.007462 0.006992 Middle Socioeconomic Group White Females Communication 0.623943 0.059139 0.482071 0.140589 Evaluation 0.051127 0.045058 0.105878 0.140443 Control 0.210572 0-020535 0.185357 0.015510 Decision 0.031553 0.042768 0.066109 0.046520 256 TABLE II-7 (continued) Mothers Fathers Category Mean Standard Mean Standard Proportion Deviation Proportion Deviation Tension management 0.005208 0.010417 0.009259 0.018519 Integration 0.026648 0.015815 0.075451 0.067068 Mumbling 0.000000 0.000000 0.008906 0.010349 Note.--N = 4 for each cell. Each proportion indicates proportion of observations designated by that category. these behaviors, totals do not necessarily equal 1.00. Since not every observation was characterized by one of 257 TABLE II-8 Demographic Group Means for Fantasy Verbalization on the Parent-Child Interaction Rating Procedure Mothers Fathers Demographic Group Mean Standard Mean Standard Proportion Deviation Proportion Deviation Lower socioeconomic group black males 0.000000 0.000000 0.004902 0.009804 Middle socioeconomic group black males 0.137259 0.152424 0.000000 0.000000 Lower socioeconomic group black females 0.000000 0.000000 0.000000 0.000000 Middle socioeconomic group black females 0.009259 0.018519 0.000000 0.000000 Lower socioeconomic group white males 0.047680 0.095361 0.000000 0.000000 Middle socioeconomic group white males 0.001736 0.003472 0.017789 0.026691 Lower socioeconomic group white females 0.001059 0.002119 0.000000 0.000000 Middle socioeconomic group white females 0.009328 0.018657 0.006679 0.008788 Note.--N = 4 for each cell. Each proportion indicates proportion of verbalizations which were fantasy-oriented. 258 TABLE II-9 Demographic Group Means for Voice Tone on the Parent-Child Interaction Rating Procedure Mothers Fathers Demographic Group Standard Standard Mean . . Mean . . DeViation DeViation Lower socioeconomic group black males 1.918225 0.107104 1.977718 0.047645 Middle socioeconomic group black males 2.076836 0.060589 2.026447 0.078820 Lower socioeconomic group black females 2.021042 0.134671 2.000336 0.010957 Middle socioeconomic group black females 2.080440 0.122800 2.047383 0.023964 Lower socioeconomic group white males 2.017793 0.047892 1.986201 0.015941 Middle socioeconomic group white males 2.069415 0.076574 2.066464 0.097810 Lower socioeconomic group white females 2.018687 0.014995 2.008093 0.049568 Middle socioeconmic group white females 2.078286 0.136532 2.013468 0.145571 Note.--N = 4 for each cell. Means were based on a weighting of "l" for negative voice tone, "2" for neutral voice tone, and "3" for positive voice tone. 259 TABLE II-lO Demographic Group Means for Verbal Specificity on the Parent-Child Interaction Rating Procedure Mothers Fathers Demographic Group Mean Standard Mean Standard Proportion Deviation Proportion Deviation Lower socioeconomic group black males 0.431018 0.106542 0.453658 0.274178 Middle socioeconomic group black males 0.587168 0.236153 0.424465 0.165973 Lower socioeconomic group black females 0.597480 0.113321 0.345551 0.114309 Middle socioeconomic group black females 0.411839 0.120831 0.349601 0.055187 Lower socioeconomic group white males 0.483304 0.138111 0.276105 0.224223 Middle socioeconomic group white males 0.524668 0.179151 0.430636 0.123430 Lower socioeconomic group white females 0.393594 0.144312 0.384943 0.161369 Middle socioeconomic group white females 0.518734 0.222368 0.405814 0.197992 Note.--N = 4 for each cell. Each proportion indicates the proportion of verbalizations which were specific rather than global in nature. 260 TABLE II-ll Demographic Group Means for Time Orientation on the Parent-Child Interaction Rating Procedure Mothers Fathers Demographic Group Standard Standard Mean . . Mean . . DeViation DeViation Lower socioeconomic group black males 2.012422 0.024845 1.990910 0.032580 Middle socioeconomic group black males 2.031652 0.021586 1.997368 0.005263 Lower socioeconomic group black females 2.000000 0.000000 2.001592 0.003185 Middle socioeconomic group black females 2.004673 0.009346 2.052982 0.120756 Lower socioeconomic group white males 1.996105 0.008718 1.999181 0.023424 Middle socioeconomic group white males 2.011225 0.016264 2.008496 0.010199 Lower socioeconomic group white females 2.006820 0.022516 2.020819 0.035032 Middle socioeconomic group white females 2.002016 0.004032 2.012741 0.048749 Note.--N = 4 for each cell. Means were based on a weighting of "l" for past time orientation, "2" for present time orientation, and "3" for future time orientation. 261 TABLE II-12 Demographic Group Means for Task Orientation on the Parent-Child Interaction Rating Procedure Mothers Fathers Demographic Group Mean Standard Mean Standard Proportion Deviation Proportion Deviation Lower socioeconomic group black males 0.976688 0.025027 0.966882 0.040314 Middle socioeconomic group black males 0.913993 0.054421 0.933526 0.083261 Lower socioeconomic group black females 0.935438 0.096259 0.980233 0.016357 Middle socioeconomic group black females 0.992991 0.014019 0.941348 0.079917 Lower socioeconomic group white males 0.879965 0.111151 0.990715 0.006811 Middle socioeconomic group white males 0.954837 0.040056 0.957291 0.023817 Lower socioeconomic group white females 0.939827 0.049806 0.952926 0.044233 Middle socioeconomic group white females 0.916400 0.063485 0.903709 0.090339 Note.--N = 4 for each cell. Each proportion indicates proportion of verbalizations which were task-oriented in nature. 262 TABLE II-13 Demographic Group Means for Interference (Initiator) Categories on the Parent-Child Interaction Rating Procedure Mothers Fathers Category Mean Standard Mean Standard Proportion Deviation Proportion Deviation Lower Socioeconomic Group Black Males (N = 3 for mothers and 3 for fathers) Parent 0.962962 0.064148 0.277777 0.254585 Child 0.037037 0.064148 0.722222 0.254585 Outside 0.000000 0.000000 0.000000 0.000000 Middle Socioeconomic Group Black Males (N = 4 for mothers and 3 for fathers) Parent 0-248208 0.243418 0.592592 0.525089 Child 0.751792 0.243418 0.377104 0.462643 Outside 0.000000 0.000000 0.030303 0.052478 Lower Socioeconomic Group Black Females (N = 2 for mothers and 3 for fathers) Parent 0 045454 0-064280 0.466666 0.503321 Child 0-954545 0-064280 0.466666 0.503321 Outside 0.000000 0.000000 0.066666 0.115468 Middle Socioeconomic Group Black Females (N = 1 for mothers and 3 for fathers) Parent 0-666666 0 000000 0.629629 0.548097 Child 0.333333 0.000000 0.370370 0.548097 Outside 0-000000 0.000000 0.000000 0.000000 Lower Socioeconomic Group White Males (N = 3 for mothers and 3 for fathers) Parent 0.158333 0.323017 0.666666 0.577350 Child 0.841666 0.323017 0.333333 0.577350 Outside 0.000000 0.000000 0.000000 0.000000 263 TABLE II-l3 (continued) Mothers Fathers Category Mean Standard Mean Standard Proportion Deviation Proportion Deviation Middle Socioeconomic Group White Males 4 for mothers and 4 for fathers) Parent 0.250000 0.500000 0.295455 0.477453 Child 0.750000 0.500000 0.654545 0.445609 Outside 0.000000 0.000000 0.050000 0.100000 Lower Socioeconomic Group White Females (N = 3 for mothers and 4 for fathers) Parent 0.185185 0.169730 0.666667 0.471405 Child 0.814814 0.169730 0.333333 0.471405 Outside 0.000000 0.000000 0.000000 0.000000 Middle Socioeconomic Group White Females (N = 4 for mothers and 4 for fathers) Parent 0.158654 0.191263 0.777174 0.261096 Child 0.841346 0.191263 0.222826 0.261096 Outside 0.000000 0.000000 0.000000 0.000000 Note.--Cell sizes vary because observations were conditionalized on the occurrence of non-task-oriented verbalizations. Each proportion represents proportion of non-task-oriented verbali- zations which were initiated by the indicated source. 264 .muOqum Hmcofiumsuflmco: co mdoow How :m: 0:0 .ucwanu ne>cm unneedeafl so m900w How :m: .mamwuoume no em: xmmuicoc How :H: mo oneunmfloz 0 co Comma wuw3 mama: .mcoflumuaamnuo> pmucweuo lemleOC MO OOQQHH5000 0S“ GO UmNfiHMCOHUHUCOO 0H0? MCOHHM>H®mQO mmflmuwfl >Hm> MQNHm HHwUII.mUOZ mmeome.o semmoo.~ e oummo~.o mmmdom.e e mmememu muse: dsoum ueeoeoouoeuom messes mmmmmm.o mmmm0h.m w mmmmmm.o enhehb.a m mmHmEmm mafisz msoum UHEocoomOHUOm umzoq mmomae.o meoemm.~ e menace.o Noemmm.e e muses means dsoum unaccoomoeoom menus: Hmemda.o mmmmmm.m m su~m~m.o mmmed¢.e m duets uses; dsoum 6880806864666 86:63 060666.o mmmmmm.m m oooooo.o oomooo.~ H mmHmEmu semen muons oesocoomOeoom messes «sommm.o mmmmmm.m m mdemmo.e deHmH.H N mmHuEmm semen dsouo oesoeoumanom smsou memoem.o mmmmme.m m mmmmms.o mmeoHo.H 4 means sodas muons oesoeoumoeoom sauces ommmsm.o wooooo.m m Hmommm.o Hmmedm.~ m muses sumac dsoum oeeoeoomOuuom umzod COHUMH>00 Emmi HGQESZ COfiUMfl>QQ Gmwz HWQEDZ pumpcmum pumocmum QSOHU oflcmmHOOEwo mumcumm mumnuoz endomooum @cfiumm cofiuomumucH paflnulucmumm mnu co AmmochOEmm xmmev mocmumwumucH How mcmmz mdouw ofl£m0u0080o VHIHH m4m48 APPENDIX II-C CELL MEANS OF NONVERBAL COMMUNICATION VARIABLES Cell means and standard deviations are given for those variables included within the dimensions of nonverbal communication used in the analyses. To facilitate interpretation, raw proportions are given for those variables which are qualitative in nature rather than the logit transformations of the dichotomous and polychotomous ratios which were used in the statistical analyses. Dimensions of nonverbal commu- nication described are as follows: affective tone (Table II-lS), anxiety (Table II-l6), level of involvement (Table II-l7), and physical behavior (Table II-18). 265 266 TABLE II-15 Demographic Group Means for Affective Tone on the Parent-Child Interaction Rating Procedure Mothers Fathers Demographic Group Standard Standard Mean . . Mean . . DeViation DeViation Lower socioeconomic group black males 1.81545 0.15499 1.98109 0.05812 Middle socioeconomic group black males 2.30557 0.38069 1.98142 0.14675 Lower socioeconomic group black females 2.01770 0.16569 2.03001 0.04390 Middle socioeconomic group black females 2.25375 0.45346 2.10074 0.11172 Lower socioeconomic group white males 2.01320 0.12249 2.02649 0.06055 Middle socioeconomic group white males 2.25639 0.29639 2.14975 0.17274 Lower socioeconomic group white females 2.03092 0.06021 2.09493 0.25119 Middle socioeconomic group white females 2.10059 0.29986 2.16423 0.38614 Note.--N = 4 for each cell. Means were based on a weighting of "1" for negative affective tone, "2" for neutral affective tone, and "3" for positive affective tone. 267 TABLE II-16 Demographic Group Means for Anxiety on the Parent-Child Interaction Rating Procedure Mothers Fathers Demographic Group Standard Standard Mean . . Mean . . DeViation Dev1ation Lower socioeconomic group black males 1.22225 0.21143 1.05575 0.07514 Middle socioeconomic group black males 1.04167 0.08333 1.04545 0.09091 Lower socioeconomic group black females 1.10117 0.12281 1.06868 0.09659 Middle socioeconomic group black females 1.00000 0.00000 1.00000 0.00000 Lower socioeconomic group white males 1.06675 0.08124 1.10040 0.10710 Middle socioeconomic 9rOUP white males 1.00000 0.00000 1.00000 0.00000 Lower socioeconomic group white females 1.27082 0.23573 1.06173 0.07229 Middle socioeconomic group white females 1.00000 0.00000 1.06557 0.13115 Note.--N = 4 for each cell. Means were based on a weighting of "1" for low anxiety level, "2" for moderate anxiety level, and "3" for high anxiety level. 268 TABLE II-l7 Demographic Group Means for Level of Involvement on the Parent-Child Interaction Rating Procedure Mothers Fathers Demographic Group Standard Standard Mean . . Mean . . DeViation DeViation Lower socioeconomic group black males 1.98700 0.10873 1.94104 0.04432 Middle socioeconomic group black males 2.24693 0.41627 2.04068 0.09858 Lower socioeconomic group black females 1.97414 0.03301 2.18446 0.34640 Middle socioeconomic group black females 2.15333 0.35251 2.25855 0.49456 Lower socioeconomic group white males 1.98098 0.02483 1.90748 0.10852 Middle socioeconomic group white males 2.33897 0.41004 1.99020 0.01961 Lower socioeconomic group white females 1.96803 0.06971 2.23429 0.54699 Middle socioeconomic group white females 2.18041 0.32833 2.21609 0.31050 Note.--N = 4 for each cell. Means were based on a weighting of "1" for passive involvement, "2" for moderate involvement, and "3" for high involvement. 269 TABLE II-18 Demographic Group Means for Physical Behavior Categories on the Parent-Child Interaction Rating Procedure Mothers Fathers Category 7 Mean Standard Mean Standard Proportion Deviation Proportion Deviation Lower Socioeconomic Group Black Males Social physical 0.00031 0.00062 0.00000 0.00000 Negative physical 0.01610 0.02498 0.00250 0.00500 Defensive posture 0.00294 0.00588 0.00000 0.00000 Approach gesture 0.01316 0.02632 0.00500 0.01000 Neutral motion 0.64066 0.17460 0.56778 0.18840 Middle Socioeconomic Group Black Males Social physical 0.00630 0.00927 0.00293 0.00586 Negative physical 0.00000 0.00000 0.00207 0.00413 Defensive posture 0.00000 0.00000 0.00000 0.00000 Approach gesture 0.00000 0.00000 0.00000 0.00000 Neutral motion 0.50548 0.22546 0.44676 0.30435 Lower Socioeconomic Group Black Females Social physical 0.00000 0.00000 0.00498 0.00935 Negative physical 0.03611 0.04178 0.03125 0.06250 Defensive posture 0.00000 0.00000 0.00000 0.00000 Approach gesture 0.00862 0.01724 0.00000 0.00000 Neutral motion 0.52701 0.16831 0.41446 0.04577 Middle Socioeconomic Group Black Females Social physical 0.00720 0.01440 0.01572 0.03119 Negative physical 0.00000 0.00000 0.00000 0.00000 Defensive posture 0.00000 0.00000 0.00000 0.00000 Approach gesture 0.00463 0.00926 0.00000 0.00000 Neutral motion 0.41106 0.33563 0.64058 0.19200 Lower Socioeconomic Group White Males Social physical 0.00003 0.00005 0.00009 0.00017 Negative physical 0.02822 0.03288 0.00949 0.01150 270 TABLE II-18 (continued) Mothers Fathers Category Mean Standard Mean Standard Proportion Deviation Proportion Deviation Defensive posture 0.00513 0.00592 0.00581 0.01163 Approach gesture 0.00000 0.00000 0.00000 0.00000 Neutral motion 0.41269 0.11297 0.56445 0.15518 Middle Socioeconomic Group White Males Social physical 0.00994 0.01343 0.00024 0.00049 Negative physical 0.00000 0.00000 0.01190 0.02381 Defensive posture 0.00000 0.00000 0.03475 0.06424 Approach gesture 0.00000 0.00000 0.00595 0.01190 Neutral motion 0.51139 0.27666 0.49537 0.17761 Lower Socioeconomic Group White Females Social physical 0.00007 0.00014 0.00520 0.00603 Negative physical 0.02560 0.02107 0.00407 0.00813 Defensive posture 0.00712 0.00947 0.00000 0.00000 Approach gesture 0.00000 0.00000 0.00000 0.00000 Neutral motion 0.56350 0.29702 0.46853 0.12940 Middle Socioeconomic Group White Females Social physical 0.00282 0.00484 0.01095 0.01485 Negative physical 0.04167 0.05893 0.00000 0.00000 Defensive posture 0.00000 0.00000 0.00000 0.00000 Approach gesture 0.03125 0.06250 0.02500 0.05000 Neutral motion 0.44440 0.16067 0.37006 0.24081 Note.--N = 4 for each cell. Each proportion indicates proportion of observations designated by that category. Since not every observation Was characterized by one of these behaviors, totals do not necessarily equal 1.00. APPENDIX II-D CELL MEANS OF TEACHING PROCEDURE VARIABLES Cell means and standard deviations are given for those variables included within the dimensions of teaching procedure used in the analyses. To facilitate interpretation, raw proportions are given for those variables which are qualitative in nature rather than the logit transformations of the dichotomous and polychotomous ratios which were used in the statistical analyses. Dimensions of teaching proce- dure described are as follows: concept specification (Table II-19) and teaching method (Table II-20). 271 272 TABLE II-19 Demographic Group Means for Concept Specification Categories on the Parent-Child Interaction Rating Procedure Mothers Fathers Category Mean Standard Mean Standard Proportion Deviation Proportion Deviation Lower Socioeconomic Group Black Males New concept 0.070359 0.018969 0.051768 0.029524 New approach 0.121423 0.071976 0.113894 0.079294 New aspect 0.425603 0.095599 0.348041 0.187066 General orientation 0.016368 0.011814 0.013935 0.009817 General summary 0.035175 0.036274 0.002500 0.005000 Ongoing concept 0.180341 0.062332 0.192527 0.143127 Middle Socioeconomic Group Black Males New concept 0.069710 0.033146 0.080658 0.079584 New approach 0.062150 0.064590 0.106601 0.092313 New aspect 0.265720 0.043270 0.262067 0.103948 General orientation 0.027841 0.048403 0.018000 0.014769 General summary 0.035970 0.043129 0.054693 0.053116 Ongoing concept 0.212301 0.074533 0.220829 0.071019 Lower Socioeconomic Group Black Females New concept 0.056026 0.031529 0.077103 0.040003 New approach 0.131472 0.035573 0.060738 0.050352 New aspect 0.401712 0.118912 0.275465 0.058426 General orientation 0.006669 0.008334 0.002907 0.005814 General summary 0.007600 0.008932 0.043649 0.034846 Ongoing concept 0.227973 0.037095 0.289489 0.030935 Middle Socioeconomic Group Black Females New concept 0.114977 0.034037 0.068056 0.019450 New approach 0.107407 0.065682 0.151329 0.050258 New aspect 0.299005 0.210264 0.176049 0.015055 General orientation 0.040764 0.023931 0.036592 0.040275 General summary 0.037801 0.016343 0.019378 0.033303 Ongoing concept 0.189954 0.142365 0.227265 0.052664 273 TABLE II-19 (continued) Mothers Fathers Category Mean Standard Mean Standard Proportion Deviation Proportion Deviation Lower Socioeconomic Group White Males New concept 0.034752 0.017792 0.064659 0.051516 New approach 0.142198 0.114075 0.108478 0.111546 New aspect 0.233574 0.081493 0.222265 0.159073 General orientation 0.005115 0.007215 0.007327 0.006336 General summary 0.032581 0.020749 0.033654 0.032243 Ongoing concept 0.225792 0.054071 0.285855 0.111483 Middle Socioeconomic Group White Males New concept 0.068600 0.020698 0.075387 0.014026 New approach 0.103900 0.087848 0.083463 0.048714 New aspect 0.265649 0.105687 0.232247 0.133924 General orientation 0.019331 0.011637 0.015279 0.012410 General summary 0.024073 0.031168 0.039097 0.016160 Ongoing concept 0.244749 0.043229 0.233504 0.076073 Lower Socioeconomic Group White Females New concept 0.053508 0.012676 0.047205 0.014439 New approach 0.109829 0.131974 0.140312 0.140452 New aspect 0.328447 0.110382 0.319579 0.134350 General orientation 0.003178 0.006356 0.030874 0.014480 General summary 0.025923 0.029081 0.015536 0.019490 Ongoing concept 0.225034 0.123529 0.174573 0.126334 Middle Socioeconomic Group White Females New concept 0.089793 0.008581 0.071857 0.035613 New approach 0.051298 0.024228 0.068157 0.069578 New aspect 0.269154 0.130940 0.379551 0.150638 General orientation 0.019139 0.017592 0.027352 0.021378 General summary 0.052596 0.030003 0.043999 0.037795 Ongoing concept 0.206670 0.050038 0.183836 0.028188 Note.--N = 4 for each cell. Each proportion indicates proportion of observations designated by that category. Since not every observation was characterized by one of these behaviors, totals do not necessarily equal 1.00. 274 TABLE II-20 Demographic Group Means for Teaching Method Categories on the Parent-Child Interaction Rating Procedure Mothers Fathers Category Mean Standard Mean Standard Proportion Deviation Proportion Deviation Lower Socioeconomic Group Black Males Demonstration 0.034939 0.047793 0.016530 0.014091 Explanation 0.760967 0.072359 0.644831 0.195910 Illustration 0.051893 0.046739 0.040102 0.034200 Middle Socioeconomic Group Black Males Demonstration 0.001157 0.002315 0.014105 0.017861 Explanation 0.608379 0.009151 0.634852 0.121750 Illustration 0.062701 0.010853 0.072933 0.028309 Lower Socioeconomic Group Black Females Demonstration 0.008621 0.017241 0.008210 0.012840 Explanation 0.765170 0.117026 0.662099 0.076196 Illustration 0.029034 0.034686 0.044067 0.013515 Middle Socioeconomic Group Black Females Demonstration 0.008333 0.016667 0.034455 0.060665 Explanation 0.701829 0.033364 0.577561 0.050435 Illustration 0.079745 0.082278 0.058841 0.018866 Lower Socioeconomic Group White Males Demonstration 0.000000 0.000000 0.009763 0.016862 Explanation 0.612167 0.127326 0.632888 0.146034 Illustration 0.043995 0.029208 0.054999 0.034130 Middle Socioeconomic Group White Males Demonstration 0.001736 0.003472 0.001488 0.002976 Explanation 0.676673 0.083435 0.560506 0.049973 Illustration 0.029288 0.050293 0.086736 0.089862 27 5 TABLE II-20 (continued) Mothers Fathers Category Mean Standard Mean Standard Proportion Deviation Proportion Deviation Lower Socioeconomic Group White Females Demonstration 0.010300 0.009380 0.001263 0.002525 Explanation 0.614670 0.079096 0.645628 0.178795 Illustration 0.104243 0.092458 0.026179 0.004998 Middle Socioeconomic Group White Females Demonstration 0.000000 0.000000 0.006148 0.012295 Explanation 0.605628 0.100518 0.716662 0.162101 Illustration 0.080418 0.073337 0.051942 0.040245 Note.--N = 4 for each cell. Each proportion indicates proportion of observations designated by that category. Since not every observation was characterized by one of these behaviors, totals do not necessarily equal 1.00. APPENDIX II-E CELL MEANS OF COVARIATES Cell means and standard deviations are given for those variables included within the analyses as covariates. These are described as follows: length of session (Table II-21) and age of child (Table II-22). 276 emu-:11 .312 Tr!) 277 TABLE II-21 Demographic Group Means for Length of Session on the Parent-Child Interaction Rating Procedure Mothers Fathers Demographic Group Standard Standard Mean . . Mean . . DeViation DeViation Lower socioeconomic group black males 47.5000 32.3574 51.7500 34.9130 Middle socioeconomic group black males 67.0000 36.2399 57.7500 44.8135 Lower socioeconomic group black females 37.2500 11.3248 61.5000 29.4675 Middle socioeconomic group black females 29.5000 16.9411 44.5000 48.7203 Lower socioeconomic group white males 72.5000 30.3590 76.2500 32.1079 Middle socioeconomic group white males 57.7500 21.7773 73.2500 39.6936 Lower socioeconomic group white females 58.7500 41.1613 74.5000 43.7607 Middle socioeconomic group white females 44.7500 17.8769 33.5000 18.5921 Note.--N = 4 for each cell. Means were based on number of observation intervals for the session, where each interval represented one twenty-second period of time. 278 TABLE II-22 Demographic Group Means for Age of Child at Time of Parent's Assessment on the Parent—Child Interaction Rating Procedure Mothers Fathers Demographic Group Mean Standard Mean Standard Deviation Deviation Lower socioeconomic group black males 3.214 0.261 3.258 0.263 Middle socioeconomic group black males 3.166 0.380 3.208 0.376 Lower socioeconomic group black females 3.312 0.368 3.317 0.366 Middle socioeconomic group black females 3.207 0.374 3.247 0.424 Lower socioeconomic group white males 3.395 0.237 3.440 0.198 Middle socioeconomic group white males 3.025 0.072 3.099 0.040 Lower socioeconomic group white females 3.253 0.165 3.320 0.108 Middle socioeconomic group white females 3.062 0.211 3.162 0.200 Note.--N = 4 for each cell. Means were based on child's age in years.