n.,,_ \ obI- ..__ .1. ‘- ‘\ -‘ ‘s - t‘n . ABSTRACT MOTHER-SON INTERACTION AND THE COPING BEHAVIOR OF YOUNG BOYS By Thomas S. Rowland The present research examined the relationship between mother-son interaction and the c0ping behavior of boys. Three questions were of concern. What child var- iables are involved in the boy's ability to cope with frustration? Is the boy's manner of coping with frustra- tion related to his mother's attitudes and behavior toward him? And what is the nature of the relationship between mother's behavior and son's manner of coping? 55 were 32 preadolescent boys and their mothers, who were divided into two groups on the basis of teachers' ratings of the boys' classroom behavior. has. were boys rated low in self-control, self-sufficiency, and achieve- ment motivation, while highs had received high ratings on these three variables. A high degree of pair-wise matching was achieved on other presumedly relevant factors. Mothers l ‘.-~ ‘ you: 0". n :- -..n I uoa‘ Ita- I.. ~ I" “n a... I: Thomas S. Rowland filled out an abbreviated version of the Stanford Parent Questionnaire, and they participated in two interaction situations with their sons. In the first session (frus- tration session) the boy was intentionally frustrated by offering him a prize for completion of a puzzle which was too difficult for him, while the second session (verbal session) required the boy's mother to teach him three proverbs . Observing the frustration session through a one-way mirror, two judges rated the boy's reactions to frustration and his mother's responses to them. Both sessions were tape-recorded, and the verbal interaction was coded using the interpersonal rating scheme and system of categories developed by Freedman, Leary, Ossorio, and Coffey. Interjudge agreement was very high for both rat— ing methods. Analysis of the "reactions to frustration" data revealed that low boys demonstrated aggressive and regres- sive reactions to a significantly greater extent and con- structive reactions to a significantly lesser extent than high boys. The two groups did not differ with regard to yithdrawal and intr0punitive reactions. Likewise, the two 2 .VA..“: Gav ‘J' , n . m5. -, 0'" o—\ mgr - 5' V. t a npn- a. uv..~.u p. . g — ‘ n on, . . .J‘ van It...‘ u..._J - u u" Q“‘ ”’t-«- . _.‘. _- +- Thomas S. Rowland groups of mothers differed significantly in their responses, with mothers of lows giving more negative responses and fewer nondirective and positive responses than high_mothers. Consideration of son-mother sequences indicated that lgw_ and high_mothers showed different patterns of reinforcement for the same son behaviors, with the differences suggesting that lg! mothers tend to give negative and high mothers positive reinforcement for their sons' efforts of mastery. Similar findings were reflected in a comparison of the groups with regard to their verbal interpersonal be- lmvior. L 3 boys demonstrated significantly more behavior that was poorly controlled, passive—aggressive, and nega- tively dependent, while high_boys showed significantly more positively assertive behavior. These differences were cmmplementary to those for the two groups of mothers. ng_ nmthers showed significantly more dominant and protective twhavior and a trend toward greater rejection, while high, anthers demonstrated significantly more autonomy granting Imhaviors that were both positive and negative in nature. Interaction sequences were analysed separately for son- nmther and mother-son interaction. When son sender 3 .9 .- Thomas S. Rowland behavior was statistically "equated" for the two groups, lg! and high_mothers evidenced different patterns of re- sponse to the same stimulations. Similarly, lgw_and high boys responded differently to the same mother sender be— havior, suggesting that the differences in interaction for the two groups were a function of the boys as well as their mothers. In their responses to each other, 19!, anther-son pairs tended to maintain a pattern of negative interaction and unconstructive activity. Both mother and son acted in ways which would serve to perpetuate a com- ;flementary "dependent son--controlling mother" relation— ship, which stands in contrast to the more symmetrical, nmtually assertive relationship of high_mother-son pairs. Comparison of the two groups of mothers on the Stanford Parent Questionnaire (SPQ) revealed that lgw anthers scored significantly higher on rejection, incon- EiéEsggy, punitiveness and physical punishment, demands £E£ aggression, democragy, and demands for conformity than did high mothers. Integration of the SPQ and behavioral interaction findings led to the interpretation that the love relationship for lgw_mother-son pairs is not a secure 4 uh: I ".5, ‘ 4.0-1- en'vo . u can . u h.“. urn-e Iii. n Ina... I’h‘ . "'1‘ a a... 5-..- . a... ‘uu. ‘ ' Q; t”, ‘.‘ ’b Thomas S. Rowland one, that consequently mother and son must be continually involved with each other. The lgw_mother's ambivalent feelings are manifest in her inconsistent and dominant, protective behavior toward her son, while the lgw_boy's anxious dependency results in a passive-aggressive and negatively dependent stance toward his mother. While the results lend support to the notions that boys' poorly controlled behavior is related to an aggressive parental model and that dependency in boys is related to maternal inconsistency and dominance, it was concluded that an interaction of parent effects and child effects can best account for the "dependent son--controlling mother" rela- tionship observed in $93 mother-son pairs. Several meth- odological issues were discussed in the light of the Present study, and a new approach was suggested for family interaction research. % / Approved: »A‘<»1fi2 X4J\Q5LA&N\ Committee Chairman Date: 0 c/on/epvx lOr Lab? MOTHER-SON INTERACTION AND THE COPING BEHAVIOR OF YOUNG BOYS BY :f' Thomas SibRowland A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1968 To Polly With Gratitude To Heather With Hope If '1’ ‘1) (I. r) ACKNOWLEDGEMENTS I would like to thank Dr. Lucy Ferguson, Committee Chairman, for her contributions to this research as well as to my growth as a clinical psychologist and person. I'm also grateful to Dr. Clarence Winder, Dr. Bill Mueller, and Dr. JOhn Hurley for serving on my committee and for their helpful suggestions regarding research design and data _analysis. Research dealing with observed interaction could not be accomplished without reliable behavioral ratings, and for this task I am indebted to Steve Bogen and Gary Davis. Thanks likewise goes to Jerry Musgrave for developing com- puter programs which made possible an economical and thorough analysis of the data. I would also like to thank Dr. Edward Remick, Re- search Director for Lansing Public Schools, and the Prin- cipals and teachers of Wainwright and Pleasant View Schools for their assistance in obtaining research subjects. Finally, I would like to express my special grati- tude to my wife Polly for her creative ideas and for the long hours spent in preparing the manuscript. iii TABLE OF CONTENTS Page ACKNOWLEDGEMENTS. . . . . . . . . . . . . . . . . . . iii LIST OF TABLES. . . . . . . . . . . . . . . . . . . . vi LIST OF APPENDICES. . . . . . . . . . . . . . . . . . ix Chapter I. INTRODUCTION. . . . . . . . . . . . . . . . . 1 Statement of the Problem. . . . . . . . . . l Parent—Child Research Methodology . . . . . 3 Interpersonal Framework . . . . . . . . . . 12 Boys' Coping Behavior . . . . . . . . . . . l8 Self-Control . . . . . . . . . . . . . . 20 Self-Sufficiency . . . . . . . . . . . . 25 Achievement Motivation . . . . . . . . . 30 Hypotheses. . . . . . . . . . . . . . . . . 35 II.METHOD.................... 40 Subjects. . . . . . . . . . . . . . . . . . 40 Procedure . . . . . . . . . . . . . . . . . 48 Interaction Session. . . . . . . . . . . 48 Rating Techniques. . . . . . . . . . S4 Stanford Parent Questionnaire. . . . . . 68 iv ~~‘\. .. -:eoav In" '96 n QC. (1 a... v I TABLE OF CONTENTS (continued) Chapter III. RESULTS . . . . . . . . . . . . . . . Boys' Behavior. . . . . . . . . . Hypothesis I: Reactions to Frustration. Hypothesis II: Sons' Behavior Toward Mother . . . . . . . . . . . . Mothers' Behavior . . . . . . . . . Hypothesis III: Responses to Sons' Frustration. . . . . . . . . . . Hypothesis IV: Mothers' Behavior toward Sons. . . . . . , . . . . . . Hypothesis V: Mothers' Evaluation of Sons' Ability. . . . . . . . . Mothers' Attitudes. . . . . . Mother-Son Interaction. . . . . . . Reactions to Frustration . . . Interpersonal Behavior . . . IV. DISCUSSION. . . . . . . . . . . . . . Reactions to Frustration. . . . . . Interpersonal Behavior. . . . . . . Mothers' Attitudes. . . . . . . . ‘Vfl SUMMARY . . . . . . . . . . . . . . . . . REFERENCES. . . . . . . . . . . . . . . . . . APPENDICES . Page 70 7O 7O 74 81 82 85 92 93 102 102 107 120 124 129 141 152 157 163 LIST OF TABLES Table Page 1. Wainwright Sample: Intercorrelation among Teachers' Ratings of Boys Classroom Behavior and Two Independent Judgments of Socio— Economic Status (SES) . . . . . . . . . . . . 43 2. Comparison of High and Low Groups and the Wainwright Sample on Relevant Group Characteristics . . . . . . . . . . . . . . . 46 3. Reliabilities for Two Independent Raters on Son and Mother Categories for Reactions to Frustration . . . . . . . . . . . . . . . . . 58 4. Total Agreement (TA). Octant (Oct). and One- Step (l-S) Per Cent Agreement Reliabilities for Two Independent Raters. . . . . . . . . . 66 5. Total Agreement (TA). Octant (Oct). and One- Step (l-S) Per Cent Agreement Reliabilities for the Sixteen Circumplex Categories . . . . 67 5. Comparison of Low and High Boys on their Reactions to Frustration. . . . . . . . . . . 72 7. Comparison of Low Boys with High Boys on the Proportions of their Behavior in the Major Sections of the Circumplex. . . . . . . . . . 76 3- Comparison of Low Boys with High Boys on the Proportion of their Behavior in the Circum- plex Octants. . . . . . . . . . . . . . . . . 8O 9- Comparison of Mothers of Low Boys with Mothers of High Boys on their Responses to Sons' Frustration . . . . . . . . . . . . . . . . . 84 vi LIST OF TABLES (continued) Table 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. Comparison of Mothers of Lows with Mothers of Highs on the Proportions of Maternal Behavior in the Major Sections of the Circumplex. . . . . . . . . . . . . . Comparison of Mothers of Lows with Mothers of Highs on the Proportions of Maternal Behavior in the Subcategories of Affiliation-Control . Spearman Rank Correlation Coefficients Between Circumplex and Reaction to Frustration Categories for the Frustration Session Ratings . . . . . . . . . . . . . . . . . Comparison of Mothers of Lows with Mothers of Highs on Sixteen SPQ Scales . . . . . . . . Intercorrelations of 16 SPQ Scales. Ordering of Mean Scores on Eight SPQ Scales for Mothers of Poorly Adjusted (PA). Well- Adjusted (WA). Low (Lo). and High (Hi) Boys Mean Proportions of Son-Mother Interaction Occurring in the Reaction to Frustration Dyads for Low and High Mother-Son Pairs . Distribution of the Mean Proportions of Mothers' Response in the Low and High Groups for Each of the Sons' Reactions to Frustraé tion. . . . . . . . - . . . . . . . . . Mean PrOportions of Son-Mother Interaction Occurring in the Various Circumplex Quadrant Dyads for Low and High Mother—Son Pairs . Comparison of the Response Distributions of Low and High Mothers for Quadrants of Sons' Sender Behavior . . . . . . . . . vii Page 86 90 92 95 97 101 104 106 109 111 225? OF 1 table Me LIST OF TABLES (continued) Table Page 20. Mean Proportions of Mother-Son Interaction Occurring in the Various Circumplex Quadrant Dyads for Low and High Mother-Son Pairs . . . . . . . . . . . . . . . . . . . 114 21. Comparison of the Response Distributions of Low and High Boys for Quadrants of Mothers' Sender Behavior . . . . . . . . . . . . . . 116 viii ix ‘u-‘I“ ‘1‘ 'II n .guio. Q41 ‘Qi Hr. Kit. .u. . it. LIST OF APPENDICES Appendix Page A-l RATING INSTRUCTIONS TO TEACHERS. . . . . . . 165 A-2 LETTER TO PARENTS. . . . . . . . . . . . . . 171 A-3 HIGH-LOW PAIRS AND THEIR MATCHINGS ON RELEVANT CHARACTERISTICS . . . . . . . . . 172 ‘B-1 PROVERBS . . . . . . . . . . . . . . . . . . 174 3-2 CIRCUMFLEX CATEGORIES FOR MOTHER AND SON BEHAVIOR . . . . . . . . . . . . . . . . . 175 B-3 RATING SHEET FOR INTERPERSONAL MECHANISMS. . 185 C-1 M.S.U. PARENT QUESTIONNAIRE. . . . . . . . . 187 C-2 16 SPQ SCALES AND THEIR RELIABILITIES. . . . 205 D-l COMPARISON OF FRUSTRATION (F) AND VERBAL (V) SESSION RATINGS OF BOYS' BEHAVIOR. . . . . 207 D-2 COMPARISON OF FRUSTRATION (F) AND VERBAL (V) SESSION RATINGS OF MOTHERS' BEHAVIOR . . . 208 E-1 DISTRIBUTION OF Low MOTHERS' RESPONSES ACROSS CIRCUMFLEX CATEGORIES FOR MAJOR CATEGORIES OF SONS' SENDER BEHAVIOR. . . . 210 3-2 DISTRIBUTION OF HIGH MOTHERS' RESPONSES ACROSS CIRCUMFLEX CATEGORIES FOR MAJOR CATEGORIES OF SONS' SENDER BEHAVIOR. . . . 211 ix LIST OF APPENDICES (Continued) Appendix Page E-3 DISTRIBUTION OF ALL MOTHERS' RESPONSES ACROSS CIRCUMPLEX CATEGORIES FOR MAJOR CATEGORIES OF SONS' SENDER BEHAVIOR. . . . 212 E-4 DISTRIBUTION OF LOW BOYS' RESPONSES ACROSS CIRCUMPLEX QUADRANTS FOR INDIVIDUAL CATEGORIES OF MOTHERS' SENDER BEHAVIOR . . 213 E-S DISTRIBUTION OF HIGH BOYS' RESPONSES ACROSS CIRCUMPLEX CATEGORIES FOR MAJOR CATEGORIES OF MOTHERS' SENDER BEHAVIOR. . . . . . . . 214 E-6 DISTRIBUTION OF ALL BOYS' RESPONSES ACROSS CIRCUMPLEX QUADRANTS FOR INDIVIDUAL CATEGORIES OF MOTHERS' SENDER BEHAVIOR . . 215 &.".E 6,, ....e ¢.' 2» “A vtg u Ill ’ I K). CHAPTER I INTRODUCTION Statementgof the Problem Impetus for the present research was provided by an exploratory study of mother-son interaction (Rowland. 1966) which compared the behavior of mothers of poorly adjusted (PA) boys with that of mothers of well-adjusted (WA) boys under conditions that were frustrating to the boy. Boys were offered a desirable toy as a reward for solving a puzzle which. because of its difficulty. was un— likely to be solved within the allotted time. There were three major findings of this exploratory study. First of all. frustrating the boy proved to be a profitable tech- nique for eliciting interaction between mother and son. Secondly. two independent techniques for assessing ma- ternal attitudes and behavior. the observational method and the self-report. questionnaire method yielded data that were both consistent and supplementary. And finally. three dimensions of maternal attitudes emerged as bearing a significant relationship to the boy's adjustment. these being Warmth-Hostility. encouragement of Independence— Dependency. and Self-esteem. The exploratory research raised several questions that will be considered in the present study. What child variables are involved in the boy's ability to cope with frustration? In the exploratory study the investigator's frustration of the boy served as a technique for "pulling" spontaneous behavior from the mother. The boy's response to frustration was not studied in its own right. Although the boys in the pilot research differed with regard to a global assessment of "school adjustment." evidence as to their standing on less global behavioral variables was lacking. Since the research is embedded within the frame- work of parent-child relations. the second question is ob- vious. Is the boy's manner of coping with frustration re- lated to his mother's attitudes and behavior toward him? These two questions do not present any special problems for research design. but a third and most crucial question does. If mother's behavior toward her son is related to the ways in which he deals with frustration. what is the nature of that relationship? This question asks for a closer look at what "goes on" between mother and son in their interaction. whether there exist between them con- sistent patterns of interaction which set up and maintain the boy's coping behaviors. The answer to this question may be obtained by a sequential analysis of mother—son interaction. Parent4Child Research Methodolggy How has existing research dealt with the subject of mother-son interaction? The classic approach has been to interrogate mother. either by means of a questionnaire or an interview. This approach offers the advantages of being able to gather data on a large number of variables and on both present and past behavior. and in the case of the questionnaire. often permits greater economy and pre- cision of measurement than do behavioral rating schemes. However. interrogation of the mother suffers several dis- advantages (Yarrow. 1963). the most serious of which would seem to be that mother supplies the information about her child's behavior as well as her own. Such a confounding of variables does not occur in observational studies of Rmther-son interaction. but this method also has its difficulties. Smith (1958) talks about the problem of insufficient and unreliable sampling of mother and child behavior. and Vidich (1956) deplores the mechanical and "forced" nature of husband-wife interaction elicited under laboratory conditions. It is the conclusion of this writer that self—report and observational techniques must be em- ployed as supplements to each other. if the difficulties inherent in both are to be overcome. Most of the studies which have attempted to observe mother-child interaction in the laboratory have focused primarily on methodological issues. The first of these studies was done by Merrill Bishop (1946). She observed mothers and their preschool children interacting in a play- room setting. and she recorded the mothers' behavior by making a behavioral rating every 5 seconds. Mothers who were told that their children's play was of inferior qual— ity showed an increased amount of interference with and direction of that play. This manipulation of instructions in order to elicit the particular maternal behavior of interest has also been accomplished by frustrating the Child (Rowland. 1966) and by flunking the child on a test (Hilton. 1967). In a subsequent study Merrill Bishop (1951) in- creased the sc0pe of the observational approach by look- ing at the child's behavior as well as the motherzs. For this purpose she developed a very elaborate system of be- havioral categories. The category totals for mother be- havior and for child behavior were correlated with each other. with the finding that certain types of behavior tended to coexist in mother and child. It was observed. for example. that aggressive behavior by the child was positively associated with negative behavior on the part of the mother. Moustakas. Sigel. and Schalock (1956) have further elaborated the Merrill BishOp system of categories. and they have used the system with a high degree of relia- bility to code mother—child interaction in the home. and therapist-child and mother-Child interaction in the play- room. Neither Merrill Bishop nor Moustakas et al. have examined the association of mother and Child behavior as they occur in same time sequence. One reason why such Contingencies have not been explored may be due to the Practice of coding interaction at 5 second intervals. Interval scoring breaks up the interaction into artificial units and may disrupt the uncovering of what follows what in that interaction. Another reason why the Moustakas et a1. system of categories is unsuitable for examining con— tingencies of response in mother—child interaction is that it affords too many possible mother—child or child-mother interactions to look at. the exact number being 14.596 dyadic sequences. Other researchers have developed more economical coding Systems. using a small number of categories to cap— ture that portion of the total interaction which bears on a specific problem. Ruebush. Byrum. and Farnham (1963). in a study of problem solving in low and high defensive preadolescent boys. coded only that maternal behavior which was "helping" or "evaluative." Bing's (1963) study of the effects of maternal attitudes and behavior on pre— adolescent children's cognitive development focused only on mothers' helping behavior. Both studies failed to con- Sider the child's participation in the interaction. Hil— ton's (1967) research on maternal behavior and ordinal POSition of the child also explored a limited range of tmhavior and did take into account the behavior of the tfluld toward his mother. Yet. as has been the case with the other observational studies of mother-son interaction. this research did not examine sequential contingencies or patterns of interaction. Turning now to studies which have relied on interro- gation of the mother as the primary source of data on the mother-son relationship. one finds conflicting results. These studies cannot provide reliable data on the actual interaction between mother and son. but they can determine what maternal attitudes and behaviors are related to be- haviors of the boy. Schaefer and Bayley's (1963) longi- tudinal research. which used both observational and self— report measures of maternal behavior. intercorrelated the mother and child behavior of 54 Ss from infancy through adolescence. A very high degree of consistency over time was found in maternal behavior and in boys' behavior. but this was not true of girls' behavior. The correlations between observational data on mothers' behavior when their children were infants and interview data collected when their children had reached adolescence were quite high for the love—hostility dimension (r = .68) and only moderate for the autonomy-control dimension (r = .26). Two of the behavioral variables for boys showed extremely high «uI\ IQ consistency from early childhood through adolescence. These are facility. which includes responses to a task that facilitate performance. and attentiveness. which also includes persistence. Taken together these two var- iables are very similar to the index of constructive COp- ing behavior used in the present study. With regard to the relationship of mother behavior to child behavior. the highest correlations were between Inother and son behavior when the boy was 9 to 12 years old. Mothers' ignoring. punitivengss. and irritability showed liigh negative correlations and mothers' affection. eguali- jgarianism. and_pgsitive evaluation of son showed high pos- .itive correlations with boys' social behavior (friendly. cnaoperative) and with boys' attentiveness and facility. Phathers' standing on the autonomy-control dimension was Iinrelated to boys' task-oriented behavior. but mothers' iflatonomquas negatively associated and mothers' achieve— }Egnt demand and emotional involvement were positively associated with boys' social behavior. Medinnus (1961) used the Fels Parent Behavior Rating Scales and the Parent Attitude Research Instrument (PARI) to compare mothers of poorly adjusted and well— adjUSted first graders. His finding that mothers of well-adjusted children are more encouraging of dependency than mothers of poorly adjusted children is not consistent with the results of this author's research on mothers of 7 to 9 year old boys (Rowland. 1966). These latter results showed mothers of poorly adjusted boys to be either very encouraging of dependency or very encouraging of independ- ence. while mothers of well-adjusted boys demonstrated be- havior and attitudes falling between these extremes of the dependency:independence encouraging dimension. Use of a different sample of Ss as well as use of a different be— havioral rating scheme and attitude questionnaire (Stan- ford Parent Questionnaire) may be responsible for the dis— crepant findings. Several studies which have used other parent atti- tude questionnaires have found little relationship between Inothers' attitudes and teachers"ratings of boys' behavior. ILeton (1958) employed Shoben's Parent Attitude Survey (PAS) and the Minnesota Parent Attitude Inventory (MPAI). Gildea (1961) used the PAS. and Friedman (1964) relied on Here- fOrd's model for parent attitude measurement. and none of these instruments adequately differentiated mothers of well-adjusted and poorly adjusted boys. Brody's (1963) 10 use of the PARI and the Maryland Parent Attitude Survey (MPAS) to assess the relationship of maternal attitudes to mother-child interaction yielded similarly disappoint— ing results. This review of some of the parent-child research in relation to the specific area of mother—son interaction sets the stage for a few evaluative comments. First of all. there is disagreement among studies regarding whether or not mothers' behavior and attitudes are related to boys' 'behavior. Those studies which point to such a relation- ship have employed behavioral ratings and interviews of Inothers. the SPQ being the only questionnaire to yield such results. Secondly. observational studies of mother- son interaction are still in the exploratory stage. with this body of reSearch supplying few substantive findings regarding the relationship between mothers and their school-age boys. Thirdly. the results of parent-child research have almost always been interpreted within a "parent effect" framework. The child has been considered a passive recipient of parent behavior which molds him into one form or another. Although researchers have been Careful to substitute "associated with" in place of "caused." 11 the usual design of parent-child research implies this orientation of one-way effect from parent to child. For the sake of argument Bell (1968) presents the case for "child effect" by reinterpreting some of the ex- isting research and by pointing to some maternal behaviors of animals which appear to be a function of the stimulat- ing and selective effect of the young. He theorizes that parents have hierarchies of actions. and that different children induce responses from different portions of the hierarchy by reinforcing or failing to reinforce different parental behaviors. Parents demonstrate two kinds of con- trol. Upper limit control behavior (UCL) serves to reduce or redirect children's behavior which is too intense. 'while lower limit control behavior (LLC) stimulates chil- dren's behavior which is too low in intensity. The ULC parent is well-known in parent-child research as "punitive" and."restrictive" while his LLC counterpart is often re- ferred to as "demanding." A final point relevant to much parent-child re- search is that the mother-son relationship has been Studied in isolation from other intra-family relation- Ship3. Murrell and Stachowiak (1965) regard this practice ‘ us up... a man ...J.,.. euulsA-u . 9-2: a {been 5 . § '1‘“ ”T L»; ..C I ‘ . "C h‘ "chi y. TO‘AQr “V v.‘\—. “YAeny revue“ III’ '1'! in (I) 12 as undesirable because the child must not only cope with individuals (mother-son. father-son interaction). but he must also COpe with a two—person group (mother and father) who have formed a system for meeting their needs prior to his birth. Although this argument does not invalidate the mother-son approach. one should be aware that such an ap- proach does not give a complete picture. Interpersonal Framework Further amplification of the interactional point of view referred to in the Bell (1968) and Murrell and Stachowiak (1965) studies may be found in the works of both therapists and researchers who Operate within the framework of interpersonal thery. From Haley (1963) comes the notion that both participants in a relationship exer- Cise "maneuvers" on the other in order to gain control of the relationship. Although it is easy to perceive a dom- ineering parent as very much in control of a Child. per- liaps less obvious but no less real is the relationship in ‘Mhich a child uses his helplessness to control a parent. The parent's dominance is forthright. the child's Law ”baht- D nan. ,- by“... 13 helplessness much more subtle. but both are attempts to control the parent~chi1d relationship. Kell and Mueller (1966) invoke the concept of "eliciting behaviors" to de— scribe the client's attempts to set up certain patterns of interaction with the counselor. On the basis of his past relationships with significant others. the client has learned to deal with people by expecting and thus eliciting certain behaviors from them. Martin (1967) calls these elicitors "cues" and points to the expected response as "reinforcement" of the pattern of interaction. In his work with families he tries to point out highly repetitive interaction sequences which are "set off" in more or less automatic fashion. Freedman. Leary. Ossorio. and Coffey (1951) View interaction in a similar way. Using as a cornerstone the concept of interpprsonal mechanism. which describes behavior in terms of its interpersonal function. these authors have designed a two-dimensional (love-hostility. dominance-submission). Circumplex arrange- ment of behavioral categories. This system would seem to be the most successful attempt thus far to capture behav— ioral interaction. The Circumplex rating scheme has been used by RauSh. Dittman. and Taylor (1959) to investigate the n' \. (f (t) '. J ‘ t): 56' I 56—55; :P‘Ilr owns 7" b. y .0» 4 I *znfia‘ “an-nu... Vi‘a v evoke acts 0 ‘"~v . Quflge: 'v 5-H... l4 behavior of hyperaggressive preadolescent boys. They ob- served the boys for brief periods in various naturalistic settings and coded the dictated observations into the four quadrants of the Circumplex. In their interaction with peers. those boys who acted in a hostile-submissive manner were responded to in a hostile—dominant way and Visa versa. Hostile-submissive behavior tended also to evoke friendly-dominant behavior. while hostile-dominant acts often resulted in friendly-submissive responses. The same procedure was employed to compare hyper- aggressive boys to well-adjusted controls with respect to their behavior toward adults (Raush. Farbman. and Llewel- lyn. 1960). The hyperaggressive boys were observed on two occasions separated by 18 months. with one same—age control group being observed before and another after this time span. The hyperaggressive boys showed significantly more hpstile-dominant and hostile-submissive behavior toward adults than did the controls. while the controls exceeded the hyperaggressive boys with regard to friendly-submissive behavior. Friendly:dominant behavior. which occurred very infrequently. failed to differentiate the groups. After 18 months these differences were much less marked. due to ~ 5" ”A (I. :I ‘Ff‘! no. kHL. 1:") k“ C ‘- ‘ p . ."au P r- . r p a w p- .4 I)! (I) C P» l s.“ “a; ‘ s .‘ ! 1 Nd “A. “'0 15 the decreases which hyperaggressive boys demonstrated in both kinds of hostile acts and their corresponding in- crease in friendlyfsubmissive acts. Both groups showed an increase in friendly-dominant behavior. which is in keeping with their greater proximity to adulthood. Adults "responded" in complementary fashion by decreasing their friendly-dominant behavior and increasing their friendly- submissive behavior. At all times hostility begot hostil- ity. but the adults tended to "send" less hostile behavior than they "received." Raush (1965) has attempted to examine the sequen- tial aspects of the data generated by his research with hyperaggressive boys by utilizing multivariate information analysis. He found that the most significant determinant of the friendly Vs hostile behaviors of these boys toward their peers was the antecedent act or interpersonal stimu- lus to whiCh they were responding. To a lesser but signif- icant extent these behaviors were a function of group mem- bership and situational factors. However. the boys' friendly or hostile responses to ggglg'behavior were de— termined less by the specific antecedent acts of the adult than by group and situational variables. Although Raush pen: a! .- .b.».-.. 36 t3: ' ‘ 7C?Hfl.nr “Vb-.vu” fill ‘ MM ‘ til t. I. n- m. “‘SF‘SS {.7 Q. Q,‘ \I 16 has worked only with gross behavioral variables. his pioneer effort illustrates a very important point. One must do more than describe interaction in order to de- termine why 55 behave in that interaction as they do. Statistically holding variables constant or partialing out their effects so that the effects of one variable may be studied by themselves is an important addition to the methodology of research on interaction. MacKenzie (1968) also used the Circumplex rating scheme of Freedman et al. in her comparison of the inter- action of normal and clinic families. Although she exam- ined three—person interaction (mother-father—son) as well as parent-child interaction (mother-son. father—son). only the data relevant to mother—son interaction will be discussed here. The clinic boys ranged in age from 7 to 11 years. and they had been referred to a psychological clinic because of aggressive. poorly—controlled behavior and underachievement in school. With one exception. com— parison of the boys' groups regarding their behavior to- ward mothers yielded the same pattern of results that Raush et al. (1960) found with his maladjusted and normal groups. MacKenzie's aggressive clinic boys demonstrated 17 a significantly greater amount of hostile-dominant and hostile—submissive behavior than did the normal controls. while the normal controls exceeded the clinic boys on ‘pggg types of friendly behavior. The friendlyrdominant quadrant differentiated these two groups but failed to differentiate the groups of Raush et a1. However. the results of both studies indicated that boys' behavior is not very often scored in the friendly—dominant quadrant. possibly because judges used an adult frame of reference which prevented them from viewing children as showing dominant behavior in relation to adults. This problem suggests a need for separate definitions of the categories for adults and children in cases where the object of the research is to compare same-age groups with each other rather than comparing adults with children. As was the case with their sons. mothers of clinic boys showed significantly more hostile-dominant behavior and significantly less friendly behavior of both types than did mothers of normal sons. The two groups of mothers did not differ with regard to hostile—submissive behavior. An analysis of the different types of interaction in the two groups revealed the following broad patterns. Normal 18 sons exhibited primarily friendly—submissive behaviors. to which their mothers responded with behaviors that were friendly-dominant. The clinic sons often behaved in a hostile-submissive manner. and their mothers reciprocated with both hostile—dominant and friendly—dominant behaviors. This "pattern" also existed in reverse. whereby clinic mothers behaved in a dominant way and their sons reacted with hostile-submission. Thus. this analysis of interac- tion suggests the existence of a circular and seemingly self—perpetuating pattern of behavior between the clinic sons and their mothers. There did not appear to be any such "locked in" patterns of interaction for the normal mother-son pairs . Boys: Coping Behavior Frustration may be conceived as an increase in tension when the satisfaction of a need is blocked (Rosen- zWeig. 1944). Applied to the preadolescent boy in a PrOblem—solving situation. two things must be present flu: frustration to occur. The boy must be motivated to- ‘Ward.some goal connected with the task. successful comple- tion of the prOblem and its accompanying rewards. and he 19 must perceive a more or less insurmountable obstacle to lie in his path. Several factors may cause the same set of circumstances to precipitate differing degrees of frus— tration in different boys. factors such as differing per- ceptions as to the insurmountability of the obstacle and the value of the reward. different aspiration levels. and different feelings about success and failure. However. any boy who is both motivated and blocked experiences frustration to some extent. Frustration. then. is an intrapsychic phenomenon *which must be inferred from the child's behavior. Barker. JDembo. and Lewin (1941) observed regression in the play of ‘preschool children after they had been denied use of some new and exciting toys. Yarrow (1948) found that frustrated ;preschoolers react with increased aggression in doll play asessions. The relatively uninhibited. nonstereotyped na- ‘ture of the resultant aggression caused the author to con- '=1ude that frustration also results in a disorganization ‘15 the child so that he can less effectively control his impulses. Preadolescent boys showed a variety of reactions ‘13 frustration. ranging from tension releasing behaviors thatwere poorly controlled (aggressive. regressive) to 20 those that were rigidly controlled (withdrawing. intr0pun- itive). Present also were constructive means of releasing tension such as laughing or talking about the obstacle. and continuing efforts to solve the problem (Rowland. 1966). It is felt that the boy who reacts aggressively in a frustrating situation is not necessarily experienc- ing greater frustration than the boy who reacts construc- tively. The two boys are presumedly coping with pretty much the same situation in different ways that are a func- tion of differences in personality organization for the two. Frustration tolerance. the ability to cope with frustration so that constructive activity continues. would seem to be primarily a function of three variables. These three--appropriate impulse control. self-sufficiency. and achievement motivation--are conceived as different. though interdependent dimensions of the boy's personality. §elf Control Block and Martin (1955) classify children into three classes of ego-control. Under-controllers tend to 21 be impulsive. distractible. and unable to delay gratifica- tion. They react to frustration with direct and unmodu- lated impulse expression. Over—controllers are constrained and allow themselves only very indirect impulse expression. And appropriate-controllers selectively bind and discharge tensions with respect to reality considerations. In a study of the play behavior of preschool children under conditions of frustration. under-controllers regressed in their play. while overegontrollers maintained integrated. constructive play. However. the writer's observations of frustrated preadolescent boys (Rowland. 1966) suggest that the overly controlled boy may not cope with frustration any more adequately than the boy who lacks sufficient con— trols. While he inhibits direct and obviously regressive and aggressive expressions. the overly controlled boy seems to develOp "emotional blocks." isolation tendencies. and to direct his anger inward. The net result appears to be a halt in constructive activity. What maternal behaviors and attitudes are related to the self-control dimension of boys' behavior? Becker (1964) brings the results of many studies together in his comparison of two major types of parental discipline. 22 Disciplinary techniques which utilize the love relation- ship with the child are highly correlated with internal— ized reactions to transgression and with nonaggressive and cooperative social relations. Praise and reasoning seem to be the most effective of these techniques. Becker theorizes that reasoning is effective in three ways. First of all it provides a model of restraint. Secondly. it provides the child with an understanding of what he did wrong. enabling anxiety about misbehavior to become con- nected to the proper cues. Thirdly. reasons can be used by the child to build up his own internal means for eval- uating his behavior. Power-assertive disciplinary tech- niques. on the other hand. are highly correlated with ex- ternalized reactions to transgression and with behavior that is aggressive and noncooperative. Hoffman and Saltzstein (1967) further differen- tiated techniques utilizing the love relationship. .13- duction is the practice whereby the parent points out the painful consequences of the child's transgression for the Parent and others (e.g.. indicate his disappointment in the child). Love withdrawal is a direct but nonphysical eXpression of anger or disapproval toward the child 23 (e.g.. ignore. isolate). Utilizing several different moral indices on a very large sample of normal 7th grade boys and girls the following results were obtained. Mothers' discipline was highly related to boys' moral development but fathers' was not. Mothers who used ig— ggggigg had boys who were high on all of the moral indices. and mothers who used power assertion had boys who stood low on the indices. Mothers' love-withdrawal showed little relationship to the moral indices. The authors conclude that power assertion is ineffective because it makes the Child intensely angry. it frustrates his need for autonomy. and it gives him no available means of reparation. The findings of Goldstein. Judd. Rodnick. Alkire. and Gould (1967) also point up the ineffectiveness of power asser- tion. Parents of two groups of externalizers. antisocial. aggressive adolescents and passive-negative adolescents. made many direct and implied demands on them and very rarely asked them for opinions. Many studies have demonstrated a relationship be- tween power assertion by parents and aggression in chil- dren. In a study of eight-year olds which employed peer ratings of aggression. Lefkowitz. Walder. and Eron (1963) 24 found parents' physical punishment to be highly related to children's aggressiveness. This relationship was most marked in the case of mothers; and both mothers' physical punishment and children's aggressiveness were significantly and negatively associated with children's confessing be— havior. These findings are consistent with those of McCord. McCord. and Howard (1961) on lower-class. nondelinquent boys. Mothers' physical punishment. rejection. and incon- sistent discipline were significantly related to boys' aggression. A comprehensive study by Becker. Peterson. Hellmer. Shoemaker. and Quay (1959) explored the factor structure of parental behavior and its relationship to school chil— dren's personality problems. The strongest relationship existed between a "general family maladjustment" factor and children's aggression in school. The Specific factor loadings suggest the following explanation. In the parent— child relationship. mother is dictatorial and thwarting of the child. and father does not enforce regulations. Because Of these personality differences and due to marital dishar- mony the total impact of both parents on the child is one Of inconsistent and ineffective discipline plus a poor 25 model for controlling his impulses. The net "effect" is aggressive and poorly controlled behavior in the child. This interpretation is entirely consistent with that of McCord et a1. (1961); and the finding of Sears. Maccoby. and Levin (1957) that preschoolers who were highest in aggression had mothers who were both punitive and permis— sive may also be relevant. in the sense that both parents may accomplish the same end. Employing a similar research design with five-year olds and their parents. Becker. Peterson. Luria. Shoemaker. and Hellmer (1962) found much the same results. The findings were most marked for the mother—son relationship. with maternal hostility. use of physical punishment. and overprotection all significantly related to children's aggression. §elf-Sufficiency Self-sufficiency would seem to involve three ele- ments: coping with problems or tasks with relatively little help: absence of a continual need for affection and reassurance; and a tendency toward mastery for its intrinsic rewards (Heathers. 1955). The overly dependent child is not likely to develop the ability to Cope 26 adequately with frustration. largely because he is exposed to few situations which require coping. Unlike the more self-sufficient boy. he does not develop an adequate system of self-reassurance to support him when things are not go- ing well. At the other extreme is the boy who is too inde- pendent. who does not seek help or emotional support when it is appropriate to do so. Since he learns not to ask for help. he carries the entire burden of coping by himself. Much of the research which has explored the rela- tionship between maternal behavior and attitudes and chil- dren's dependency has focused on the early years of the child's life. In a study of mothers of overdependent six- year olds. Stendler (1954) examined both their techniques of infant rearing and their present behavior toward the children. There were no differences between mothers of overdependent children and mothers of normal controls with respect to permissiveness—strictness in their handling of infant weaning and toilet training. nor was there a differ- ence between the two groups of children regarding early frustration in these situations. Less than a third of the mothers of overdependent children were overprotective. The majority of the overdependent Children needed mother's try-A avecpn U. A eral s OH“. L-1 4 q n v H ‘Eee H t" My .51 y- “W". ~ ~'.“ LL ‘2'”: .i. 92:? 1 he has ch ""5 e n_. I“ I A, t s .7 27 presence but not her help. They had all experienced sev- eral stressful events (e.g.. illness. loss of parent) dur- ing the period from 9 months to 3 years. Their mothers were disturbed at the children's dependency and attempted to force independence. However. they were also inconsis- tent. sometimes giving in to the child's demands and some- times not. The author concludes that there is a critical period in the child's life (9 months to 3 years) during which an abnormal amount of stress may cause the child to turn more and more to his mother and develop a habit of overdependency. The research of Sears et a1. (1957) likewise pro- vides evidence against the psychoanalytical theory that frustrations connected with early weaning and toilet train- ing are the basis of overdependency in the child. A very important finding was that mothers who both punished and rewarded the child's dependency had overdependent children. The authors offered the following interpretation. Depend- ency is sometimes the child's way of expressing fear that he has lost his mother's affection. Mother's reaction to this expression with both punishment and reward serves only to increase the child's uncertainty and produce 28 conflict in him regarding future expressions. Although he anticipates reward. he also is fearful lest he be pun— ished. Thus. he inhibits much of his dependent behavior. But when it is finally expressed. it is likely to be over— determined and intense. and because it is so intense mother is likely to reject the child's overture and thereby per- petuate the conflict. Hilton (1967) studied first-born. later-born. and only children (4-year olds) and their mothers in a struc- tured interaction situation. The child was given a puzzle to work. and mother-child interaction was observed across two different sessions which differed with respect to mother's instructions. During the first session the mother was told that her child was doing well. and during the second one she was informed that he was failing the "test." Several indices of dependent behavior significantly differ— entiated the groups of children. with first-born and only Children running to mother's side contrary to instructions. aSking for direct help. and asking for reassurance to a greater degree than later-born children. Mothers of first- borns were rated as significantly more involved and more interfering than mothers of later-born children. while 29 mothers of both first-born and only Children were more likely to initiate work on the puzzle. to make suggestions. and to disregard instructions and give direct help. While they did not Show inconsistency toward the child within either of the sessions. mothers of first—horns did show a decrease in affection and verbal support in going from suc— cess to failure conditions. How do mother's interference and inconsistency re- late to child dependency? Hilton suggests that interference and inconsistency prevent the child from developing refer- ence points for internal evaluation. If mother reacts in an inconsistent manner. there are no stable guidelines to enable the child to internalize correct standards or ways of proceeding. Prediction on the basis of past experience is thwarted. and the Child must ask anew for reassurance that he is proceeding correctly. The child whose mother is constantly interfering. setting his goals for him. and praising him only when he pleases her. fails to develop a perspective for praising himself. In a study of social deviance in preadolescent boys. Winder and Rau (1962) found parental attitudes toward child- rearing to be significantly related to boys' dependency. 30 The Peer Nomination Inventory (PNI) provided data on bOys' behavior. while the Stanford Parent Questionnaire (SPQ) was used to assess parental attitudes. Dependency in boys was related to maternal attitudes indicative of rejection and sex anxiety. Maternal restrictiveness. deprivation of privileges. punitivenss and physical punishment were asso- ciated with dependency and aggression in boys. The total measure of deviancy in boys was related to ambivalence (high rejection and high demonstrated affection) and low self-esteem in the mother. Rau. Mlodnosky. and Anastasiow (1964) found very similar results with second grade boys. Using ratings of boys' behavior from teachers. parents. and peers. and the SPQ for measurement of mothers' atti— tudes. they found boys' self—sufficiency to be associated with maternal attitudes reflecting high acceptance of the boy. high self-esteem. and high rewarding of independence. Ashievement Motivation Somewhat less circumscribed and clear—cut is the third dimension of boys' behavior. achievement motivation. The boy who is high in achievement motivation is concerned about his performance and tries to do his best. A related 31 variable is "avoidance of failure." which can be defined as a reluctance to try things for fear of being unsuccess— ful at them. All new situations arouse both approach and avoidance tendencies. mixed feelings about the probability of success or failure. The observable outcome of such a conflict. an index of the strength of achievement motiva- tion. would seem to be the boy's task persistence (Feather. 1962). Under conditions of frustration. a boy will persist if his achievement motivation is high. and he will give up if it is insufficient. Achievement motivation and self- sufficiency in the boy undoubtedly are closely related. Studies concerned with the acquisition of achieve- ment motivation give conflicting results. While Sontag and Kagan (1963) found that the period from 6 to 10 years is the crucial time in the child's life for development of n achievement. the results of Winterbottom's research (1958) suggest that the preschool years may be even more important. Since the same parental practices may have dif- ferent "effects" at different ages of the child. the re— sults of studies using children of different ages are not strictly comparable. 3231'. . 8-6 riders a , . . ‘AK‘ wvn 1‘ cfiuoe.t ‘ s 3-7-35 ”#326 ‘ 1 A“. R " I ate: ..e. «I ; ‘ '~ : J i~~‘ “aggrv 4 a: ‘ U; is“, L t..€l¥ S“: v 5?- . “1 am; ‘A.‘ ”tug mfl‘ «‘3‘ ’ HAP I 3") Q [I ‘ . " i “y. 2'. “:9 32 Winterbottom used a story-telling measure of n achievement with eight-year old boys and questioned their mothers about past and present Child—rearing practices. Results indicated that mothers of high n achievers made significantly more early (through age 7) demands on their child for independence and mastery than did mothers of low n achievers. while the latter group made more of these de- mands when the boy was older (8 to 10). These age differ- ences held for maternal restriction. too. with mothers of high n achievers showing more restriction of the boy prior to 8 years and mothers of low n achievers demonstrating more after the boy had reached 8 years. However. mothers of high n achievers made more demands for independence and Inastery than restrictions at the early ages. and mothers (Df low n achievers did not. Mothers of high n achievers were also more likely to give a positive evaluation of tflieir son's abilities. The author concluded that a neces- Sary antecedent condition to adequately learning achieve- ment motives is early training in independent mastery. The findings of Crandall. Preston. and Rabson (1960) with respect to preschool boys and girls are not in agreement with those of Winterbottom. They observed that 33 maternal reactions to both children's help-seeking and support-seeking behaviors were not predictive of children's achievement efforts. However. it should be pointed out that the sample of children included girls as well as boys. and that only dependency behaviors were studied. Mothers' acceptance-rejection was not related to Children's achieve- ment efforts. but direct positive rewards for achievement behaviors did bear a significant relationship to achieve- ment motivation. A study by Rosen and d'Andrade (1959) provides evidence that is consistent with the findings of Crandall et al. They observed preadolescent boys and their parents interacting in a situation where the boy was required to solve a problem and his parents had some of the informa— ‘tion crucial to solution of the problem. Independence ‘training turned out to be less important than achievement ‘training as a factor in the development of n achievement. iParents of boys who were high n achievers tended to have liigher aspirations for him. have higher regard for his competence. set up standards of excellence where there were none. and tended to react to a good performance with warmth and approval and to a bad one with disapproval. 34 Mothers were more involved in their sons' achievement efforts than were fathers. and they tended to dominate the boys to a large degree. The authors concluded that fathers carry the responsibility for independence train- ing. while mothers have a greater influence on boys' achievement efforts. These findings of Rosen and d'Andrade regarding the mothers of high n achievers are very puzzling. It would be easier to View mothers of low n achievers as dominating. intruding. and discouraging of self-reliance. By interfering with her son‘s independent efforts at prob— lem solving. such a mother might prevent him from obtain- ing rewards and gaining confidence in achievement situa- tions. The present study may yield some data on this point. The final study to be reviewed is one which con- Siders all three of the child behavior variables presumed to relate to coping behavior. Watson (1957) studied a large number of normal children from homes that were clearly permissive or clearly strict. Several measures were used to examine the child's personality: teachers' ratings: free play; Rorschach; TAT: and a performance test. 35 The performance test was one which gradually became too difficult. thus permitting observation of the child's reac- tion to frustration. Children from permissive homes were significantly more independent and cooperative than were strictly reared youngsters. while the latter showed sig— nificantly more hostility than the former. Under condi- tions of stress. children of permissive parents demon- strated a moderate degree of persistence compared to the unusually high or unusually low persistence of strictly reared children. Although there were no group differences with respect to emotional responses to frustration. chil— dren of strict parents showed a significantly greater de- terioration in the intellectual quality of their responses. Hypotheses The review of research on the mother—son relation- Ship makes it possible to reconsider the three broad ques- tions with which the present study is concerned and to formulate some relevant hypotheses. These hypotheses aPPly to two groups of young boys (7 to 9 years) and their mOthers. The lows are boys who have been rated low in 36 self—control. self-sufficiency. and achievement motivation. while the highg are boys who have received high ratings on these three variables. The first question asked was what Child variables are involved in the boy's ability to COpe with frustration. The two groups of boys in the present study have been chosen on the'basis of their standing on the variables of self-control. self—sufficiency. and achievement motivation. and their coping behavior is expected to differ in the following ways. I. Boys' reactions to frustration: Low group boys will show more aggressive and regressive reactions to frustration than will high group boys. while high group boys will show more constructive reactions to frustration than will i2! group boys. The two groups will not differ regarding withdrawal and intr0punitive reactions to frustration. II. Boys' behavior with respect to mother: Low group boys will demonstrate more negatively assertive and poorly controlled behavior .. ‘N "a "91' 1 ti. ... 4 “ ‘5 reidt I... beyr‘ Cf I? ‘e 37 (disaffiliation-dominance) and more passive- aggressive and negatively dependent behavior (disaffiliation-submissiveness) than will high group boys. while high_group boys will demon- strate more positively assertive and friendly behavior (affiliation-dominance) and more co- Operative behavior than will low group boys. The second question which this research asks is whether or not the boy's manner of coping with frustration is related to his mother's behavior and attitudes toward him. The research which has been reviewed prompts expecta— tion of the following relationships. III. Mothers' regponses to sons' frustration: Mothers of i2! group boys will give more negative responses and more restrictive help to their son's frustration than will mothers of high group boys. while mothers of high group boys will give more positive responses and more nondirective responses to their son's frustration than will mothers of low group boys. VI. 38 Mothers' behavior with regpect to son: Mothers of low group boys will Show more rejecting be- havior (disaffiliation-control, and. complaint- suspicion) and more dominant and protective be— havior (affiliation-control) than will mothers of high group boys. while mothers of high group boys will show more accepting and democratic behavior (affiliation-autOhomy) than will mothers of low group boys. Mothers' evaluation of sons' ability: Mothers of high group boys will Show a higher evalua— tion of their son's ability (they will predict greater success for their son on the puzzle. and they will choose more difficult proverbs for him to learn) than will mothers of igg group boys. Mothers' attitudes toward child-rearing: Mothers of low group boys will demonstrate attitudes toward their sons that are more in- dicative of high rejection. high demonstrated affection. high sex-anxiety. low self-esteem. 39 high inconsistency. high punitiveness and physical punishment. high pgrental aggression. and high demands for aggression than are the attitudes of mothers of high group boys. Mothers of high group boys will demonstrate attitudes toward their sons that are more in- dicative of acceptance (low rejection ). high seif-esteem. good marital relationship (high positive father-mother relationship). high democracy. and high use of reasoning and‘gghr tingent rewards than are the attitudes of mothers of low group boys. The third question of concern in the present study asks about the way in which the separate behaviors of Inother and son are related. Are there consistent patterns ()f mother-son interaction which serve to maintain the boys' ‘COping behaviors? This is an exploratory question which Will be investigated by examining for the two groups sep— arately the sequential contingencies of mother-son behavior. CHAPTER II METHOD Subjects Subjects were 32 boys between the ages of 7 years and 9 years and the mothers of these boys. The boys were selected from two Lansing. Michigan. elementary schools which provided a sample of boys and mothers that was rep- resentative of the predominantly middle- and lower middle- class population of that community. All second and third grade teachers of Wainwright and Pleasant View schools rated all of the boys in their classrooms on three dimen- sions of boys' behavior: self-control; self-suffigiency: and.achievement motivation. In addition. the teachers Inade a glObal estimate of each boy's intelligence. using 1.0. and achievement test scores to supplement their im- pressions. The teachers had been acquainted with the boys approximately five months prior to making the ratings. A forced-choice technique previously used by Rau. Mlodnosky. and Anastasiow (1964) was employed. whereby the teachers 40 41 were instructed to place their boys on a four-point scale (low; medium-low; medium-high: high) for each of the rated variables so that the distribution of their ratings would approximate a hypothetical distribution with fewer ratings in the extreme categories than in the medium categories. The teachers were also instructed to rate each variable independently of the others. Instructions to the teachers and definitions of the variables may be found in Appendix A-l. 8 Self-control. self-sufficiency. and achievement motivation are defined as global Variables. Such an approach is clearly supported by the finding of Rau et al. (1964) that global variables like "impulse control" are Inore reliable and also apparently more valid than those derived from breaking down the global variables. The ‘teachers in the present study found it relatively easy to tnake judgments based on such variables. Only one out of seventeen teachers deviated markedly from the instructions. and her ratings were discarded. All Of the teachers found it difficult to separate out from the larger group of boys those who are "overly controlled" and "overly self—sufficient.” However. it is unlikely that any 42 of these boys found their way into the high and low ad- justment groups used in the study. Definitions for "appro- priately controlled" and "appropriately self-sufficient” explicitly eliminated them from the high adjustment group. and ratings of “under-controlled" and "dependent" were necessary for inclusion in the low adjustment group. Most teachers likewise had difficulty in distributing their ratings as instructed. so that there was a disprOportion- ate number of boys in the high category for all four var- iables. The relationships of the rated variables to each other and to two indices of socioeconomic status are de- picted by Table 1. The Wainwright sample upon which these data are based consisted of 118 boys. and it was from this sample that the majority (78%) of the final sample of 32 boys was chosen. The very high intercorrelations among the three behavioral variables suggest that with regard to teachers' perceptions of classroom behavior self-control. self-sufficiency. and achievement-motivation represent asPacts of an adjustment syndrome. This finding is consis- tent with that of Rau et al. (1964) who described a general sYndrome of maturity or adjustment in second grade boys 43 TABLE 1. Wainwright Sample: Intercorrelations among Teachers' Ratings of Boys' Classroom Behavior and Two Independent Judg- ments of Socio—Economic Status (SES). (n=1l8) SES Ratings SC SS AM Father's Father's Occupation Education Intelligence .176 .538**** .411**** .068 .128 Self-control .560**** .638**** .137 .079 Self- sufficiency .750**** .181 .124 Achievement motivation .164 .138 Note.-—Pearson product-moment correlations. All tests are two tailed. ****p < .0005 which included impulse control. self-sufficiengy. and mastery. Although the teachers were directed to make independent rat- ings of the four variables. the participation of "halo effects" in this apparently high degree of interrelationship cannot be rUled out. The fact that ratings of intelligence were signif- icantly and positively related to ratings of self-sufficiency and gghievement mohivaiion made it difficult to find high and low adjustment pairs of boys that were matched with respect to intelligence. A most interesting finding is that none of the rated v ‘ 1 h- .C‘A'EI m "PA-u“ f‘Hy W o Q YA! 1‘ ‘N ‘Vcig'fi‘ 'Q r .9 A ‘1‘~|‘ b_‘ 'S' 44 rated variables was significantly related to SES. which is probably due in part to the fairly homogeneous middle- and lower middle—class sample. On the basis of the teachers' ratings. two groups of boys were selected from the larger sample. The iggg were boys who had received below average (1 or 2) ratings on self-control. self—sufficiency. and achievement motiva- high. and the high§_were boys who had received above average (3 or 4) ratings on the same variables. Each boy in one group was matched with a boy from the other group on the following variables: age; grade; intelligence; birth order; number of siblings; mother's education; father's education; and father's occupation. The Hollingshead and Redlich (1964) occupational index of social class was used to classify father's occupation into: professional (1): business or managerial (2); administrative. technical. or clerical (3): Skilled or semi-skilled manual worker (4); unskilled manual Worker (5). A letter1 was then sent to the parents of each boy aSking that mother and son participate in a "study of boys ¥ 1For the full text of the letter sent to the boys' Parents. see Appendix A-2. 45 in a problem solving situation." Enclosed with the letter was a questionnaire for the mother to fill out.which both stimulated and tested her motivation to participate. The percentage of those contacted who agreed to participate was 64% for the iggg and 62% for the highg, The final sample consisted of 16 matched pairs. Since the final sample of boys was selected "pair- wise" according to the qualification of the members as suit— able matches for each other. it cannot be considered repre- sentative of the total pOpulation of middle- and lower middle-class pairs of ABELand high boys. However. follow- ing Holt (1965) it is felt that two things can be done in lieu of a representative sample. One of these is to repli- cate the study with a different sample (this research is a replication of Rowland's 1966 study). and the other is to describe the sample as completely as possible. To this Second end relevant group characteristics for igyg, highg, and the larger Wainwright sample are presented in Table 2. Detailed information for the 16 pairs of Ss and an indica- tion of the high degree of pair-wise matching may be ob- tained by consulting Appendix A-3. QUE: AEUmCCHCA. nvmumw \ 3.»..qu QTK . mevrarumiurfi \ :HV .W, .18 but. .I. th Hz. .9 Viv-9v -IN “IFN'U N,“~HU .h ".U . . . . . r. ... ~::u.u - I ...~.h.~p\.~. In . x . THAN-unshP-Hv 3a~.h fiv-Im‘ «qu‘hh Han -«u-L..u ul . v . . .. . . app-En. lbs-‘4‘ .3n-qlkz Nina u Pv.-l..q Hahn")?! ”HUME nufiv flvdn 46 .naamo humsw on» now mHmEmm unmflnzcamz may mcfloummmn mannaflm>m uoc mum mumnll.wuoz mm.H ae.~ om. Ho.a oo.H mo.H mm. m semen; Isaak om.aa mn.~a Hm.m Ho.m ~m.~ em.~ mo.~ m . a v Ha H mm s mm H em m mm mm mm om mm mm m 30A ma.m mm.~m Hm.aa mm.~H me.m me.m om.a ~0.H om.a nm.~ m 0H e mm on e mm co m mm me me me as on m swam om.~ om.~m oo.ma om.NH ma.m om.m mn.m m>.m Hm.m ma.m m S 0 no H mm .05 .0: .msoo mm . w sumam mnwm m N m Mm n mm m mo m hum z¢ mm Um HWH» . . . . . - “SONG sumo HMOflmoHOAUOm mmcflumm .mumsomma ucm>mamm :0 mamfimm uflmHH3cflm3 Gnu Cam mQSOHG 30A cam nmflm MO GOmHHmmEOU .mOsumHumuomumno msouo . m 39% 47 Comparison of the igy§_and highg_with regard to the matching variables utilized two-tailed tests of the difference between the means of matched groups (Hays. 1963). There was only one significant difference. with the highg_ showing a trend toward higher ratings on intelligence than the iggg_(p < .10). It can be seen from Table 2 that both highg and'ig!§_have higher mean ratings on intelligence than does the larger Wainwright sample. Appendix A-3 indi- cates that most of the pair-wise differences in intelligence occurred when the i2! member received a rating of "3." the high member a rating of "4." When this difference is con- sidered in light of the fact that teachers made many more "4" ratings than would be expected for an "average" sample of boys. it would seem to be of little consequence. To what extent do the boys represent two extreme groups with respect to the independent variables? The highg_ are probably not as "extreme" as their ratings imply because teachers made a disproportionate number of "4" ratings. Likewise. the iggg do not qualify for such labels as "Clinic" or "deviant." None had been referred for psycho— logical help. Their mothers defined both the boys and themselves as "normal." something which became apparent 48 in informal interviews with the mothers. Past experience with a group of boys with poor school adjustment (Rowland. 1966) had suggested that mothers of such boys volunteer for "psychological" experiments in order to get psycho- logical help. The mothers of iggg in the present sample seemed only to want verification that they and their sons were "normal" and the school to recognize that they were "interested" parents. The quite frequently occurring re- mark of higthothers. "Johnny is doing fine. but we are having trouble with his brother." is evidence that high mothers also do not perceive themselves as an extreme group. Procedure .theraction Session Mother-son interaction was observed in two struc- tured situations. The first or frustration session lasted from 30 to 45 minutes. and the second or verbal session t00k 15 minutes and immediately followed the first. Both Sessions were conducted in a room equipped with a ceiling 49 microphone and a one-way observational window. Mother and son sat opposite each other at a table which was situated at right angles to the observation window. The table con- tained only an ashtray. Other chairs situated about the room and carpeting on the floor gave the room a somewhat home-like atmosphere. There was a clock on the wall which faced the boy and a low table covered with toys and books located directly beneath the observation window. The ceil— ing microphone was placed just slightly higher than the seated boy's head. Although this procedure resulted in the boy's awareness that he was being listened to. it was necessary in order to hear him adequately. In no case was the boy aware that he was being observed. Both sessions were tape recorded from the microphone amplifier in the Observation room. Frustration session: The procedures for this ses- Sion were aimed at inducing frustration in the boy while leaving the situation relatively unstructured for the mother. It was hoped that in what manner and to what de- gree she responded to his frustration would only minimally be a function of the instructions. 50 When the mother—son pair arrived for their appoint- ment. the investigator met them in the clinic waiting room. The mother was given some written instructions which she was to read while the investigator took her son to the room and started him on his task. Her instructions were as follows: This is a study of boys with regard to how they deal with a problem-solving situation and why they behave in the situation as they do. You and your son will be participating in two problem-solving situations. designed to be simi- lar to situations that might occur in your own home. During the first session your son will be given a chance to win a prize by solving a puzzle in 30 minutes. Since there is only one correct solution to the puzzle. the time taken to solve it is considered to reflect his ability at solv— ing such puzzles. I will not be present during this session. If your son seems to need infor- mation. help. etc.. then you may give it to him. Feel free to act as you would in your own home under similar circumstances. There is no right or wrong way for parents and children to act in such situations. I am only interested in observ- ing and recording how peOple do behave and in finding out how well your son can solve problems. As with the questionnaire which you filled out. any records of these sessions are kept with your code number and not your name. Before going to the room in which your son is located. please estimate how well you think your son can do on the jigsaw puzzle. and write the estimate down on the blank sheet of paper. Select one of the following alternatives only: high score; somewhat above average score; somewhat below average score; low score. 51 The investigator introduced the Child into the room and asked him to sit at the table. The boy was then shown a box containing several new toys (models. toy guns. toy automobiles. army toys. etc.) and was instructed as follows: I want to see how good you are at solving a puzzle. First. choose one of these toys for your- self. Now. let's put the toy right here (on the table in front of the boy). You can have this (toy) Ii_f_ you can solve a puzzle for me. But you'll have to work fast in order to win the (toy). See this puzzle? It works like this (investigator completes the puzzle in three easy moves). If you put it together right. you get a picture like this one. It's Superman. You put the puzzle together so that it's just like the picture. Here is the one which you will do. and it's all scrambled up. It has to be just like the picture of YOgi Bear. I am going to leave for 30 minutes. but your mother will be here with you. If you have any questions to ask. you'll have to ask them to her. If you hurry real fast and finish the puzzle before I get back. you can keep the prize. The investigator brought the boy's mother to the room and then left the two of them alone while the boy ‘worked the puzzle. The puzzle was the Sliding Squares Puzzle Game. It is a miniature jigsaw puzzle. the solu- tion of which involves the manipulation of sliding mosaic Squares within a nondetachable frame. in such a way that the correct picture is produced. Although the puzzle aPPears to be very easy to solve. it is. in fact. very 52 difficult. The average adult would take at least 30 minutes to complete it. Even more frustrating for the boy is the fact that the course of solving the puzzle generally comes to within one step of a correct solution. a point at which the final step entails an entire reworking of the puzzle to achieve the solution. In most cases it took longer than 30 minutes to solve the puzzle. so the investigator delayed his return another 15 minutes. If after 45 minutes the boy Still had not finished. the investigator returned anyway. told the boy he had done well considering the difficulty of the puzzle. and awarded him the prize. Verbal session: The purpose of the procedures for 'this second session was to stimulate verbal interaction (a large proportion of the rated behavior in the frustration session was nonverbal) under circumstances of minimal stress to the boy. It was realized that the behavior observed in this session might reflect carry-over effects from the pres Viously frustrating situation as well as effects elicited by the verbal session itself. Because it was not of inter- est to determine how mother-son pairs interacted in a verbal PrOblem—solving session per se. no attempt was made to counter-balance the order of the two sessions for the Ss. 53 The verbal session followed the frustration session for all S3. The investigator again gave some written instruc- tions to the mother which read as follows: During this second 30-minute session. which will begin as soon as I return. you are asked to teach the meanings of 3 proverbs to your son. You may do this in any manner which you wish. It is only asked that you try to complete all 3 proverbs and that they be taught by you and learned by your son to your satisfaction. Please choose any 3 proverbs from the list and write down your choices on the blank sheet of paper. The numbers preceding the proverbs indicate the comparative difficulty involved in teaching and/or learning them. For example. those proverbs preceded by a #1 are relatively easy. while those preceded by a #4 are relatively difficult. Using these numbers as a guide. try to select those proverbs which your son. with instruction from you. will be able to master. Since you may have questions regarding the research etc.. let's discuss them after this second session. Whether or not your son is in- cluded in the discussion afterward is a decision which I would prefer to be reached during this second session. The list of proverbs and their "level of diffi- CUlty" may be found in Appendix B-2. The more or less ‘hyPothetical numerical values were attached to the proverbs in order to determine if the two groups of mothers would differ regarding the supposed difficulty of the proverbs WhiCh they chose. It was presumed that the more difficult were the proverbs selected. the higher mother's aspirations 54 for or confidence in her son or herself. Mother and son were also alone for this second session. which was termin- ated by the investigator after 15 minutes- Rating Techniques Two undergraduate psychology majors used two dif- ferent rating schemes to code the mother-son interaction. They were not aware of the research design or the group identity of the Ss. Although it is still possible that some of the Ss gave behavioral clues from which the raters could have inferred some kind of "good vs. bad" identity. this probably did not happen for two reasons. The raters ‘were not aware of the design: and the Ss. themselves. were nOt.aware that they had a group identity. Reactions to frustration: Both raters observed the frustration session and independently coded the boy's reactions to frustration as they occurred. Only those Verbal or nonverbal behaviors of the boy were rated which fell into one of the five categories defined below. --Son Categories-- A - Aggression—~Includes aggressive acts directed to- ward other objects or people. Examples are: hit- ting the table; throwing or banging the puzzle; 55 hostile remarks to mother. or about the in- vestigator or task. Regression--Inc1udes behaviors more appro- priate to a younger age level. that is. to a preschool age child. Examples are: suck— ing his thumb: sitting on mother's lap; cry— ing or whining; making faces. etc.. in the mirror; stroking self; silliness; playing with prize. toys. or microphone. etc. In some cases A and R will overlap. such as when a boy's remarks are both silly and hostile. Rate one or the other depend- ing upon which seems to predominate. Withdrawa1--Includes behaviors which clearly remove the boy from his work on the puzzle. but which are not aggressive or regressive in nature. Essentially. "flight" is the only impulse indulged. Examples are: sitting back away from the puzzle or pushing his chair back without doing anything else right away; trying to leave the room: spend- ing some time looking around the room; stOpping work on the puzzle and humming to himself or ask- ing questions irrelevant to the task. These be- haviors differ from the others in that the uncon- structive activity is not manifested in an impul- sive. poorly controlled. or infantile manner. Intropunitiveness—-Includes behaviors which reflect anger expressed toward the self. Examples are: reddening of the face without following activity; silent crying: stuttering; self-criticism or degra— dation; self-defeating behavior such as deliberately messing up the puzzle. Constructive reactions——Includes behaviors which seem to enable the boy to overcome blocks and con- tinue his efforts toward solution of the puzzle. Examples are: Boy says. "This is a hard puzzle." and renews his efforts: he verbalizes his determ- ination: he reasons himself through a difficult spot; he boasts. reassures himself: when he reaches inscrat me of t 56 an impasse. he laughs at the situation or at the incomplete state of the puzzle; he glances at the prize and then renews his efforts. The raters also independently coded the mother's reaponse. whether it was verbal or nonverbal. to each be- havior of the boy which they had coded under reactions to frustration. These responses of the mother were coded into one of the five categories defined below. --Mother Categories-- 1 - Ignoringe-Mother ignores son's remarks or discom- fort. Turning away. failing to look up. or "tuning out" son are examples. 2 - Negative response--Mother restricts. dominates. is critical of. or is otherwise hostile toward son. Examples: "StOp your crying": M takes the puzzle from son and works on it: "Look what you have done now." 3 - Help--Mother works on the puzzle. instructs son on what to do (direct but nonrestrictive help). makes suggestions. or gives task relevant information. 4 - Nondirective response-—Mother responds in a passive. positive way. She acknowledges or accepts son's remakrs. frustrations. etc.. but she does not at— tempt to direct him or to rush to his rescue. S - Positive response-—Mother responds in a loving. nurturant way. She reassures. encourages. praises. or is affectionate. 57 Because the hgip category for mother's behavior was defined to include all kinds of helping behavior. a separate rating was made by a third observer of mother‘s restrictive help. This rating was a global estimate of the extent to which mother actually worked on the puzzle and was made on a four-point scale ranging from "no such help" (4) to "she did it for him" (1). Reliabilities for reaction to frustration categories for son and mother are presented in Table 3. It should be pointed out that the correlations between the two independent ratings are with respect to the proportion of S's total rated behavior falling in each category. not the frequency counts of such ibehavior. Since rated prOportions in a closed system of categories are being correlated. the reliabilities for in- dividual categories are not entirely independent of each Other. If reliabilities are high for four categories. the reliability must also be high for the fifth category. Aside from this consideration. the reliabilities are cer- tainly sufficiently high to warrant confidence in the ob- jectivity of the present rating scheme. 58 TABLE 3. Reliabilities for Two Independent Raters on Son and hkmher Categories for Reactions to Frustration. (n=20)a Son Totalb r Mother Total r Category Category Aggression (A) 94 .966 Ignoring (1) 326 .944 Regression (R) 73 .798 Negative response (2) 193 .948 Intropuni— tiveness (I) 125 .979 Help (3) 392 .922 Withdrawal (W) 242 .946 Nondirective response (4) 332 .970 Constructive reaction (C) 852 .984 Positive response (5) 143 .806 lens control (A+R) 167 .945 Total negative response (1+2) 519 .976 High control (I+W) 367 .979 Total positive response (4+5) 475 .956 _. Note.--Re1iabilities are Pearson product-moment correlations. Frequency ratings for each category were converted to propor- tions of S's total ratings. Correlations are between "rated" PrOPOrtions across 20 Ss. . aTwenty out of thirtybtwo Ss were observed by both raters. One rater only observed Ss 4—6. 9-12. 23.-25. 28. 31. 32. Number of ratings made by both raters across 20 Ss. L... yup” A? of -ui 0V- 9 "H. ”5;. «I - :ul Eve. ‘ g "pro, v CC '0tt...) I . ‘v. n if.“ .c.....0: o =‘a‘l W ‘- It“); 139 Dersoe awh‘n‘“ A "‘Wb.y., U . . V ' R "Yy‘f‘y ”GAG. “I- ‘. :a“l ac . 1 ‘c’fln ‘ ~~:5_: LO n l'\‘ ‘ Ce ratw o P- 61“ -:\-._4 arid ‘- ~ ‘ A '1 .. I“: “F!- ‘ ‘yv‘0 . n «.8 sh". ‘ ‘ V R H ‘A §." i“ -‘7 LL 1’ Lnev] "« e“ a " w h the ,, 59 Mother-son interaction: The tape recorded verbal be- havior of both the frustration and verbal sessions was rated using the continuous scoring technique and the system of be- havioral categories developed by Freedman. Leary. Ossorio. and Coffey (1951). The Circumplex rating scheme is suffi— ciently comprehensive and flexible to permit coding of all behavior occurring in mother—son interaction into a relatively small number of meaningful categories. Behavior is rated from the perspective of interpersonal mechanism. "the interpersonal function of a unit of social behavior." Therefore. coding of a behavior depends neither on its form of expression nor its Imedium of expression. Although the authors rate the intensity of each act as well as the mechanism involved. the Circumplex seenmsto have a built-in intensity gradation which makes sep- arate ratings of intensity unnecessary. The authors also recommend that the coder empathize with the person being acted.upon and ask himself what that behavior means to him. The two raters in the present study found it very difficult to maintain this empathic set while continuously shifting baCk and forth from mother to son. and because of this diffi- cultY they rated largely from the perspective of observers. Since the mechanics of scoring may differ widely among 60 studies using the same rating scheme. precise delineation of scoring procedure is highly important. One may consult the paper by Mueller and Dilling (1968) for considerations cm'scoring as they apply to the Circumplex rating scheme. In the present study. the basic unit of verbal in- teraction is the meaningful speedh. comprising one or more wordS'which serve an interpersonal function and which are uninterrupted by the other person. The speech may serve more than one function (e.g.. to dominate and to punish). in which case only the predominant function or affect will be scored. A speech may also include more than one state- ment» separated by brief silences. with the additional state— ment(s) either representing a change in function or affect. or a continuation of the affect reflected in the original Statement. This speech comprising a sequence of statements is Still scored for the predominant affect only. A state- ment(s) followed by Several seconds of silence and then another statement by the original speaker presents a special case. If in the rater's opinion. the original statement(s) embodies the expectation of a response from the recipient. the recipient's silence is scored as an "implicit Speech." With the rater necessarily having to infer the function of .I'E :.1 n ". on ":1: s; elk-I 0‘ .5... .. ». st; AY‘R‘ \l. «.5 V.-4 n y :- I‘V‘R :.E:D...: . .": ‘1! C ' e “k u. . é . ”':‘ ”3" "MU-u»- JG gfig‘ we “Vooa. . .0.“ w ‘ ‘ 'V.._: A : nu ._ Q '- ..:'Or‘ no...‘ h.‘ .- ns..‘. ‘ h‘ a PC" ‘."‘-.A\,\‘ n ‘IA “L y... :aLeQ a ‘ s X" “:5. _ k. I “‘s‘. -A tLA y Wau'L‘. be .1- i Q P R . "v, -. p ' . ye‘.a :1" EPA.“ - U...‘ C : a. "a . ‘i n ‘A ‘=\ II: N‘*a} .h.‘ ‘ ~ D“. Q ‘9- C . 5.. as "I“ ““4, ‘ \VHNN ‘u 61 the silence (e.g.. resistance. withdrawal). Since this im— plicit speech has intervened. the original speaker's next statement(s) represents a new speech. If the rater feels the original statement(s) does not embody an expectation of a response from the recipient. the silence is not scored. and the following statement(s) by the original speaker is considered a part of the original speech. It was felt that the interpersonal mechanisms should be defined separately for mother and son. since they have somewhat different meanings and are best illustrated with different examples in the two cases. These definitions are contained in Appendix B-2. In order to facilitate coding. the categories were also arranged on a Circumplex diagram and defined in terms of appropriate verbs. The diagram for Imnflier behavior is presented in Figure l and the diagram for Son behavior in Figure 2. All of the recorded behavior was rated independently by bOth coders. The rating sheet which they used may be f0und.in Appendix B-3. The tape of the frustration session wasalways rated before the verbal session tape. and these ratings occurred on an average of one week after the raters had coded "live" the reactions to frustration. In the cases Demdi muozil (52 Figure l.--Maternal Behavior Authoritarian, Dictatorial B ‘ Possessivo sanding iviiy Do . res t Dominate Antagonistic Bo narcis- Take over he agtho ~ tar an Actively Overpro- D --E 0 tective Roasct scarc- Rofuso sol: deect- i Protec- ing tive E n uPPO Oriticiss. Attack Smother, Protect Indul- Bhow contempt, disust Svmpathize. Pity gent Complain Lovs Massively Identify with r lag Feel with Bulk q.. Cooperct- Praise M Neklcct- O A:rg:ate Accept- Depend Participate in“ Free with Ask gng Accept opinion Express Confidc I. (Admire. need Indifferent respect \\ Cooyerative Question passively /, K Detached __. Democratic I J Freedom Ecto.- Descriptive adjectives outside the circle adopted from Schaefer's Circumplex model for Maternal Behavior (1961). '20:: RebeLl Iefiant , / :55:11. N \. h. I—i Tn: Lfi ( 3 r’”" "I-/'\ i) (I) m n hln _’ A n~ ‘1' : hk5..lc‘llou 63 Figure 2.--Child Behavior Assertive n A Actively Resist Dominate Emotional, Interrupt a s w h . Rsbollious Ba 3011- authority Sociable centered Change . Be solf- subject " stim- Correct Aggiziiz ulatinsJ Direct 9 D Bo rud Order 0 Boast . “88 Work with Do agrsssivo Defiant. Lose tamper Levin: Hostile to . D tilisto 3 Attack N Criticiss Glare Bo igritablo. Be crggi__i m lain Lova- ( ass vs resistance) Express admiration, 1 Obstruct. Foul up appreciaiian N Bo difficult Be affectionat Bulk. Whin- Be Identify -oaso Sus act Coo crate with Distrusf .elp- UoIIaBorato Indirozgiy n‘iloss Ds-snd Confide in on '0 can -s or Accuso -01: Pas- hel Coopisg- G ’ithdraw sively Ask for Back down Submit ucstio. Suspicious Give in s or Responsible Apologize Obey Be afraid Comply Whimpor H Worry Ask Opinion for di. Withdrawn Calm. Compliant I J Submissive Noto.-Desoriptivc adjectives outside of thb circle adopted from Becker's Circumplex Model for Boys' Behavior (1962). 33295 CC the abse between izations tes: ta; EEEt 0n 1 filVAJ “USED ¢ 64 where only one rater observed the actual interaction. his notes concerning nonverbal behavior were made available to the absent rater. Tapes were rated in the same order in which they were recorded. an order which was based on the Ss' preference for appointment times. Tapes were identi— fied only by 85' code number. Although the raters worked toqether so that they would always be rating the same unit (typescripts are costly and often confusing). communication between them was limited to clarifying hard-to-hear verbal- izations and which unit they were on. Prior to rating the test tapes. the coders had to demonstrate complete agree- Iment on a pilot tape as to how the interaction was to be 'broken down into units. Because the sequential contingencies of mother and son behavior are of interest in the present study. per cent agreement reliability was calculated on the basis of unit by unit agreement between raters. One coder was arbitrarily d€8ignated as the base rater. and the other coder's rating was scored according to its degree of deviation from the base rater's. The possible scores were: total agreement (Same category); octant agreement (other category in the 65 octant: A goes with P. O with N. etc.); one-step agree- ment (either of the immediately adjacent categories); and disagreement (category that is not immediately adjacent). Each rated unit was scored in this fashion. the frequency of each score was tallied for the session. and these fre- quencies were divided by the total number of rated units for the per cent agreement reliabilities. In tallying the octant agreement scores. cases of total agreement were naturally included. and cases of both were included in the count of one-step agreement scores. Table 4 shows per cent agreement reliabilities for the three degrees of inter rater agreement. The reliabili- ties for the verbal sessions are slightly higher than those for the frustration sessions. Reliabilities for the total 'behavior 0f.l2!§ and highs are almost identical. Assessment of reliability was not done separately for mothers and sons. If the Circumplex arrangement of categories is taken at face Vlalue (data are lacking on this point). the one-step relia— wlbilities would seem to be the most appropriate ones for the ‘present ratings. The obtained per cent agreement of 90% and.above is regarded as sufficient evidence of the objec- tiVity of the rating. The reliabilities for individual categories can be found in Table 5. 66 v~.~m mm.mm m~.om naom no.mm ¢~.mm m~.~m mnmm mo.am on.mm mm.nn ovom Hobos oa.~m mo.mm om.om doom Ho.mm mm.vm mm.am mmoa 0H.Hm nm.mm mm.mh Hmm mnmflm om.mm ¢~.¢m hm.om 05mm Ha.mm ~¢.mm mv.~w nmma om.om ma.mm mv.on mmoa mBOA mua poo as z mla poo <9 2 mid ace «9 z msouo Hmuoa Hmfluo> codumnumsnm ucmfimmum< ucmo mom Amlav moumlmco Ucm .Auoov ucmuoo .Aflev acmEmmHm< Hmuoe .mumumm sconcmmmch 038 How mowuwaflnmflamm .v mamas - .: r TABLE 5. 67 Total Agreement (TA). Octant (Oct). and One-Step (l-S) Per Cent Agreement Reliabilities for the Sixteen Circum- plex Categories . W Category N TA Oct l-S B - Active resistance 110 g 60.90 61.81 79.09 C'— Competition 10 80.00 80.00 90.00 D - Punishment (Antagonism) 6 83.33 83.33 83.33 E - Hate 0 -- —- -- F - 'Complaint 84 78.57 78.57 79.76 C - Suspicion 7 . 71.42 71.42 85.71 H - Helplessness 538 89.77 90.14 91.26 I - Submission 32 65.62 68.75 68.75 J - Democracy (Admiration) 180 78.33 85.55 85.55 1<- Dependency 108 41.66 75.00 78.70 3L- COOperation 325 89.84 90.46 94.76 M - ' Love 5 100 .00 100 .00 100 .00 N - Reassurance 149 86.57 89.93 89.93 0 - Holp 1705 84.16 87.97 95.07 P - Structuring & Teaching 1518 77.14 80.76 92.02 A - Dominance 208 57.69 77.88 88.46 68 Stanford Parent Questionnaire Mothers' attitudes toward child—rearing were assessed with an abbreviated version of the MSU Form of the Stanford Parent Questionnaire (SPQ). The SPQ was de- signed by Winder and Rau (1962) for the purpose of assess- ing parental attitudes believed to be associated with so- cial adjustment in preadolescent boys. Sixteen of the scales. comprising a total of 208 items. were used in the present study (see Appendix C-l). The internal consistency and test-retest reliabilities for the 16 scales may be found in Appendix C-2. The questionnaires were mailed to the mothers when their c00peration was initially sought. and the mothers re- turned the completed questionnaires when they arrived to participate in the study. The questionnaire was introduced to each mother as a series of statements made by other par- ents about their children. For each item. the mother was asked to indicate her own attitude toward the issue as it applied to her son. by checking Strongly agree. Agree. Disagree. or Strongly disagree. In a few cases. the mother had difficulty answering some of the items. When this 69 situation arose. the investigator clarified the statements for her and urged her to answer them. The "continuous scoring" method was used such that when agreement with a statement indicated presence of the characteristic denoted by the scale. Strongly agree was scored 3. Agree was scored 2. Disagree was scored 1. and Strongly disagree was scored 0. If disagreement with the statement indicated presence of the characteristic. the scoring was reversed. The sum of the scores on each of the individual items of the scale was the score for that scale. CHAPTER III RESULTS Boys' Behavior Two major hypotheses dealt with the question of how boys who are low in self—control. self-sufficiency. and achievement motivation differ in their behavior from boys who are high on these three variables. One of these hypotheses concerns only boys' reactions to frustration. while the other applies to all of the boys' verbal inter— personal behavior manifest in the frustration and verbal sessions. Hypothesis I: Reactions to Frustration This hypothesis predicted that low group boys would show more aggressive and regressive reactions to frustration than would high group boys. while high group boys would show more constructive reactions to frustration than would low group boys. The two groups were not expected 70 71 to differ regarding withdrawal and intr0punitive reactions to frustration. The boy's "score" for each of the five categories of reaction to frustration was derived by counting the number of times he had been rated in a particular category and dividing this count by the total number of ratings for him across all five categories. Using the differences be- tween these prOportions for matched pairs. the Wilcoxon Matched-pairs Signed—ranks Test (Siegel. 1956) was employed to compare the lgg_and high groups. This test takes into account the magnitude as well as the direction of differ— ences between pairs and does not require assumption that the sampled population is normally distributed. Results of the group comparisons are presented in Table 6. All of the predictions contained in the first hy- pothesis were borne out. The lg! boys reacted to frustra- tion with poorly controlled. aggressive and regressive be- haviors significantly more often than did higthoys (p < .03). while high_boys demonstrated a significantly . . 1 greater amount of constructive reactions (p < .06). The ¥ 1A departure was made from the standard procedure of reporting as significant those results with a probabil- ity of less than .05. Differences with p < .10 are con- Sidered significant in this paper. 72 TABLE 6. Comparison of Low and High Boys on Their Reactions to Frustration. ==== .4: q====fi====== Category iLoa 2H1. Tb NC Pd Aggression (A) .079 .021 67 12 .04 Regression (R) .084 .032 76 13 .04 Intropunitiveness (I) .103 .097 69 16 NS Withdrawal (W) .257 .201 94 16 NS Constructive reaction (C) .477 .649 105 16 .06 Low control (A+R) .162 .053 79 13 .03 High control (I+W) .360 .298 87 16 NS TOtal reactions to frustration (A+R+w+‘1+c)e 67.10 62.20 62.5 15 NS Efiazzle solution timef 40.75 38.67 27 9 NS aMean proportion of total reactions to frustration. Sum of the positive ranks or those ranks for differences in favor of the group with the greater group mean (Wilcoxon MEltchedupairs Signed-ranks Test). cPairs whose difference = 0 are dropped. thereby reducing n.‘ P values are two-tailed. eNumber of ratings made by both raters. When only one rater ‘31>served. his total was doubled. Solution time is in minutes. If S took more than 45'. his s<=c>re was 45. 73 two groups of boys did not differ with regard to intrqpuni— Eiyg and withdrawal reactions. nor did they differ with re— spect to the total number of rated reactions. Although the 123 group demonstrated fewer reactions indicative of low control (i = .162) than reactions reflect- ing high control (i = .360). this result should be evaluated within the context of the total situation to which the boys were exposed. They were brought to a strange place. sup- posedly to be tested. and were spied on by a stranger. One would expect these boys to be something less than entirely free and spontaneous in their behavior. The finding that nearly 50% (x = .477) of the 123 boys' reactions were 223- structive in nature should not be viewed as indicating that these boys as well as the high_boys characteristically react to frustration with constructive activity. The present study compares groups of boys. and any interpretations con- cerning what is or is not characteristic of either group ‘would require wider sampling of situations. For all of the boys the amount of rated constructive behavior was partly an artifact of the rating method. Whenever a boy withdrew from active involvement with the puzzle. his eventual re- turn automatically assured him of a "constructive" rating. 74 Since all boys at least occasionally reacted with githy drawal. the constructive activity scores are somewhat in- flated. Time taken to solve the puzzle was recorded mainly for the purpose of determining. in a general way. how dif- ficult the puzzle was for the boys. Valid comparison of the times for the two groups is not possible because there were some cases in which mothers completed the puzzle for their sons. Most of these cases occurred in the lgg_group. The average solution times for both groups were almost a full 10 minutes beyond the allotted time. suggesting that the "average" boy was indeed thwarted (frustrated) in his attempt to reach the goal. .Hypothesis II: Sons' Behavior toward Mother This hypothesis specified that 122 group boys would demonstrate more negatively assertive and poorly controlled behavior (disaffiliation-dominance) and more passive-aggressive and negatively dependent behavior (gig- affiliation-submissiveness) than would high group boys. while high group boys would demonstrate more positively 75 assertive behavior (affiliation-dominance) and more coOper- ative behavior than would low group boys. Datayrelevant to this hypothesis were ratings of the tape-recorded. verbal behavior of the boys in both the frustration and verbal sessions. The rated frequencies for each of the 16 behavioral categories of the circumplex were converted to proportions of S's total rated behavior. In most cases the proportions for single categories were not sufficiently large for tests of group differences. However. none of the hypotheses dealt with single category compari- sons. Single category prOportions were combined for circum- plex octants and combined again for the circumplex quadrants. Although raters were trained to differentiate adjacent cate- gories and to code behavior in terms of individual categories rather than large divisions of the circumplex. reference to the category definitions (Appendix B-2) will indicate that larger sections of the circumplex are meaningful and appro- priate. Results of group comparisons for the major sections of the circumplex may be found in Table 7. Tests were con- ducted for the frustration and verbal sessions separately and for the two sessions combined. The prediction that low TABLE 7 . 76 Comparison of Low Boys with High Boys on the Propor- tions of their Behavior in the Major Sections of the Circumplex. —. —; Section Session ‘xlo {xHi Ta Nb Pc Ifisaffiliation (B-I) T .378 .211 109 15 .006 F .383 .154 82 13 .02 V .389 .239 105 15 .02 Submissiveness (F-M) T .511 .379 95 15 .06 F .490 .340 69.5 13 NS V .517 .379 87 15 NS Ifisaffiliation- T .058 .025 69.5 12 .02 Dominance (B—E) F .102 .021 59 ll .03 V .035 .027 18.5 8 NS Ifisaffiliation- T .320 .185 104 15 .02 Submissiveness (F-I) F .281 .133 77.5 13 .03 V .353 .211 101 15 .03 Affiliation- T .191 .194 67 15 NS Submissiveness (J-M) F .209 .207 40.5 13 NS V .163 .167 72.5 15 NS Affiliation— T .441 .591 102.5 15 .02 I)ominance (N-A) F .417 .646 - 77 13 .04 V .443 .603 96.5 15 .05 \ fii liote.--Because of recording failures. the F session data for “’0 83 had to be thrown out. reducing the # of pairs to 13 for F session comparisons. 3Sum of the positive ranks (Wilcoxon Matched-pairs Signed- ramks Test) . cPairs whose difference : 0 are drOpped, thereby reducing n. P values are two-tailed. by mi se- 1‘- 77 boys would exhibit more negatively assertive and poorly controlled behavior (disaffiliation-dominance) was con- firmed for the frustration session (p < .03). and for both sessions taken together (p < .02). but not for the verbal session alone. The remaining predictions concerning cir- cumplex quadrants were confirmed for both frustration and verbal sessions. ‘hgy_boys exhibited a significantly greater amount of passive-aggressive and negatively dependent be— havior (disaffiliation-submissiveness; p < .02). while high boys showed significantly more positively assertive and friendly behavior (affiliation-dominance; p < .02). No predictions were made for the affiliative—submissive quadrant since it includes both positively dependent and cooperative behavior. A further combination of quadrants into the two circumplex hemispheres revealed that igy_boys were signifi- cantly more disaffiliative (p < .006) and submissive (p < .06) in their total behavior for the two sessions than were high. boys. In examining the mean proportions for the two groups in the various quadrants and hemispheres. it should be kept in mind that the categories were defined separately for 78 mothers and for sons. It is undoubtedly because of this redefining of son categories that boys in the present study showed a relatively greater amount of affiliative-dominant behavior and a relatively lesser amount of affiliative- submissive behavior than did the preadolescent boys in the research of Rausch et a1. (1960) and MacKenzie (1968). Group comparisons for the circumplex octants are presented in Table 8. The number of significant differences and the levels of significance are lower for the octant com- parisons than they were for the quadrant comparisons. prob- ably because division of boys' behavior into eight parts makes for relatively small pr0portions. The prediction that ig! group boys would exhibit a greater amount of inter- .personal behavior that was passive—aggressive (complaint— _§gspicion) and negatively dependent (helplessness-submission) \was confirmed for the verbal session (p < .10). and for both isessions taken together (p < .10). but not for the frustra- ;E§£h_session. The expectation that high group boys would be more cooperative than i_o__w_ group boys was not borne out. Tmhe difference was in the right direction. but it was not 'Significant. Although no prediction was made concerning Positively dependent behavior (admiration-dependengy). the 79 finding that igg group boys did not show a significantly greater amount of this behavior should be considered in relation to the findings for helplessness-submission. Thus. the boys who were selected for their low self-sufficiency were significantly more dependent ghiy with respect to de- pendent behaviors that were negative and indirect. While the quadrant comparisons indicated a greater amount of affiliative-dominant behavior for high group boys regardless of the session. a different pattern of findings emerges from the octant comparisons. hggh_group boys showed significantly more behavior indicative of hgih only for the jgustration session (p < .02). Although they demonstrated more structuring-dominance than igg. group boys. this dif- ference was significant only for the verbal session (p < .03) and for both sessions together (p < .06). hgyg and reassur- fighgg were not included with their respective octants due to their relatively infrequent occurrence. Only the high group Ixoys displayed behaviors scored in these two categories. The fact that group differences on some of the be- lmavioral variables depended upon which of the two sessions ‘Was being considered suggests that some changes occurred in the behavior of the two groups from the frustration to o“ .n 80 TABLE 8. Comparison of Low Boys with High Boys on the Pr0por- tion of their Behavior in the Circumplex Octants. w======fi==== Octant Session XLo xHi Ta Nb Pc Complaint- Suspicion (F-G) T .055 .021 80 14 .10 F .054 .025 21.5 8 NS V .051 .015 72.5 13 107 Helplessness- Submission (H—I) T .265 .164 63 12 .07 F .227 .108 69.5 13 NS V .306 .197 90 15 .10 Admiration- Dependency (J-K) T .109 .092 56 15 NS F .138 .108 43.5 13 NS V .073 .063 68.5 14 NS Cooperation (L)d T .083 .101 75 15 NS F .071 .095 63.5 13 NS V .090 .105 71.5 15 NS Help (0) e 'r .319 .405 74.5 14 NS F .239 .428 70 12 .02 V .360 .423 79 15 NS Teaching+Structuring- Dominance (P-A) T .114 .187 94.5 15 .06 F .176 .213 59.5 13 NS V .087 .179 89 14 .03 Note.--The breakdown of dissaffiliation-dominance (B-E) into its octants is not presented because of its relatively infre- quent occurrence. aSum of the positive ranks (Wilcoxon.Matched-pairs Signed- ranks Test). bPairs whose difference = 0 are drOpped. thereby reducing n. °P values are two-tailed. ' eLove (M) occurred only rarely. Reassurance (N) occurred only rarely. 81 the verbal session. Comparisons of the boys' behavior in the two sessions may be found in Appendix D-l. The behavior of the high group boys was highly consistent across the two sessions-~there were no significant changes for this group in any of the behavioral octants. Significant changes did occur for the i9! group boys. however. the most important of which were the decrease in negatively assertive and poorly controlled behavior (disaffiiiation-dominance: p < .06) and the increase in negatively dependent behavior (helplessness-submission: p < .04). These changes will receive consideration in a subsequent section. when they can be compared to changes in mothers' behavior. One should not interpret these results as indicating that high group boys are more consistent in their behavior or less influenced by situational factors than are igg group boys. Although it is an interesting proposition. several situa— tions would have to be sampled to permit a valid test. Mothers' Behavior There were three hypotheses concerned with maternal behavior. One dealt with mothers' responses to sons' 82 frustration. and it applied to ghiy that behavior of the mothers which directly followed (as a "response") rated reactions to frustration of the sons. All of the mothers' tape-recorded (verbal) behavior for frustration and verbal sessions served as data for another hypothesis. which dealt with mothers' behavior toward their sons. The data for these two hypotheses were analyzed in the same way that sons' data were. Frequency counts were made for the var- ious categories of behavior. these were converted to pro- portions of the total behavior. and groups were compared using Wilcoxon's Matched-pairs Signed-ranks Test. A final hypothesis for mothers' behavior concerned their evaluations of the sons' abilities. Data relevant to this hypothesis were derived from mothers' predictions of sons' success on the puzzle and mothers' choice of difficult or easy proverbs for their sons to learn. Hypothesis III: Responses to Sons' Frustration It was predicted that mothers of low group boys would more often react to their son's frustration with flggative responses and restrictive heip than would mothers 83 of high group boys. while mothers of high group boys would more often react with positive responses and nondirective reSponses than would mothers of i2! group boys. Results of the group comparisons for responses to frustration are presented in Table 9. and it is apparent that they are very much in agreement with hypothesis III. Mothers of igg group boys showed significantly more 2233? tive responses (p < .03) and restrictive help (p < .008) than mothers of high group boys. and the latter exhibited significantly more nondirective responses (p < .02) and positive responses (p < .09) than the former. It should be noted that the restrictive help rating was made by a third observer (the investigator) and was a global rating on a four-point scale of the extent to which mothers di- rectly worked on the puzzle. This additional rating was necessary'because of the heterogeneous nature of the hgip category. which included both indirect and direct kinds of helping behavior. The findings indicate. then. that mothers 0f i2! group boys responded to their sons' frustrations in a negative and restrictive. intrusive manner much more often than did the mothers of high group boys. Mothers 0f m. on the other hand. were much more inclined to TABLE 9. Comparison of Mothers of Low Boys with Mothers of High Boys on their Responses to Sons' Frustration. Category 3 a iHia b c d LO T N P Ignoring (1) .286 .257 70 15 NS Negative response (2) .210 .083 111 16 .03 Help (3) .275 .219 88 16 NS Nondirective response (4) .150 .299 115 16 .02 Positive response (5) .081 .142 91 15 .09 Total negative response (1+2) .495 .340 112 16 .03 Total positive response (4+5) .230 .441 123 16 .005 Restrictive helpe 2.13 3.25 84 13 .008 aMean proportion of mother's total responses. bSum of the positive ranks or those ranks for differences in favor of the group with the greater group mean (Wilcoxon Matdhed-pairs Signed-ranks Test). CPairs whose difference = 0 are dropped. thereby reducing n. dP values are two-tailed. eRated independently by a third observer. gldbal rating on a 4 pt. scale (1 is high. 4 is low). Refers to a . w..- I... "w "a “.1.“- Fl 85 respond with positive. supportive behaviors or in a manner which was minimally interfering and restrictive. Hypothesis IV: Mothers' Behavior toward Sons This hypothesis specified that mothers of igngroup boys would demonstrate more rejecting behavior (disaffilia- tion-control. and complaint-suspicion) and more dominant. protective behavior (affiliation-control) than would mothers 0f.hlflb group boys. while mothers of high group boys would demonstrate more accepting and democratic behavior (affilia- tion-autonomy) than would mothers of igy group boys. The results of comparisons of mothers' behavior for the major sections of the circumplex are indicated in Table 10. As predicted. mothers of iggg significantly exceeded mothers of high§_with regard to disaffiliation—control (p < .02) and affiliation-control (p < .03). and the mothers of‘highg demonstrated a greater amount of behavior indica- tive of affiliation-autonomy (p < .006). With the exception of the disaffiliation-control quadrant. these differences held for both frustration and verbal sessions. Since only 2%.of the total behavior of low mothers was scored in the TABLE 10 . 86 Comparison of Mothers of Lows with Mothers of Highs on the Proportions of Maternal Behavior in the Major Sections of the Circumplex. m ——== a Section Session Lo xHi Ta Nb Pc Disaffiliation (B-I) T .045 .054 59 13 NS F .047 .067 48 11 NS V .038 .043 23 10 NS Autonomy (F-M) T .067 .167 83 13 .01 F .096 .217 68 12 .03 V .042 .133 103.5 15 .02 Disaffiliation- Control (B-E) T .019 .003 53 10 .02 pd __ _- -_ _- -- Ve .025 .004 -- -- -- Disaffiliation— Autonomy (F—I) T .025 .050 83.5 14 .06 F .041 .065 51.5 11 NS V .013 .039 23 8 NS Affiliation- .Autonomy (J-M) T .041 .113 109.5 15 .006 F- .061 .131 67.5 12 .03 V .029 .094 89.5 14 .03 Affiliation- <30ntrol (N—A) T .903 .812 89.5 14 .03 F .874 .779 76.5 13 .04 V .935 .855 101.5 15 .03 \ N Highs in all. 87 disaffiliation-control quadrant. their demonstration of a larger amount of these behaviors. while representing a sig- nificant difference. should be regarded only as a trend. When all of the disaffiliative behaviors are considered to— gether (disaffiliation). the two groups do not significantly differ. These behaviors represent only about 5% of mothers' behaviors toward their sons. Consideration of all the au- tonomy granting behaviors together (autonomy) indicated that mothers of highg showed significantly more of these behaviors than did mothers of iggg for both of the experi- mental sessions. The finding that high mothers significantly exceeded mothers of lows with regard to disaffiiiation-autonomy (p < .06) was unexpected. When this quadrant was broken down into its octants. only one group difference emerged. ’thhers of highg showed significantly more helplessness- .Submission (p < .07) for both sessions taken together. Mk’thers of‘igy§_demonstrated more behaviors scored as Sfiflflplaint-sugpicion. but these behaviors did not occur frequently enough to permit statistical analysis. Thus. it: is the passive more than the negative aspects of giggi- -§3£Liation—autonomy which account for the high mothers' larger scores in this quadrant. 88 The greatest proportion of maternal behavior was scored in the affiliation-control quadrant. Although this distribution of mothers' behavior may indicate that mothers were very inhibited in the experimental sessions and were responding largely as mothers "should" (the test-like nature of the sessions might be expected to "pull" this type of maternal behavior). the possibility also exists that wider sampling of situations might yield much the same results. While MacKenzie's (1968) mothers showed less behavior indi- cative of affiliation-control than did the present sample of mothers. there is only one valid comparison between the subjects of the two studies. Her mothers of normal 7-11 year old boys were rated in the affiliation-control quadrant 62% of the time as compared to 78% for the high mothers in the frustration session of the present study. The fact that some of the boys in MacKenzie's sample were older than those used in the present study may account for some of this dif- ference. since Raush et a1. (1960) found that adults show a decrease in amount of affiliation-control as boys grow older. However. the magnitude of the difference would seem to sug- gest that the particular method of scoring interpersonal mechanisms has some influence on the findings of interaction 89 research. Studies are not strictly comparable unless they use identical methods. The results of group comparisons for the individual categories of affiliation-control are presented in Table 11. None of the individual category differences were significant except with respect to dominance. with mothers of'iggg demon- strating greater dominance (p < .07) of their sons than mothers of highs. This finding suggests that it is the controlling rather than the affiliative aspects of affiliation-control which account for the i2! mothers' higher scores on this quadrant. Several changes occurred in mothers' behavior from the frustration to the verbal session. The results of Com- parisons of mothers' behavior between the two sessions may be found in Appendix D-2. Both groups of mothers demon— strated a significant decrease in reassurance (p < .004) and a significant increase in structuringyand teachihg (p < .002). which is what one would expect from the nature of the two different situations. As was the case with the boys' groups. the other significant changes occurred ex- clusively in the iggg' behavior. Mothers of iggg showed a decrease in disaffiliation—autonomy (p < .02). affiliation- autonomy (p < .03). and all passive behaviors taken as a TABLE 11. 90 Comparison of Mothers of Lows with Mothers of Highs on the Proportions of Maternal Behavior in the subcategories of Affiliation—Control. M fi§mm Category Session xLo xHi Ta Nb Pc Reassurance (N) T .059 .080 83 15 NS F .136 .165 41.5 12 NS V .022 .021 17.5 8 NS Help (0) T .311 .319 63.5 15 NS F .381 .317 63 13 NS V .273 .289 67.5 15 NS Structuring & Teaching (P) T .456 .383 86 15 NS F .286 .248 46 12 NS V .561 .511 73.5 15 NS Dominance (A) T ..093 .050 82 14 .07 F .093 .059 40 11 NS V .084 .041 86.5 14 .04 aSum of the positive ranks (Wilcoxon Matched—pairs Signed-ranks Test). bPairs whose difference = 0 are dropped. thereby reducing n. cP values are two-tailed. whole (autonomy: p < .003). and these mothers demonstrated an increase in affiliation-control (p < .02). These changes cor- respond in complementary fashion to those demonstrated by low boys. who showed a decrease in dominant behavior and an in- crease in passive. negatively dependent behavior. More will be said about the relationship between mother and son behavior when the exploratory questions are discussed. '91 One additional question concerning the behavioral ratings for sons and mothers stems from the use of two sys- tems of behavioral categories in the present study. To what extent ék>they yield a consistent picture of behavior? Some of the categories in one of these systems do not have paral— lels in the other system. However. several categories that did seem to parallel each other were examined to see if they were really related. It should be noted that the two systems were not applied to exactly the same behavior. "Reaction to frustration" categories were used to code the largely non— verbal behavior in the frustration session while "circumplex" categories dealt exclusively with verbal behavior. ‘hgg.and high_Ss were combined into one group. and the relationship between their rated behavior for the two systems was assessed for the frgstration session by computing Spearman rank corre- lation coefficients (Edwards. 1961) between pairs of ratings. These correlations may be found in Table 12. With the excep- tion of disaffiliation-submissiveness and the intrgpunitive and withdrawal reactions to frustration. the categories se- lected for their apparent relationship to each other are. in fact. significantly correlated. 92 TABLE 12. Spearman Rank Correlation Coefficients Between Cir— cumplex and Reaction to Frustration Categories for the Frustra- tion Session Ratings. (n = 28) Mothers Circumplex Category Reaction to Frustration rs Affiliation-Autonomy (J—M) Nondirective response (4) .460** Help (0) Help (3) .668** E; ____________________________________________________________ Sons Disaffiliation- Dominance (B—E) Aggression (A) .714** Regression (R) .578** Disaffiliation- Submissiveness (F-I) Intropunitiveness (I) .380* Withdrawal (W) .157 Affiliation—Dominance (N—A) Constructive reaction (C) .597** Note.—-Coefficients are corrected for ties. Tests are one— tailed. *p < .05 **p < .01 Hypothesis V: Mothers' Evaluation of Sons' Ability It was predicted that mothers of high group boys would show a higher evaluation of their son's ability (they would predict greater success for their son on the puzzle. 93 and they would choose more difficult proverbs for him to learn) than would mothers of 12E group boys. Mother's prediction of her boy's success in solving the puzzle turned out to be a non-discriminating measure. She was asked to choose among four alternatives: high score; somewhat above average score: somewhat below average score: low score. Nearly all mothers chose the second alternative. "somewhat above average score." The second measure of mother's evaluation of her son's ability was the difficulty of the pro- verbs she chose for him to learn. The level of difficulty (1— easy through 4-difficult) was totaled for the three proverbs she selected. and these scores for the two groups of mothers were compared by means of the Wilcoxon Matched—pairs Signed- ranks Test. There was no significant group difference on this measure. Hypothesis V is clearly not supported by the evidence. Mothers' Attitudes Hypothesis VI dealt with mothers' attitudes toward child-rearing. It predicted that mothers of low group boys would demonstrate attitudes toward their sons that were more 94 indicative of high rejection. high demonstrated affection. high sex anxiety. low self-esteem. high inconsistency. high pmnitiveness and physical punishment. high pgrental aggression. and high demands for aggression than were the attitudes of mothers of high group boys. high group mothers were expected to demonstrate attitudes more indicative of acceptance (low rejection). high self-esteem. good marital relationship (high positive father-mother relationship). high democragy. and high use of reasoning and contingent rewards. The mothers' scores on the 16 scales of the Stanford Parent Questionnaire (SPQ) were compared for high_and igg. groups by using two-tailed t tests of the differences between the means of matched groups (Hays. 1963). The results of the group comparisons are presented in Table 13. There were 7 scales on which the igy_mothers were ex- pected to obtain higher scores than the high mothers. and 5 of these scales significantly differentiated the groups in the Luedicted direction. Mothers of igyg scored significantly higher on rejection (p < .10). affection demonstrated (p < .10). figmands for aggression (p < .10). ppnitiveness andyphysical .gnushment (p < .01). and inconsistengy (p < .05) than did uwthers of highs. Of the scales on which higher scores were TABLE 13. Comparison of Mothers of Lows with Mothers of Highs on Sixteen SPQ Scales. = l4)a Scale filo 2H1 t Affection demonstrated (Ad) 32.57 29.00 2.02* Rejection (R) 23.21 17.36 2.00* Self-esteem (E) 17.64 20.71 1.71 Rewarding independence (Ri) 28.71 27.43 .94 Achievement standards (As) 24.43 22.07 1.80* Contingent reward (Cr) 15.36 14.00 1.74 Restrictiveness (T) 22.71 21.21 1.07 Sex anxiety (Sa) 21.07 19.57 1.15 Reasoning (Rg) 26.64 26.14 .34 Demands for aggression (Da) 41.71 35.29 1.97* Parental aggression (PA) 17.14 17.00 .09 Democracy (Dem) 22.93 20.79 2.35** Demands for conformity (DC) 22.79 18.92 2.70** Punitiveness and physical punishment (PP) 17.07 12.57 3.95*** Positive father—mother relationship (FM) 48.64 51.14 .59 Inconsistency (In) 15.79 12.64 2.74** Note.--t tests are for paired observations aOne mother in each group failed to return the number of pairs by two. *p < .10 **p < .05 ***p < .01 and are two-tailed. her Q. reducing 96 expected for high mothers. only democracy (p < ;05) signifi— cantly differentiated the groups. with the difference being in the direction opposite to prediction. Self-esteem. posi- tive father-mother relationship. and sex-anxiety were all in the predicted direction but not significant. No predic— tions were made with regard to achievement standards and demands for conformipy. However. these scales did differ- entiate the two groups of mothers. with mothers of igyg demonstrating significantly higher scores on both. In an attempt to clarify the meaning of these mixed findings the relationships among all 16 scales were explored by computing their intercorrelations for each group separ- ately and for both groups combined. The intercorrelations for igyg and high§_combined may be found in Table 14. An inspection of the intercorrelations suggests that there are three relatively independent groups of scales. The scales within each group have high correlations with each other and relatively low correlations with scales outside the group.1 One group is depicted in the upper left corner of 1It was felt that the small and unrepresentative nature of the sample would not warrant a factor analysis or some other. more sophisticated exploration of the inter— correlations among the scales. 97 90‘ d Iuogu «00. V A_eeee moo. v Gees do. v lee no. V A. One eoeou -oazut 20.. .13. 1.8.- 2.0.- cocci-door. ocean-sou uno.u can. one.. no". eeoco>auouuueon rod. 000. ~v~.u cad. cusses assocaucou ooo.u an“. 000. ne~.u veg-nuecallc coavoeuu‘ can. «O~.| e~.n.| enov. cede-luove no. eve-lea Len. ana.- .44.- own. roan-ouoo- season's 28.- o3. 1.3.- one». 335.- use usilneacan ~e04e>ca unn. uo~.u eo~e.n .000... one oeoce>4oecaa Illlllll eeevOn.| eeunv.n eeeehhn. >0¢Ooeae600c~ manecOAosaeu manual eeeuov. eeeeo‘n.| thenulu I>daaeos eeeeulh . I 'OUIOIU den odeuu I E v ecu-on a h 0 o u ICOdUIdONhOOHOUp—u . v a n d I 4 h 98 Table 14 and is comprised of high self-esteem. high positive father-mother relationship. low rejection. and low inconsis— tency. This group of scales is very similar to the parental adjustment factor of Winder and Rau (1962) and to the rejec- high factor of Rau et a1. (1964). A second and very large group of scales is isolated in the lower right corner of Table 14. This group contains a mixture of scales which seem to reflect "parenting" from the perspective of technique and will be given the label. parental involvement. The high intercorrelations among these various "technique" scales sug- gest that the present middle- and lower middle-class sample of mothers does not differentiate strict from democratic parenting. High punitiveness and physical punishment. high demands for gggression. and high sex anxiety constitute a third group of scales which shows some moderate relation- ships to the other two. Parental anxiety about closeness may adequately describe this group of scales. The results of the intercorrelations among SPQ scales for each of the groups separately suggests that both igy and high mothers polarized the scales into good vs. bad parent. The fact that there were 51 significant (p < .05. two-tailed) correlations among the scales for the high group 99 probably indicates that these mothers belong to a highly homogeneous. narrow subculture in which there is high agree- ment regarding what constitutes a good parent. Considerably fewer significant correlations among SPQ scales were found for the igg group. Perhaps these mothers are less consis— tently in agreement with respect to what attitudes are appro- priate to the good parent. The findings that mothers of igy group boys demon- strated attitudes more indicative of parental maladjustment (high rejection and high inconsistency) and of parental anxiety about closeness (high punitiveness and physical punishment and high demands for aggression) are consistent with prediction. But how can their higher scores on parental involvement (high affection demonstrated. high democracy. high demands for conformipy. and high achievement standards) be explained? The results of a previous study (Rowland. 1966) using several of the same SPQ scales may provide an answer to this question. The boys represented more extreme groups with respect to school adjustment than do the boys in the present study. If the mothers also represented groups that were more extreme (re: child-rearing attitudes). then one would expect the following ordering of the groups (from 100 low to high) on parental maladjustment (high rejection. low self-esteem): mothers of well-adjusted boys (WA); mothers of high group boys; mothers of low group boys; and mothers of poorly adjusted boys (PA). Reference to Table 15 indi- cates that such an ordering exists for the rejection scale and with one exception. for the self—esteem scale. too. For the other SPQ scales common to both studies one would expect much the same ordering of mean scores. with the §h_and‘ig! mothers scoring together and the Eh and high mothers scoring together. Examination of the scales which comprised the Aparental involvement group in the present study (Ad. Cr. T. Ri. & As) indicates that. with the exception of achievement standards. low mothers obtained the highest scores. 2h mothers the lowest scores. and the scores of high and flh mothers fell between these two extremes. If this group of scales does. in fact. reflect pg;- ental involvement. or probably more accurately. the degree to which mothers are attempting to create the impression that they are doing a good job of parenting. then this order— ing of the mean scores of the groups may be explained as follows. Mothers of PA boys participated in the 1966 study for the purpose of obtaining help. Some referred themselves TABLE 15. Ordering of Mean Scores on Eight SPQ Scales for 101 Mothers of Poorly Adjusted (PA). Well-Adjusted (WA). Low (Lo). and High (Hi) Boys. Scale Order Rejection (R) 1:84 1314 2&02 ZSAG .2: .5: .22 1"}. Sex anxiety (Sa) 1:85 1:16 2383 2&01 Affection demonstrated (Ad) 2:81 2310 2:16 3%?3' Contingent reward (Cr) 1384 1210 1:94 1285 Restrictiveness (T) 2868 2112 Egg; ‘3??? Rewarding independence (Ri) 2:80 2:81 2??4 .3597 Achievement standards (As) 22%1 2:83 2385 2:94 Note.-—PA and WA group boys (n = 9) were of approximately the same age range and were from the same two schools as are the Low and High group boys (n = 15) in the present sample. to psychological clinics. most expressed concerns about their having done a poor job of rearing their sons. The mothers of low group boys in the present study gave no such indications 102 that they saw themselves as "bad" parents. On the contrary. these mothers impressed the investigator as having volun- teered for the study in order to demonstrate to both the investigator and their son's teacher that they were "good" parents. Although this interpretation is highly specula- tive. it is suggested that ipp_mothers scored higher than high mothers on the scales subsumed under_pgrental involve- hghh_in the present study because they felt it necessary to create an impression of good parenting. Mother-Son Interaction In addition to the hypotheses concerning the separ- ate behaviors of sons and mothers. an exploratory question was asked for which the present study provides relevant data. Are there consistent patterns of mother-son interac— tion which serve to maintain the boys' coping behaviors? Reactions to Frustration The boys' reactions to frustration and their mothers' responses to those reactions are one source of information 103 on interaction. Since the rating method specified that maternal behavior be coded only when it was a response to boys' behavior. these data apply solely to son-mother in— teraction. One would suspect on the basis of the results presented so far that there would be differences between the son-mother interactions of'ipy and high mother-son pairs. Table 16 gives the proportions of son-mother inter- action occurring in the various possible reaction to frus- tration dyads. These data are only descriptive in nature and do not permit statistical inferences to be made concern- ing the differences between groups. The most frequently occurring interaction for the mother—son pairs in the group of "constructive c0pers" was constructive reaction—nondirective response. 0n the other hand. mother—son pairs in the group of boys who were "less constructive copers" demonstrated a relatively high frequency of two quite different son-mother interactions: constructive reaction-negative response and withdrawal-helpful response. The most frequently occurring interactions in the two groups suggest that mothers of highs and mothers of igpg tended to follow different patterns of reinforcement of their sons' behavior. The nondirective response which occurred so 104 TABLE 16. Mean Proportions of Son-Mother Interaction Occurring in the Reaction to Frustration Dyads for Low and High Mother- Son Pairs. f ‘—1: G Son r Mother Response Reaction 0 u Ignor- Nega— Help- Nondi- Posi- p ing tive ful rective tive A ression Lo .05 .Ol .00 .01 .00 99 Hi .01 .00 .00 .01 .00 Re ression Lo .05 .02 .01 .01 .01 9 Hi .01 .01 .00 .01 .00 . Lo .05 .04 .13 .02 .01 Withdrawal Hi .03 .02 .08 .04 .02 Intropuni- Lo .04 .02 .02 .01 .01 tiveness Hi .03 .02 .00 .02 .03 Constructive Lo .08 .14 .11‘ .10 .05 reaction Hi .17 .04 .13 .22 .09 often in mothers of highs might be expected to help maintain constructive activity because of its implicit approval and noninterference. The negative response which l2! mothers demonstrated so frequently undoubtedly serves to discourage further constructive activity. while the helpful response to sons' withdrawal probably accomplishes the same end. by rein- forcing helplessness. 105 Mothers' responses. however. are at least partly a function of the boys' behavior. and these two groups of boys did not provide their mothers with the same stimuli. In an attempt statistically to "equalize" sons' behavior for the two groups. each category was taken separately to represent the universe of boys' behavior. For example. sons' aggressive reactions were considered to represent all of their reactions. and the distribution of mothers' responses to aggression across the five maternal response categories were then comp- arable (equal to 100%) for both groups of mothers. However. since i2! boys showed more aggressive reactions than high boys. the distribution of responses to aggression for ipg' mothers is probably more reliable than that for high mothers. Distribution of i2! and high mothers' responses to each of the sons' reactions to frustration are presented in Table 17. Comparisons of the two groups of mothers suggest some apparent differences in the ways they responded to the various coping behaviors of the boys. When the sons reacted to frus- tration with aggression and regression. mothers of ippg_were more likely to ignore their sons' behavior than were mothers of highs. Mothers of highs. on the other hand. were more apt to be nondirective in their responses to such stimulations TABLE 17. 106 Distributions of the Mean Proportions of Mothers' Response in the Low and High Groups for Each of the Sons' Reactions to Frustration. m?“ i n G Mother Response Son r Reaction 3 Ignor- Nega— Help— Nondi- Posi- p ing tive ful rective tive A ression L0 .677 .139 .046 .111 .024 99 Hi .496 .033 .033 .377 .058 Re r ssion Lo .546 .178 .109 .086 .079 9 8 Hi .201 .215 .062 .363 .156 . Lo .200 .148 .498 .096 .056 Withdrawal Hi .156 .107 .422 .206 .105 Intropuni- L0 .403 .146 .187 .144 .117 tiveness Hi .320 .188 .019 .228 .243 Constructive L0 .176 .288 .236 .202 .096 reaction Hi .260 .066 .203 .328 .139 Note.-—The proportions across the five categories of maternal response have been set equal to unity for each of the categor- ies of sons' reaction. from their sons. While responses to withdrawal followed much the same pattern for the two groups of mothers. they responded to boys' intropunitiveness somewhat differently. hpp_mothers more frequently helped their sons. and high mothers more often responded in a positive and supportive manner. Probably the most important group differences were with regard to mothers' 107 responses to sons' constructive activity. As was inferred from the analysis of son-mother interactions. mothers of iggg exhibited negative reSponses to a greater degree than mothers of highs. while the latter group showed a stronger tendency to react nondirectively to the same stimulations. Interpersonal Behavior The tape-recorded verbal behavior of the frustration and verbal sessions provides data for both son-mother and mother—son interaction. Unlike the reactions to frustration. this behavior is continuous. There are essentially two ways of looking at these data. Son may be regarded as the sender and mother as the respondent in their interchanges. or these roles may be reversed. with mother viewed as the sender and son as the respondent. These two perspectives lead to quite different descriptions of interaction. and both will be uti- lized in the present study. Son-mother interaction: Since use of single categories or octants would spread the son—mother interaction too thinly over the very large number of possible dyadic interactions. only the quadrant divisions of the circumplex were used in 108 the description of interaction. The quadrants and the let- ter designations of the categories comprising them are: affiliation-dominance (NOPA); affiliation-submissiveness (JKLM): disaffiliation-submissiveness (FGHI); and disaffiliation-dominance (BCDE). Table 18 contains the mean proportions for each group of the various possible son-mother interactions. Particularly relevant to the exploratory question are those interactions which occurred in the frustration session. Both groups demonstrated the symmetrical NOPA: NOPA interaction most frequently. with this sequence com- prising 39% of the igggf and 43% of the highs' total inter- action in the frustration session. Some of the less fre- quently occurring sequences reveal some apparent group dif- ferences. While NOPA:JKLM occurs fairly often (12%) in the high group. it is not very common (4%) in the ip!_group. As will become apparent in a subsequent section. this inter- action is very important because it served to perpetuate sons' positively assertive (NOPA) behavior. Although the two groups demonstrated JKLM:NOPA. the typical interaction for son-mother pairs. to approximately the same degree. two other complementary interactions were present more often in 109 TABLE 18. Mean Proportions of Son-Mother Interaction Occur- ring in the Various Circumplex Quadrant Dyads for Low and High Mother-Son Pairs. -======-==== Lows Highs Son Mother Total F V Total F V NOPA NOPA .42 .39 .45 .48 .43 .48 JKLM .03 .04 .02 .09 .12 .08 FGHI .01 .03 .01 .02 .04 .02 BCDE .01 .00 .01 .00 .00 .00 JKLM NOPA .17 .22 .14 .18 .17 .17 JKLM .01 .01 .01 .01 .02 .01 FGHI .01 .01 .00 .01 .02 .00 BCDE .00 .00 .00 .00 .00 .00 FGHI NOPA .27 .20 .31 .18 .12 .20 JKLM .01 .00 .01 .00 .01 .01 FGHI .01 .01 .00 .01 .01 .01 BCDE .01 .01 .01 .00 .00 .00 BCDE NOPA .04 .06 .03 .02 .02 .02 JKLM .00 .01 .00 .00 .00 .00 FGHI .00 .01 .00 .00 .00 .00 BCDE .00 .00 .00 .00 .00 .00 110 the ip!_group. These were FGHI:NOPA (20%.for iggg, 12% for M) and BCDE:NOPA (6% for ipls. 2% for h__i_g_l_1_s_). In order to determine if the two groups of mothers responded differently to the same sender behavior in sons. each quadrant of sons' sender behavior was taken separately to represent the universe of sons' behavior. The distribu— tion of mothers' responses to these quadrants of sons' sender behavior may be found in Appendices E-l (ippg) and E-2 (highg). The distributions for the two groups of mothers were compared by using the high mothers' distribution as the basis for expected frequencies in the various maternal re- sponse categories. Then. the chi square test could be used' to assess the extent to which the observed frequencies for ipg_mothers differed from (were independent of) these ex- pected frequencies. Since it is desirable to have expected frequencies of at least 5 in each category (Hays. 1963). related categories were combined into the smallest divisions that would adequately meet this condition. The following divisions of the circumplex for categories of mothers' re— sponses were employed in the comparisons: disaffiliation (B-I): affiliation-autonomy.(J-M): reassurance (N): hgip (0): teaching (P) and dominance (A). The distributions of 111 responses for the two groups of mothers were compared for each quadrant of sons' sender behavior separately. Table 19 gives the results. TABLE 19. Comparison of the Response Distributions of Low and High Mothers for Quadrants of Sons' Sender Behavior m— mm Son 2 Category X P BCDE 34.77 .002 FGHI 37.71 .002 JKLM 23.84 .002 NOPA 100.44 .001 Note.—-P values are for 5 degrees of freedom and are two-tailed. The distributions of mothers' responses to all of the quadrants of sons' sender behavior show significant differences between ippg and‘highg, Appendices E—l and E-2 indicate more specifically the ways in which the responses of the two groups of mothers differed. When sons demon- Strated poorly controlled behavior (BCDE). ipp_mothers were more apt to respond in a negative manner (15%.for ipgg, 4% fOr‘highg) and less likely to £2222 or structure (29%.for %. 46% for highs) than were high mothers. In response :0 son: havior 112 to sons' passive-aggressive and negatively dependent be- havior (FGHI). low mothers showed less affiliative-submis- sive behavior (2%.for lows. 6% for highs). less reassurance (5%.for ipgg. 1r% for highg). and more dominance (9% for ‘iggg. 5% for highg) than did high mothers. The two groups of mothers were most alike in their handling of sons' affiliation-submissiveness (JKLM). the only quadrant of behavior which did not differentiate i2! and high boys. hpp_mothers tended to respond with greater dominance (7% for ipgg, 3% for highg) and with less affiliative—submissive behavior (4% for iggg. 9% for highg) than did high mothers. The greatest group difference occurred with respect to mothers' responses to affiliative-dominant (NOPA) stimu— lations of their sons. ‘hpp_mothers demonstrated more gphr inance (10% for i235. 5% for highg) and teaching (44% for ippg. 34% for highp, while high mothers responded with a greater amount of affiliative—submissive acts (17% for highg, 6% for i933). These differing patterns of rein- forcement are of particular interest since boys' NOPA be- havior includes many of the same behaviors included under constructive reactions to frustration. Whether or not these different responses of the two groups of mothers £19! mot :ain b0: only b behavio be cons respons behavic in Appe respons 113 (ipg.mothers' dominance and teachihg. high mothers' affiliation—submissiveness) serve to discourage or main- tain boys' constructive (NOPA) behavior can be determined only burexamining how the boys respond to those maternal behaviors. Boys' responses to mothers' stimulations will be considered below. Although this analysis of mothers' reaponses has dealt only with broad classes of sons' sender behavior. data for individual son categories may be found in Appendix E-3. which presents the distribution of maternal reSponses for ipy and high mothers combined. Mother—son interaction: The mean proportions for ipg and high_groups of the various possible mother—son in- teractions are presented in Table 20. As was the case with son—mother interactions. the most frequently occurring se- quence in the frustration session was NOPA:NOPA for both groups. This symmetrical interaction occurred somewhat more often in the high group (46%) than in the ipy group (39%). Three kinds of complementary mother-son sequences were present to a greater extent in the ipg group than they were in the high group. and all involved sender behavior of the mothers which was affiliative-controlling. The sequences were: NOPA:JKLM (22% for lows. 17% for highs); NOPA:FGHI :81? 2f ir. the \ Mather-i Mouser “U ’4 A 1'.“ '1 30:91 KIM F o. I ~ ‘ G4.I 114 TABLE 20. Mean Proportions of Mother-Son Interaction Occurring in the Various Circumplex Quadrant Dyads for Low and High Mother-Son Pairs. ====================s ,:================ “#E Lows Highs Mother Son Total F V Total F V NOPA NOPA .42 .39 .44 .48 .46 .50 JKLM .15 .22 .15 .18 .17 .17 FGHI .29 .19 .32 .17 .12 .20 BCDE .04 .08 .03 .02 .02 .03 JKLM NOPA .02 .04 .02 .08 .14 .06 JKLM .01 .02 .00 .02 .02 .02 FGHI .01 .01 .01 .01 .01 .01 BCDE .00 .00 .00 .00 .00 .00 FGHI NOPA .01 .01 .01 .02 .04 .00 JKLM .01 .01 .00 .01 .01 .00 FGHI .01 .01 .00 .01 .01 .01 BCDE .00 .01 .00 .00 .00 .00 BCDE NOPA .01 .00 .01 .00 .00 .00 JKLM .00 .00 .00 .00 .00 .00 FGHI .01 .01 .01 .00 .00 .00 BCDE .01 .00 .00 .00 .00 .00 (19% 2% f: t10n is J] .3 u (2 J ..o , .. , c .3011”. was ‘ hav1< quad. Octa: nifi. 115 (19% for‘iggg. 12% for highg); and NOPA:BCDE (8% for i235. 2% for highg). One quite different complementary interac— tion was found more often in the high group. This sequence is JKLM:NOPA (14% for highp, 4% for iggg).'which involves a reversal of the standard adult-child roles. In an attempt to determine whether or not.ipg and high_boys responded differently to the same stimulations from mothers. each quadrant of mothers' sender behavior was taken separately to represent the universe of her be- havior. Because of the very large frequencies in the NOPA quadrant. it was possible to divide it into the N0 and PA octants. The distributions of sons' responses to these categories of mothers' sender behavior are presented for the‘ig! group in Appendix E-4 and for the high group in Appendix E-S. Using the distribution of responses for high.boys as the source of expected frequencies. the degree of independence between the distributions for high_and igy_ boys was assessed by means of the chi square test of sig- nificance. To insure sufficiently high frequencies in son response categories. the following circumplex divisions were used: disaffiliation (B-I); affiliation-submissiveness (J-M): help (0); and structuring & teaching-dominance (PA). Table ' distrl N0 tail anal WEIE I710 t} Sult sugg tior and 8am 116 Table 21 indicates the results of comparisons of the response distributions of the two groups of boys for each quadrant of mother sender behavior. TABLE 21. Comparison of the Response Distributions of Low and High Boys for Quadrants of Mothers' Sender Behavior. W ==== Mother 2 P Categorya X FGHI 10.16 .05 JKLM 10.93 .05 NO 53.25 .002 PA 92.56 .001 Note.--P values are for 4 degrees of freedom and are two- tailed. aBCDE occurred too infrequently to be included in the analysis. The response distributions for low and high boys were significantly different for all of the classes of mother sender behavior. This finding along with the re- sults of the comparisons for the two groups of mothers suggest that the differences in various types of interac— tion for low and high groups are a function of both mothers and sons. Low and high mothers re5pond differently to the "same" stimulations. and so also do low and high boys. that 1: mother relativ fare. t1 groups c negatim were CO at: “elect the a 117 Appendices E-4 and E-5 give an indication of the specific ways in which the two groups of boys differ. When mothers demon— strated disaffiliation-autonomy (FGHI). ipp_boys responded with more negative behaviors (46% for iggg. 28% for highg) and fewer helping behaviors (16%.for igpg. 39% for highp) than did high boys. This finding is understandable if one remembers that breaking FGHI down into its octants revealed that high mothers showed relatively more helplessness—submission and relatively less complaint—suspicion than i2! mothers. There— fore. the FGHI stimulations were not the same for the two groups of boys. The i9! boys apparently were responding to negative stimulations with more of the same. While high boys were coming to the aid of their "helpless" mothers. The net effect was that high boys were manifesting much more construc- tive activity (help) than were ipg_boys. There is no reason to suspect that affiliation— autonomy (JKLM) stimulations differed for high and ipg, mothers. However. the two groups of boys differed in their responses to JKLM. with igg boys more frequently responding with negatively dependent behavior (HI: 11% for igpg. 3% for highg) and less apt to respond with positively assertive be- havior (NOPA: 62% for lows. 75% for highs) than were high bays. out of :unity strated 10 any 10.“ 9n or hel} havior ously, 118 boys. Thus. when the igg‘boys were given a chance to get out of the dependent role. they were more likely to resist the change and less likely to take advantage of the Oppor- tunity than were high boys. Both groups of boys demon— strated much more NOPA behavior to JKLM stimultations than to any other sender behaviors of the mother. The fact that i9! group boys tended more often to remain in the dependent or helpless role was undoubtedly also a function of the be— havior of i2! group mothers. As has been mentioned previ- ously. these mothers responded to sons' NOPA behavior with more dominance and teaching and less JKLM.behaviors than did high group mothers. The most marked differences in the responses of ipp and high boys occurred to NO and PA stimulations of the mother. Mothers' reassurance and help were greeted with negative behavior (B-I) more often by i2! boys (31% for i235. 21% for highg) and with positively assertive be- havior (NOPA) more often by high_boys (57% for Elflhfix 43% for iggg). The same pattern of differences existed for responses to mothers' PA behaviors. with i2! boys demon- strating more negative responses (B—I: 38% for iggg. 24% for highs) and fewer positively assertive responses (NOPA; 49% for preser 5901186 IESpoz keno: APPS nt 119 48% for ipygy 57%.for highp) than high_boys. It would appear that in their attempts to keep from being controlled and protected by their mothers. ig!_group boys were more likely to demonstrate passive-negative behavior. while high group boys were more likely to attempt a symmetrical relationship. The fact that i2! mothers showed signifi- cantly more NOPA behavior than high mothers suggests that i2! boys were less successful in their attempts to keep from being dominated than were high boys. While the data presented in this section have dealt only with sons' re- sponses to broad classes of mothers' stimulations. their responses to individual categories of mothers' sender be— havior are presented for high and igg_groups together in Appendix E-6. the p certa that exam: comp: clas of t inte of m was rele grou ings PIES whie] ”P or CHAPTER IV DISCUSSION Before conclusions are drawn from the results of the present study, it would seem important to reemphasize certain aspects of the sample of Se and the methodology that make this research different from other attempts to examine mother-son interaction. The mother-son pairs comprise a highly homogeneous middle- and lower middle- class sample. The only seriously unrepresentative aspect of the sample is undoubtedly the somewhat above average intelligence of the boys. With regard to the two groups of mothers and sons, a high degree of pair-wise matching was achieved between ipg_and high Ss on all presumedly relevant variables except boys' intelligence. The high group boys show a trend toward higher intelligence rat- ings. In other ways, too. the i2! and high groups of the present study are probably more alike than are the groups which most studies have compared. Both groups are made up of "normals." differing only with respect to teachers' 120 ratings bays he time w the te mother sons 5 group behav: “Boys Likew rmar they their mothe grou; aSpec discu back regar tak 1 m SGIVQE nPoor 121 ratings of the boys' classroom behavior. The ig! group boys had not been referred for psychological help. Some time was spent in talking to the mothers subsequent to the test sessions, and these discussions suggested that mothers of ippp as well as mothers of highg viewed their sons as essentially "normal." While some of the igg' group mothers had received bad reports on their son's behavior at school. they generally dismissed these with. "Boys will be boys," or "Johnny doesn't like his teacher." Likewise, mothers of ipg_group boys apparently did not regard themselves as failures at child—rearing nor did they feel that they needed psychological help in handling their sons. When the investigator discussed with the mothers his observations of mother-son behavior, ipg. group mothers tended to deny or rationalize the negative aspects of this feedback. Their behavior during these discussions as well as their apparent distortion of feed- back from their sons' teachers may indicate suspiciousness regarding outside authorities, a negative attitude about taking help from others. and/or a need to defend them- selves against what they may perceive as accusations of "poor parenting." in com simila the ci the ca diffs: up th JKLM as in cedux by 0‘ but grea quad refe such Thus fram moth SPEC‘ Cire! 122 The a3pect of methodology which must be considered in comparing the results of the research with those of similarly designed studies is the particular way in‘which the circumplex rating scheme was employed. First of all. the categories or interpersonal mechanisms were defined differently for sons and mothers in an attempt to break up the set of viewing most child behavior as reflecting JKLM (affiliation-submissiveness) and most adult behavior as indicative of NOPA (affiliation-dominance). This pro- cedure was successful in reducing the amount of rated JKLM behavior for boys to a level lower than that reported by other studies (Raush, et a1., 1960: MacKenzie, 1968). but it also resulted in the stereotypical rating of a great preponderance of maternal behavior in the NOPA quadrant. While rater instructions contained specific reference to avoiding the JKLM stereotype for boys, no such reference was made to the NOPA stereotype for mothers. Thus, the raters seemed to have used a separate. child frame reference to code boys' behavior. but to have coded mothers' behavior by implicitly using an adult-child per- Spective. A better distribution of ratings around the circumplex can probably be best accomplished by instructing judges t each 0th dorinanc departu: tions 0 that in tive of present pathic son. ) majors Edge a] to tak. ferent the m0 The ne Sacrif high i: agreem. the hi9 123 judges to compare mothers with each other and boys with each other, particularly with regard to ratings on the dominance-submission axis. A second point concerning rating procedure is the departure, in the present study, from the rating instruc- tions of Freedman. et a1. (1951). These authors recommend that interpersonal mechanisms be rated from the perspec- tive of the one who is being acted upon. Raters in the present study found it difficult to maintain this em- pathic set while shifting back and forth from mother to son. Moreover. the raters were undergraduate psychology majors who were relatively unsophisticated in their knowl— edge about psychological processes. They were more apt to take behavior at face value and less likely to be in- ferential in their judgments than would be the case for the more experienced clinician using the circumplex system. The net effect of these differences seems to have been a sacrifice of psychological "depth" in the ratings for very high inter-rater reliabilities. The unit by unit per cent agreement reliabilities obtained in the present study are the highest thus far reported. ehhcyi with the un hehavj still hseli such rate unit spec stih res; stir weir cod mot 124 A third procedure which differs among studies employing the circumplex rating scheme is the manner in which the behavior phih is defined. In the present study, the uninterrupted speech served as the basic unit. If behavior or affect changed within the unit, the unit was still scored only for the predominant behavior or affect. Mueller and Dilling (1968) point out the disadvantages of such a global evaluation of the unit. When one does not rate in sequence different behaviors occurring within a unit. two kinds of information are lost: data about the specific behavioral response to the immediately preceding stimulus; and data concerning the manner in which the respondent reorganizes internally in response to previous stimulation. Loss of these two kinds of data should be weighed against the greater amount of work involved in coding and analyzing intra-unit sequences. Reactions to Frustration The rated verbal and nonverbal behavior of sons and mothers in the frustration session was examined with respect to its bearing on two major questions. The first asked what child var with true study we: on the v; achiever were apt (it; am rated r in both Puzzle) icantlj With it While . the iEL boys d that y suffic t0 dem likely \ middle. boys sa 125 child variables are involved in the boy's ability to COpe with frustration. The two groups of boys in the present study were chosen on the basis of their different standing on the variables of self-control. self-sufficiency, and achievement motivation. While igy and high group boys were apparently frustrated to the same extent by the task (ig! and high groups did not differ regarding number of rated reactions to frustration. and the majority of boys in both groups surpassed the time limit for work on the puzzle), the two groups reacted to frustration in signif- icantly different ways. hpngroup boys reacted more often with aggression and regression than did high group boys, while constructive reactions occurred more frequently in the high group than in the low group. The two groups of boys did not differ with regard to withdrawal and intro— punitive reactions. One may conclude1 from these findings that young boys who are high in self-control. self- sufficiency, and achievement motivation are more likely to demonstrate constructive coping behaviors and less likely to react to frustration with poorly controlled, 1Conclusions should be generalized only to the middle- and lower middle-class population of 7-9 year old boys sampled in the present study. unconstru: | these var the Egg sufficie' the inve speculat high deg cause i‘ acted w tendenc distrac puzzle could r Work or prize. same t: suffic: self fc has to of the with th 126 unconstructive behaviors than are boys who are low on these variables. Although there were no objective data to indicate the specific relationships of self-control, self- sufficiency, and achievement motivation to coping behavior, the investigator's observations of the boys enable some speculation as to what these relationships might be. A high degree of self-control would seem to be effective be- cause it keeps the boy in the situation. When boys re- acted with poorly controlled behavior, this behavior had a tendency to "snowball." Their attention shifted from one distracting stimulus to another, and their work on the puzzle became increasingly erratic. Some of these boys could not delay gratification. After a few moments of work on the puzzle, they would start playing with the prize. Occasionally a boy claimed the prize while at the same time denying any further need for work on the puzzle. It is felt that the effectiveness of high self- sufficiency lies in its enabling the boy to look to him- self for evaluation of his progress and for directions on how to proceed. This point was clearly illustrated by many of the high_boys as they virtually carried on a conversation with themselves. They would ask questions and then answer them. qu€ can progr' Atthe 0t tinually few moths ing the y from her some of remarked Although Structir she 0ij believi: on his 1 behavio. somethi: high n a as eXtEJ or not 1 or that gave Up 127 them. give directions to themselves. and evaluate their own progress, as if there were no one else in the room. At the other extreme were some of the igy_boys who con- tinually turned to their mothers for help. There were a few mothers who obliged their dependent sons by complet- ing the puzzle for them. One such mother took the puzzle from her son after he had made a feeble attempt to move some of the pieces, and as she completed the puzzle she remarked again and again on how well hhgy were doing. Although she attempted to deceive the investigator by in- structing her son to take credit for the finished puzzle, she obviously could not successfully deceive her son into believing that he was capable of coping with situations on his own. The role of high achievement motivation in coping behavior is apparently to enable the boy to persist at something when the likelihood of success is in doubt. The high n achiever seems to be motivated by intrinsic as well as external, material rewards. Whether he wins the prize or not he wants to be able to say that he solved the puzzle or that he did his best. While some of the igp group boys gave up on the puzzle, all of the high group boys continued atey oi for their to work : irustrat constru: less c0 tration Contrc “seats mOthe: Positj greate their 128 their effort until the investigator ended the session. In a few of the cases where the boys were awarded the prize for their efforts, they put the prize aside and continued to work on the puzzle. The second major question concerning reactions to frustration dealt with mothers' behavior. Do mothers of constructive copers respond differently from mothers of less constructive c0pers to their sons' reactions to frus- tration? The two groups of mothers did respond differ- ently, with ig!_mothers showing significantly more hggge tive responses and restrictive help and significantly fewer nondirective responses and positive responses than did high mothers. However, the two groups of mothers were not responding to the same child stimulations. The poorly controlled, unconstructive behavior of ig!_boys may "pull" negative and restrictive, intrusive responses from their mothers, while the greater frequency of nondirective and positive responses of high mothers may be due to the greater amount of constructive activity demonstrated by their sons. Taking each of the categories of sons' reactions separately to represent the universe of sons' behavior, itwas f ‘erehtiy ularly ‘ ity. wi aid ies tion, are af hayioz r953301 be di: 129 it was found that the two groups of mothers responded dif- ferently to the same child stimulations. This was partic— ularly true of their response to sons' constructive activ— ity. with igg mothers more likely to respond negatively and less likely to respond nondirectly than high mothers. Since there are no comparable data on mother—son interac- tion. it is not possible to determine what effects these responses of the mothers had on their sons' subsequent behavior. It may be conjectured. however. that the cumu- lative effect of high mothers' greater nondirectiveness is to facilitate further constructive activity in their sons. The greater degree of negative responses (restric- tion. criticism. etc.) in igp mothers suggests that they are ambivalent about self-sufficient and assertive be- haviors in their sons. and to the extent that their sons respond to the negative side of this ambivalence. they may be discouraged from further constructive activity. Interpersonal Behavior The circumplex ratings of the verbal behavior of mothers and sons in the frustration and verbal sessions were exam hhgyaud their int between t airs, a: that dif Let em nifican- hehavic iFG) ar boys, m5tche were n Peflder are VG SivE i 130 were examined with respect to two major questions. Do high and igy_boys differ and do their mothers differ in their interpersonal behavior? And if differences exist between the separate behaviors of high and igg_son-mother pairs, are there also consistent patterns of interaction that differ for the two groups? Comparison of the separate behaviors of ig!_and high group boys revealed that igg_boys demonstrated sig— nificantly more poorly controlled and negatively assertive behavior (BCDE) and significantly more passive-aggressive (PG) and negatively dependent (HI) behavior than did high_ boys. high boys significantly exceeded their ig! group matches on positively assertive behavior (NOPA). There were no group differences with respect to positively de- pendent behavior (JK) or cooperation (L). These findings are very similar to those of MacKenzie (1968) with aggres- sive underachievers and normal controls. The only differ- ence between the results of the two studies is that her normals showed more JKLM behavior than her clinic boys. Much of this difference is probably due to the fact that much of the JKLM behavior observed by MacKenzie would have been scored NOPA in the present study. irterpers icantly n atrend 1 Idiots de' .1. behavior These fi ‘w‘ith thr her gro' hostih the pre submisg Signif: lCdntlj 131 The two groups of mothers also differed in their interpersonal behavior, with ipg_mothers showing signif- icantly more dominant and protective behavior (NOPA) and a trend toward greater rejection (BCDE) and mothers of high§_demonstrating significantly more autonomy granting behaviors that were both positive (JKLM) and negative (HI). These findings for mothers are only partly in agreement with those of MacKenzie. While the differences between her groups of mothers were largely a function of the love- hostility axis of the circumplex, i2! and high mothers in the present study differed primarily on the dominant- submissive axis. MacKenzie's clinic mothers demonstated significantly more negative behavior (BCDE) and signif- icantly less positive behavior (JKLM and NOPA) than did her normal mothers. Because her clinic mothers comprised a more extreme group (psychological clinic referrals) than i9! mothers, much of the dominant behavior scored as NOPA for ipg mothers would probably have been scored negatively as BCDE for her clinic mothers. The findings with respect to the behavior of igg_ and high mothers are similar to those of most studies which have used comparable groups of Ss. Becker, et a1. (1959) iomd thai dictatori 'oserved behaviors cents tha observed nificant of more and d’Ar more d0] mothers ings of That a ence of bEhavio and m the 0th decreas aPPIOpr. defined Other he 132 found that mothers of aggressive school children were dictatorial and restrictive, Goldstein, et a1. (1967) observed more power-assertive and less opinion-seeking behaviors in mothers of aggressive, externalizing adoles- cents than in mothers of internalizers, and Hilton (1967) observed that mothers of dependent preschoolers were sig- nificantly more involved and interfering than were mothers of more independent children. The observation of Rosen and d'Andrade (1959) that mothers of high n achievers were more dominating of their preadolescent sons than were mothers of low n achievers is inconsistent with the find- ings of the present study. How are the behaviors of sons and mothers related? That a relationship exists is suggested by the correspond- ence of changes in mothers' behavior to changes in sons' behavior from frustration to verbal sessions. High boys and high mothers showed little change from one session to the other. The only significant changes were high_mothers' decrease in reassurance and increase in teaching, both apprOpriately cued to the instructional differences which defined these two sessions. pg! sons and mothers, on the- other hand, exhibited shifts in their behavior that were couple their and i mothe grant tecti relat is e) in se prev: exam the 081' Cons be i Spec mot} difj SOn. 133 complementary to each other. While igg_sons decreased their negatively assertive and poorly controlled behavior and increased their negatively dependent behavior, igg, mothers decreased both positive and negative autonomy- granting behavior and increased their dominant and pro- tective behavior. However, what is really crucial to determining relationships between the behaviors of mothers and sons is examination of their respective behaviors as they occur in sequences at the same point in time. As pointed out previously, interaction between mother and son can be examined from two perspectives. Son may be regarded as the sender and mother as the respondent, or the roles may be reversed for the participants. Since this research considered "child effects" as well as "mother effects" to be important, interaction was examined from both per- spectives. As one might expect from the differences between igg and high sons' behaviors and between igg and high mothers' behaviors, the two groups of mother-son pairs differed in their interactions. While the most common son-mother and mother-son interaction for both groups was the syzu compiem hOPA sc mother- Oh the the JR: in the in the would the m iS dey more < help . is of m0the 134 the symmetrical NOPA:NOPA sequence, several less common complementary sequences were present in different degrees in ig!_and high_groups. The igg§_demonstrated the FGHI: NOPA son-mother sequence and the NOPAzJKLM and NOPAzFGHI mother-son sequences more often than high mother-son pairs. On the other hand, the NOPAzJKLM son-mother sequence and the JKLMzNOPA mother-son sequence occurred more frequently' in the interactions of high§_than in the interactions of iggs, Those interactions which occurred more frequently in the i9! group than in the high group are ones that would serve to maintain a mother-son relationship in which the mother is dominant and protective (NOPA) and the son is dependent (FGHI or JKLM). Interactions that occurred more often in the high group than in the igg_group would help to foster a mother-son relationship in which the son is often positively assertive (self—sufficient) and the mother is willing to grant him autonomy (JKLM). The more symmetrical mother-son relationship for highs was consistent across the frustration and the verbal sessions, while the more complementary relationship between low mothers and their sons was more marked in the verbal session than in the frustration session. In going from the firs their do creased order 0: was not second frustra changed the dif changes the E tunity mothers involve Proverl on how 3150 p1 COncerr aQQIESs may hav their m the Per: 135 the first to the second session, low mothers increased their dominance and protective behavior and low sons in- creased their negatively dependent behavior. Since the order of presentation of frustration and verbal sessions was not counter-balanced, changes from the first to the second session may reflect a carry-over of the effects of frustration or they may reflect the Ss' reactions to the changed structure of the verbal session. With respect to the different structure of the verbal session, these changes can probably be accounted for by the fact that the verbal session presented.l2w mothers with the oppor— tunity to take over and to prove themselves as "good mothers." The lgw mothers did appear to be more "ego- involved" than the high mothers in their teaching of the proverbs, and several asked the investigator for feedback on how well they had done. However, the $2! boys were also presented with an opportunity, if their mothers were concerned about doing a good job. The increased passive- aggressive and negatively dependent behavior of these boys may have been intended to make things more difficult for their mothers. It is apparent that interpretations from the perspective of "parent effect" or from the perspective of"chiL tions fo differed well as and lcgg separat cedure and big persona high mc sponse: 131 an: reSpon. These lien a: sons. and so Obser V1 136 of "child effect" may present equally plausible explana- tions for behavior occurring between mother and son. In an attempt to examine why lgw_and high groups differed in their interactions. sons' sender behavior as well as mothers' sender behavior was "equalized" for highs and lgg§_by regarding each category of sender behavior separately as the universe of sender behavior. This pro- cedure made it possible to determine whether or not lgw. and high.groups responded differently to the "same" inter- personal stimulations. Results indicated that-lg! and high_mothers were significantly different in their re- sponses to all major classes of son sender behavior, and $93 and high_sons were significantly different in their responses to all major classes of mother sender behavior. These results suggest that differences in the behavior of high and lg! groups are a function of both mothers and sons. It is most likely that an interaction of mother and son effects may best account for the group differences observed in the present study. A comparison of the specific patterns of response or reinforcement for lggg and highs permits some specula- tion about how the mother-son relationship differs in the twogro to res; stimuli tant) : group I and po likely and hi reSpor Sponse Sponse mainte constr reSpo; posit: which Both 1 Spend while SPond behavi 9€St t] to be 1 137 two groups. While both groups of mothers were most apt to respond with NOPA behaviors to all classes of sons' stimulations, the less frequent (and probably more impor- tant) responses of the mothers showed some interesting group differences. When sons showed negatively assertive and poorly controlled behavior, l2! mothers were more likely than high_mothers to respond in a negative manner, and high mothers were more likely than lgw_mothers to respond with structuring and teaching. The negative re- sponse of lg! mothers was likely to evoke a similar re- sponse in lgw_boys approximately 50%.of the time, thus maintaining the pattern of negative interaction and un- constructive activity. However, high boys very frequently responded to their mothers' structuring and teaching with positively assertive or constructive behavior (NOPA: 53%), which served to perpetuate a quite different pattern. Both lg!_mothers and their sons were more likely to re- spond negatively to negative stimulations than were highs, while high mothers and their sons were more likely to re- spond to negative stimulations with positively assertive behaviors than were lgwg. These differing patterns sug- gest that mother—son conflict may persist in lgw§_but tends to be resolved in highs. In the previous chapter it was sported 1 asters o axi aha-'51 imestige pst‘nesize deronstr cf copin Consider aid neg; sznila: ative-5‘ often w tective tithe; quently mothers ference ence in HOME be} quently Salts a] (I) C) nnot}. 138 reported that low mothers obtained higher scores than high mothers on a group of SPQ scales (sex-anxiety. punitiveness and physical punishment. and demands for aggression) which the investigator chose to call anxiety about closeness. It is hy— pothesized that the perpetuation of conflict and the frequent demonstration and encouragement of aggression constitute a way of c0ping with anxiety about closeness in 123 mother-son pairs. Consideration of this hypothesis is a task for future research. ggw_and high_mothers' responses to passive—aggressive and negatively dependent behaviors of their sons were fairly similar. While high_mothers responded more often with affili- ative-submissive behavior and reassurance. and 12w mothers more often with dominance. both groups of mothers demonstrated pro- tective and dominant behaviors (NOPA) approximately 90% of the time. The FGHI:NOPA son-mother interaction occurred more fre- quently in lgwg (27%) than in highs (18%). Since 12w_and high mothers responded similarly to FGHI behavior. it must be dif- ferences in boys' responses which account for the group differ— ence in FGHI:NOPA. If one looks at sons' responses to mothers' NOPA behavior. it is apparent that 193 sons responded more fre- quently with FGHI than did high sons. MacKenzie's (1968) re- sults also revealed a greater perpetuation of this FGHI:NOPA son-mother pattern in the poorly adjusted group. for this hCPA beh their fr mothers, indicati dominant nation ; strated (HI) th regard by Sear ambiva] sions 1 Uncerte may ex] inten81 139 There are two plausible explanations to account for this phenomenon. Lg! boys may perceive their mothers' NOPA behavior as an attempt to control them, to take away their freedom, and so they resist or withdraw (FGHI). L23; mothers, meanwhile, may interpret their sons' behavior as indicative of helplessness and therefore respond with more dominant and protective behavior (NOPA). A second expla- nation is suggested by the finding that lgw boys demon- strated significantly more negatively dependent behavior (HI) than higthoys, while the groups did not differ with regard to positively dependent behavior. As hypothesized by Sears, et a1. (1957), dependent (lgw) boys may be highly ambivalent about expressing dependence lest such expres— sions be punished by their mothers. Because of their uncertainty about how their mothers will react, lgw boys may express their dependency in an indirect but quite intense manner (FGHI). While the frequencies are small, lg! mothers did respond to their sons' FGHI behavior with more rejecting behavior (BCDE and FG) than did.hlflh mothers. The occasional rejections of the lgngoys' dependency expressions probably served to intensify their ambivalence. However, the majority of low mothers' responses t nature. WT. eith more F fibers re central W 359 majorii iii‘dl’e. B’ EXpress thf “it they ' SPCIld with for Somet} they haVe to both 0 scnmothe and mothe asse‘rtive and 1985 than did meats!“ attemPts given a c 140 responses to FGHI were dominant (62%) or protective (31%) in nature. Why would 12! sons react to these maternal behaviors with more FGHI? As the first explanation suggests. possibly 12w boys perceive these maternal behaviors as an attempt to control them. which is certainly consistent with the fact that the majority of l2w_mothers' NOPA behaviors were dominant in nature. But it is just as likely that so long as lgw_boys express their dependency in an indirect. intensely helpless way. they will not feel satisfied even when their mothers re- spond with help or reassurance. They have not directly asked for something. and therefore they are not likely to feel that they have really been given something. One factor is common to both of the explanations of the FGHI:NOPA sequence in $91 son-mother pairs. and that is the perception of both lgw_sons and mothers that lpy_sons are very helpless. That lgw_mothers responded to sons' positively assertive behaviors more often with dominant behavior (PA) and less frequently with autonomy-granting behavior (JKLM) than did.high mothers implies that 123 mothers are either threatened by or have little confidence in their sons' attempts to be self-reliant. When sons were presumedly given a chance to assert themselves in response to JKLM stimula' more he behavrc tion tl of ther adequa ihOPA) (FGHI) tiOnr Scale C0rre moths high relat 141 stimulations from their mothers, lgw_boys demonstrated more helpless behavior (HI) and less positively assertive behavior (NOPA) than did high boys. Here is the sugges— tion that lgw_boys concur with their mothers' perception of them as helpless and unable to cope with situations adequately. The above mentioned finding that lgw_boys reacted to mothers' attempts to dominate the situation (NOPA) with more passive-aggressive and helpless behavior (FGHI) and less positively assertive behavior (NOPA) than did higthoys confirms the impression that the complemen- tary "dependent son--controlling mother" relationship in lows is a function of both sons' and mothers' behavior. Mothers' Attitudes There were three groups of Stanford Parent Ques- tionnaire scales which differentiated lg!_and high mothers. Scales were grouped together on the basis of the inter- correlations among all 16 SPQ scales for lgw_and DEED. mothers combined. Each group contained scales which had high correlations with each other and relatively low cor- relations with scales outside the group. The first group of scales inconsisti —I—'- and was c_ laws scor inconsisi] is consi: ental at served n their sc and to 1 less agl nifican ment" f Parents Rau (ls boys‘ ( bEtwee; suffici to edo} 142 of scales contained low rejection, high self-esteem, low inconsistency, and high ppsitive father-mother relationship and was given the name, parental adjustment. Mothers of lows scored significantly higher on both the rejection and inconsistency scales than did high mothers. This finding is consistent with the results of several studies of par- ental attitudes and behavior. McCord, et a1. (1961) ob- served mothers of aggressive boys to be more rejecting of their sons, to be more inconsistent in their discipline, and to have a poorer marital relationship than mothers of less aggressive boys. Becker, et a1. (1961) found a sig- nificant relationship between a "general family maladjust- ment" factor (marital conflict and inconsistency between parents) and aggression in school children; Winder and Rau (1962) a relationship between parental rejection and boys' dependency, and Rau, et a1. (1964) an association between low rejection-high self-esteem and boys' self- sufficiency. A second group of scales that were highly related to each other was labeled, parental anxiety about close- pggg, and this group included punitiveness and physical punishment, demands for aggression, and sex anxiety. Low mothers two of t and degg with st‘ asserti' (Becker et al. , Punishni Irowitz, 143 mothers scored significantly higher than high mothers on two of these scales, ppnitiveness and physical punishment and demands for aggression. This result is in agreement with studies which have found a relationship between power- assertive discipline and externalized controls in children (Becker, 1964; Hoffman and Saltzstein, 1967: and Goldstein et al., 1967) and studies which have related physical punishment specifically to children's aggression (Lef- kowitz, et al., 1963; McCord, et al., 1961; and Becker, et al., 1962). The "aggressive model" interpretation seems to account most adequately for the present findings. When mothers of lgy boys use physical punishment on their sons, they provide a model for poorly controlled behavior. The greater demands of these mothers for their sons to react aggressively (fight for their rights) with their peers provides additional reinforcement for learning that one must be aggressive in order to cope with things and peeple in general. Physical punishment might also be expected to frustrate the boy and to instigate anger in him, as well as to threaten the love relationship between son and mother, leading to an anxious dependency in the boy. The significantly greater amount of passive-aggressive 144 and negatively dependent behavior (FGHI) demonstrated by lg!_boys in the present study is consistent with these interpretations of the effects of mothers' physical pun- ishment. The third group of highly intercorrelated scales was considered to reflect parental involvement, and this group contained a heterogeneous mixture of scales dealing with parenting techniques. Of the scales in this group, lg! mothers scored significantly higher on democragy, demands for conformity,,affection demonstrated, and achievement standards than did high'mothers. Two of these scales which differentiated the groups provide results consistent with those of previous research. Winder and Rau (1962) found mothers' demonstrated affection to be positively related to social deviancy in preadolescent boys. Possibly the greater demonstration of affection operated as a reaction formation to mask underlying atti- tudes of rejection toward the boy. Lg!_mothers' higher scores on achievement standards provide evidence that is in agreement with Winterbottom's (1958) findings regarding mothers of low and high n achievers. She determined that mothers of high n achievers make more demands on their childrer such de: mothers early i that is notivat ipplice the po; vokes ; Praoti presen Peflder String depem mothe: sPond werite Until and t tery 145 children for achievement prior to eight years and fewer such demands on their children after eight years than do mothers of low n achievers. She concluded that it is gagly_independence training and stressing of achievement that is crucial to the child's deve10pment of achievement motivation. This finding illustrates a point that is applicable to many areas of child-rearing, namely, that the point in the child's development at which mother in- vokes a particular practice may be as important as the practice, itself. Several of the lg! mothers in the present study acknowledged that their sons were too de- pendent, that they had waited too long to untie the apron strings, and that present measures to curb their sons' dependency were not very successful. However, these same mothers were observed to dominate their sons and to re- spond to dependent behavior by giving their sons what they wanted. It appears quite likely that low mothers waited until their sons were relatively old to stress achievement and to initiate independence training because they were and probably still are ambivalent about their sons' mas- tery behaviors. who foun P': 43 Slhme3 more peer group of C OnVl r tha t t 146 If the scales which have been organized under parental involvement are considered together, some do not appear to belong with others. For example, low mothers' higher scores on democracy seem to be inconsistent with their higher scores on demands for conformity. The same sort of phenomenon was reported by Lefkowitz, et a1. (1963), who found that mothers who indicated higher use of physical punsihment also tended to indicate greater use of other, more permissive techniques. The fact that a heterogeneous group of parenting techniques demonstrated high intercor- relations suggests that the present sample of both $93. and high mothers did not differentiate strict and permis- sive approaches to discipline. It is felt that the higher scores of low mothers on the parental involvement scales reflect a desire on the part of these mothers to create a ”good parent" image. Their greater ego—involvement in the proverb teaching task and their demonstration of signifi- cantly more dominant and intrusive behavior than high mothers support the impression that these lgy_mothers were concerned about their performance as mothers and needed to convince the investigator, the schools, and themselves that they were doing a good job of parenting. self-re materna to c0pi who is sent mc constru- confide her hat: use phi, ably be f€latic rally 1 constru him But her son relianc aChieve and as mother. Emhers 147 Integration of the results of observational and self-report measures suggests that several interdependent maternal variables are significant in their relationship to coping behavior in young boys. The mother of the boy who is high in self-control, self-sufficiency, and achieve- ment motivation and who copes with frustration in a largely constructive manner is apparently accepting of her son and confident in her role as a mother. She is consistent in her handling of her son's behavior and is not likely to use physical punishment as a method of discipline. Prob- ably because she and her son are confident of the love relationship between them, they do not need to be contin- ually involved with each other. While the mother of the constructive c0per can relax her control over him, grant him autonomy, and respond to him in a nondirective manner, her son can take advantage of these opportunities for self- reliance and mastery. He seems to regard independent achievement as within his power, as a desirable thing, and as nonthreatening to the mother-son relationship. A much different picture is suggested for the mother-son relationship of unconstructive c0pers and their mothers. The mother of the boy who is low in self-control, self-s of her Her pi be a 1 c0ntr< punisl son a1 sive, cause her 54 incon; behav d6pen. IEfer. in an by wh and h. must . 0f thi OVer 1 munioe capabj 148 self-sufficiency, and achievement motivation is rejecting of her son and seems to be unsure of herself as a parent. Her punitiveness and physical punishment of the boy may be a response to or an instigation to aggressive and poorly controlled behavior in him. Whatever its basis, physical punishment may jeOpardize the love relationship between son and mother while at the same time providing an aggres- sive, poorly controlled model for the boy. Probably be- cause she is ambivalent about independence and mastery in her son, the mother of the unconstructive c0per is both inconsistent and dominating in her handling of the boy's behavior. Her inconsistency and dominance probably foster dependency in her son by preventing him from deve10ping reference points for internal evaluation. When she reacts in an inconsistent manner, the boy has no stable guidelines Tby which he can internalize correct ways of proceeding, and he cannot act on the basis of past experience. He Inust continually aSk his mother what to do next. Dominance of the boy's activity, setting his goals for him, taking over his attempts at mastering a situation serve to com- Inunicate to the son that he is helpless, that he is in- capable of tackling situations independently, and that self-rel Because i tive cop and help controll in the c in the C defined nation o Of the 1 are sti. tion £01 the my "SEnt" tion of as inte whEn th thEir Grant a of Pare 149 self-reliance is threatening to the mother-son relationship. Because he perceives himself as helpless, the unconstruc— tive c0per responds to his mother in a passive-aggressive and helpless manner, thus perpetuating the "dependent son-- controlling mother" relationship. Interpretations of the parent-child relationship in the case of constructive vs. unconstructive c0pers or in the case of any groups of children which differ in defined ways are bound to be speculative. Even in a sit- uation which permits close observation and reliable coding of the unit—by-unit interaction of parent and child, we are still too far away from the action. The ideal situa- ‘tion for studying interpersonal processes is one in which 'the researcher has control over the cues or stimulation "sent" to the subject and can check the subjects' percep- ‘tion of these cues to determine whether they are "received" .as intended. Do lg!_mothers intend to help or to dominate ‘Mhen they respond with NOPA? How do lg! sons perceive ‘their mothers' NOPA behavior? The answers to these ques- tions must be inferred from the respondents' behavior. Grant and Kantor (1961) have called attention to the need Of parent—child research to examine the child's perceptions of his tions t group - there really coming impact BCDE 5 child resear in igc Yield by Wh: et all SOnal: differ signii One wc hols j each C w0uld 150 of his parents' behavior and his idiosyncratic interpreta— tions of it. Even though the present study investigates group differences in responses to the "same" stimulations, there is no certainty that what is coded as the same is really the same for different people. A BCDE stimulation coming from a ig! group boy may have a quite different impact on his mother than the "same" stimulation from a high group boy has on his mother. Past experience with BCDE stimulations undoubtedly differs for the two groups. Another methodological difficulty of much parent child research is the exclusion of the father from the research picture. Study of the mother-child relationship in isolation from other intra-family relationships can yield no more than a limited knowledge about the processes by which child and parent influence each other. Becker, et a1. (1961), for example, found that the separate per- sonalities of mother and father as well as their marital differences had a combined impact on the child that was Significantly related to his deve10pment of aggression. One would predict that the mothers and fathers of igg. boys in the present study would be less consistent with each other in their attitudes toward child-rearing than Would the parents of high boys. provide presen condit: tape. separa regard and wh could affect other behav: membe: Cempa liste liste in pa latiC 0f ir tion. 151 An approach borrowed from Goldstein et a1. would provide a means by which data that were missing from the present study could be obtained. Family interaction under conditions of stress could be recorded on audio or video tape. Then the tape could be played back to each member separately, and the investigator could question the member regarding what his communications were intended to mean and what other members' communications meant to him. One could also have members role play various behaviors or affects (e.g., anger) and have the member predict how other family members will respond to him. This role play behavior could then be played back to the other family members, their responses recorded, and these responses compared to both the sender's intended effect on the listener and the sender's predicted response from the listener. Fresh approaches and new methods are needed in parent—child research, and it is felt that the manipu- lation of interpersonal stimulations and the assessment of interpersonal responses is a step in the right direc- tion. betweer boys. iables frustr tion I him? mothe: who w ratir Idler thes Was fill QUES CHAPTER V S UMMARY The present research examined the relationship between mother-son interaction and the coping behavior of boys. Three questions were of concern. What child var- iables are involved in the boy's ability to cope with frustration? Is the boy's manner of c0ping with frustra- tion related to his mother's attitudes and behavior toward him? And what is the nature of the relationship between mother's behavior and son's manner of coping? 85 were 32 preadolescent boys and their mothers, who were divided into two groups on the basis of teachers' ratings of the boys' classroom behavior. hggg were boys rated low in self-control, self-sufficiency, and achieve- E9115 motivation, while highs had received high ratings on these three variables. A high degree of pair-wise matching ‘was achieved on other presumedly relevant factors. Mothers filled out an abbreviated version of the Stanford Parent Questionnaire, and they participated in two interaction 152 situati tration session proverh one-wa; frustre sessio was co 0f cat Coffey rating 153 situations with their sons. In the first session (high: tration session) the boy was intentionally frustrated by offering him a prize for completion of a puzzle which was too difficult for him, while the second session (verbal session) required the boy's mother to teach him three proverbs. Observing the frustration session through a one-way mirror, two judges rated the boy's reactions to frustration and his mother's responses to them. Both sessions were tape-recorded, and the verbal interaction ‘was coded using the interpersonal rating scheme and system of categories deve10ped by Freedman, Leary, Ossorio, and Coffey. Interjudge agreement was very high for both rating methods. Analysis of the "reactions to frustration" data revealed that i9! boys demonstrated aggressive and regres- giyg reactions to a significantly greater extent and SEE? structive reactions to a significantly lesser extent than high boys. The two groups did not differ with regard to Eithdrawal and intrppunitive reactions. Likewise, the two groups of mothers differed significantly in their responses, With mothers of iggg giving more negative responses and fewer nondirective and positive responses than high mothers. Consi: and h; for t the c havic that ativ [:I‘ H. C) 154 Consideration of son—mother sequences indicated that igg and high mothers showed different patterns of reinforcement for the same son behaviors, with the differences suggesting that igg mothers tend to give negative and high mothers positive reinforcement for their sons' efforts at mastery. Similar findings were reflected in a comparison of the groups with regard to their verbal interpersonal be- havior. hgngoys demonstrated significantly more behavior that was poorly controlled, passive-aggressive, and neg- atively dependent, while high boys showed significantly more positively assertive behavior. These differences were complementary to those for the two groups of mothers. L ! mothers showed significantly more dominant and protec- tive behavior and a trend toward greater rejection, while high mothers demonstrated significantly more autonomy granting behaviors that were both positive and negative in nature. Interaction sequences were analysed separately for son-mother and mother—son interaction. When son sender behavior was statistically "equated" for the two groups, 12! and high mothers evidenced different patterns of response to the same stimulations. Similarly, igy_and hiflh boys responded differently to the same mother sender behavir for the their r mother- intera: son aC' mentar Which assert Stanfc mCther 155 behavior, suggesting that the differences in interaction for the two groups were a function of the boys as well as their mothers. In their responses to each other, igg mother-son pairs tended to maintain a pattern of negative interaction and unconstructive activity. Both mother and son acted in ways which would serve to perpetuate a comple- mentary "dependent son-—controling mother" relationship, which stands in contrast to the more symmetrical, mutually assertive relationship of high mother-son pairs. Comparison of the two groups of mothers on the Stanford Parent Questionnaire (SPQ) revealed that igg. mothers scored significantly higher on rejection, incon- gistengy, punitiveness and physicalgpunishment, demands fgr aggression, democracy, and demands for conformity than did high mothers. Integration of the SPQ and behavioral interaction findings led to the interpretation that the love relationship for igg_mother—son pairs is not a secure one, that consequently mother and son must be continually involved with each other. The igg_mother's ambivalent feelings are manifest in her inconsistent and dominant, Protective behavior toward her son, while the igy boy's anxious dependency results in a passive-aggressive and negati result contro model incons intera accour tions} odolog study, actior 156 negatively dependent stance toward his mother. While the results lend support to the notions that boys' poorly controlled behavior is related to an aggressive parental model and that dependency in boys is related to maternal inconsistency and dominance, it was concluded that an interaction of parent effects and child effects can best account for the "dependent son--controling mother" rela- tionship observed in igg.mother-son pairs. Several meth- odological issues were discussed in the light of the present study, and a new approach was suggested for family inter- action research. Bathe: Becker Becker Becke Becks Bell Bing Bloc? REFERENCES Barker, R. G., Dembo, T., & Lewin, K. Frustration and re- gression: An eXperiment with young children. University of Iowa Studies for Child Welfare, 1941, ig, No. 1, 1—315. Becker, W. C. Family relations and child personality. Paper presented at the University of Minnesota, April 1962. Becker, W. C. Consequences of different kinds of parental discipline. In M. L. & L. W. Hoffman (Eds.), Child development research. Vol. 1, New York: Russell Sage, 1964. Pp. 169-208. Becker, W. C., Peterson, D. R., Hellmer, L. A., Shoe- maker, D. J., & Quay, H. C. Factors in parental behavior and personality as related to problem behavior in children. Journal of Consulting Psy- chology. 1959, 22: 107-118. Becker, W. C., Peterson, D. R., Luria, 2., Shoemaker, D. J., & Hellmer, L. A. Relations of factors de— rived from parent interview ratings to behavior problems of five—year olds. Child Development, 1962, i3, 509-535. Bell, R. C. A reinterpretation of the direction of ef- fects in studies of socialization. Psychological Review, 1968, Z§_(2), 81-95. Bing, E. Effect of childrearing practices on development of differential cognitive abilities. Child De- velppment, 1963, a3, 631-648. Block, J., & Martin, B. Predicting the behavior of children under frustration. Journal of Abnormal and Social Psychologx. 1955, hi, 281-285. 157 Brody. ()1 i) q Cran Eduard Feathe Freedr Fried] Gilde Golds Gran Hale: 158 Brody, G. F. A study of the relationship between maternal attitudes and mother-child interaction. Disserta- tion Abstracts, 1963, ii (4), 1728. Crandall, V. J., Preston, A., & Rabson, A. Maternal reac— tions and the development of independence and achievement behavior in young children. Child De- velopment, 1960, ii, 243—251. Edwards, A. L. Statistical methods for the behavioral sciences. New York: Holt, Rinehart, & Winston, 1961. Feather, N. T. The study of persistence. Psychological Bulletin, 1962, £2, 94-155. Freedman, M. B., Leary, T. F., Ossorio, A. G., & Coffey, H. S. The interpersonal dimension of personality. Journal of Personality, 1951, i9, 143-161. Friedman, S. T. Parental child-rearing attitudes and so— cial behaviOr of children. Dissertation Abstracts, 1964. £4 (8), 3415. Gildea, M. C. Maternal attitudes and general adjustment in school children. In J. C. Glidewell (Ed.), Parental attitudes and child behavior. Spring- field, Ill.: Charles C. Thomas, 1961. Goldstein, M. J., Judd, L. L., Rodnick, E. H., Alkire, A., & Gould, E. A method for studying social influ- ence and coping patterns within families of dis— turbed adolescents. Paper presented at the an- nual meeting of the American Orthopsychiatric Association, Washington, 1967. Grant, Q. R., & Kantor, M. B. Some limitations of the effects of maternal attitudes on child behavior. In J. C. Glidewell (Ed.), Parental attiudes and child behavior. Springfield, Ill.: Charles C. Thomas, 1961. Haley, J. Strategies ofypsychotherapy. New York: Grune & Stratton, 1963. 159 Hays, W. L. Statistics for Psychologists. New York: Holt, Rinehart, & Winston, 1963. Heathers, G. Acquiring dependence and independence: a theoretical orientation. Journal of Genetic Pay- chology, 1955, hl, 277—291. Hilton, I. Differences in the behavior of mothers toward first- and later—born children. Journal of Per- sonality_and Social Psychology. 1967, 1, 282-290. Hoffman, M. L. & Saltzstein, H. D. Parent discipline and the child's moral development. Journal of Per- sonality and Social Psychology. 1967, h, 45—57. Hollingshead, A. B. & Redlich, F. C. Social class and mental illness. New York: John Wiley, 1964. Holt, R. R. Experimental methods in clinical psychology. In B. B. Wolman (Ed.), Handbook of clinical psy- chology. New York: McGraw—Hill, 1965. Pp. 40-77. Kell. B. L. & Mueller, W. J. Impact and change: A stugy of counseling relationships. New York: Appleton- Century—Crofts, 1966. Lefkowitz, M. M., Walder. L. O., & Eron, L. D. Punish- ment, identification and aggression. Merrill Palmerhguarterly, 1963, 2, 159-174. Leton, D. A. A study of the validity of parent attitude measurement. Child Development, 1958, g2, 515- 520. MacKenzie, M. H. The interpersonal behavior of normal and clinic family members. Unpublished doctoral dissertation, Michigan State University, 1968. Martin. B. Family interaction associated with child dis— turbance: Assessment and modification. Psycho- therapy: Theory, Research and Practice, 1967, 1, 30-35. 160 McCord, W., McCord, J., & Howard, A. Familial correlates of aggression in nondelinquent male children. Journal of Abnormal and Social PsychologY. 1961, hi, 79-93 . Medinnus, G. R. The relation between several parent mea— sures and the child's early adjustment to school. Journal of Educational Psychology. 1961, ha, 153- 156. Merrill Bishop, B. A measurement of mother—child inter- action. Journal of Abnormal and Social Psychology, 1946) Q! 37’49.» Merrill Bishop, B. Mother-child interaction and the so— cial behavior of children. Psychological Mono— graphs, 1951, h§_(11, Whole No. 328). Moustakas, C. E., Sigel. I. E., & Schalock, H. D. An ob— jective method for the measurement and analysis of child-adult interaction. Child Development, 1956. g1. 109—134. Murrell, S. A. & Stachowiak, J. G. The family group: Deve10pment, structure, and therapy. Journal of Marriage and the Family, 1965, gz_(1), 13-18. Mueller, W. J. & Dilling, C. A. Studying interpersonal themes in psychotherapy research. Journal of Counseling Psychology. 1968, in press. Rau, L., Mlodnosky, L. B., & Anastasiow, N. Child-rearing antecedents of achievement behaviors in second grade boys. USPHS COOperative Research Project No. 1838. 1964. Raush, H. L. Interaction sequences. Journal of Person— ality and Social Peychology. 1965, g, 487-499. Raush, H. L., Dittman, A. T., & Taylor, T. J. The inter- personal behavior of children in residential treat- ment. Journal of Abnormal and Social Psychology, 1959,.§§, 9-26. 161 Raush, H. L., Farbman. I., & Llewellyn, L. G. Person, setting, and change in social interaction. II. A normal-control study. Human Relations, 1960, ii, 305-333. Rosen, B. C. & d'Andrade, R. The psychosocial origins of achievement motivation. Sociometry, 1959, gg, 185-218. Rosenzweig, S. An outline of frustration theory. In J. McV. Hunt (Ed.), Personality and the behavior disorders. New York: Ronald, 1944. Pp. 379—88. Rowland, T. S. Mother—child interaction in a situation structured to induce frustration in the child. Unpublished masters thesis, Michigan State Uni- versity, 1966. Ruebush, B. K., Byrum, M., & Farnham, L. J. Problem solving as a function of children's defensiveness and parental behavior. Journal of Abnormal and Social Psychology. 1963, hi, 355-362. Schaefer, E. S. Converging conceptual models for maternal behavior and for child behavior. In J. C. Glide- well (Ed.), Parental attitudes and child behavior. Springfield, 111.: Charles C. Thomas, 1961. Schaefer, E. S. & Bayley, N. Maternal behavior, child behavior, and their intercorrelations from in- fancy through adolescence. Monographs of the Society for Research in Child Development, 1963, ga (3, Whole No. 87). Sears, R. R., Maccoby, E. E., & Levin, H. Patterns of child rearing. New York: Row & Peterson, 1957. Siegel, S. Nonparametric statistics for the behavioral sciences. New York: Holt, Rinehart, & Winston, 1961. smitjl. H. T. A comparison of interview and observation measures of mother behavior. Journal of Abnormal and Social Psychology. 1958, El, 278—282. 162 Sontag, L. W. & Kagan, J. The emergence of intellectual achievement motives. American Journal of Ortho— psychiatry, 1963, aa, 532—535. Stendler, C. B. Possible causes of overdependency in young children. Child Development, 1954, 2;. 125-146. Vidich, A. J. Methodological problems in the observation of husband-wife interaction. Marriage and Family Living, 1956, ig, 234—239. Watson, G. Some personality differences in children re- lated to strict or permissive parental discipline. Journal of PsycholoQY. 1957, fig, 227-249. Winder, C. L. & Rau, L. Parental attitudes associated with social deviance in preadolescent boys. Journal of Abnormal and Social Psychology. 1962, fig, 418-424. Winterbottom, M. The relation of need for achievement to learning experiences in independence and mastery. In J; W. Atkinson (Ed.), Motives in fantasyy_ac- tion, and society. New York: Van Nostand, 1958. Pp. 453—478. Yarrow, L. J. Effect of antecedent frustration on pro- jective play. Psychological Monographs, 1948, §g_(6, Whole No. 293). Yarrow, M. R. Problems of methods in parent-child re- search. Child Development, 1963, ii, 215-226. APPEND ICES APPENDIX A “I E UL‘UJ.“ n—J. RATING INSTRUCTIONS TO TEACHERS Please rate all of the boys in your class on the three scales for which defini- tions and rating sheets are provided. These are: self-control; self-sufficiency; and achievement motivation. Those boys who are withdrawn or who are "over-controlled" (note the sections of the definition sheet which are marked by an asterisk) should be separated into a group by themselves as you do the ratings. The majority of your boys, however, should fall into one of the four boxes on each of the rating sheets. It is not expected that a boy will necessarily fall in the same square on all three scales. That is, a boy may be rated low on one scale, medium-high on another, etc. So that the ratings on each scale will be relatively independent of each other, please rate your boys on self-control, then proceed to self-sufficiency, etc. Al- though only the end groups are defined for each scale, the scales should be seen as more or less continuous dimensions ranging from "low" through "medium low" and "medium high" to "high." The definitions of the scales are: Teachers: 1. Self-control Poor self-control - This boy shows relatively little self-control. He has difficulty following rules, sitting still, and keeping his mind on his work. He may get out of his seat and move about the room, talk when he is supposed to be working, or bother others in the room. He may show angry out- bursts, tantrums, or whining when he is displeased. Generally he appears to act on impulse, with little regard for the consequences of his acts. Appropriate self-control - This boy shows a relatively large degree of self-control, but he is not so controlled or rigid but what he can be socially outgoing with his peers and show aggressive behavior apprOpriate to boys. He respects rules, pays attention, concentrates on his work, and does not bother others. He shows restraint in his behavior, seems to think before acting. However, he can still be spontaneous and act or express him- self when it appears appropriate to do so. *There may be boys in your c1assroom.who are "over-controlled," that is, they have such rigid self-control that they are withdrawn, lack normal spontaneity, or do not act or express themselves when it is apprOpriate to do so. Please put these boys in a group separate from.the others that can be rated on this scale. 2. Self-sufficiency - In rating on this scale it should be kept in mind that some boys, because the content of the work is more difficult for them, need more help than others. Consideration of each boy's relative ability for doing school work should help on these ratings. For example, a boy of relatively low ability who asks for a moderate amount of help should be rated higher on self-sufficiency than a boy of high ability who asks for the same amount of help. Low self-sufficiency - This boy does not generally do things on his own. He seeks an unusual amoung of help from his teacher and/or peers, much more so than his abilities would suggest was necessary. Whenever things be— come difficult, he looks to others to tell him what to do or to do his work for him. He has difficulty starting things and carrying them through by him— self. He may seek a lot of reassurance and affection from his teacher. g High self-sufficiency - This boy generally goes ahead on his own and does his work without seeking an unusual amount of help from his teacher and/or peers. He can fall back on himself when the going gets rough, and he tends to carry things through to their end. He does not seek a lot of reassurance or affection from others. But he can ask for help or informa- tion when it is apprOpriate to do so. 165 .166 *There may be boys in the class who never ask for help even when it is appropriate to do so. They may be withdrawn, their rigid independence may lead them to function in an unadoptive manner. Please place these boys in a group separate from the others who can be rated on this scale. Achievement motivation - These ratings should take into consideration the boy‘s relative ability for school work. A boy of lesser ability who aspires to the same heights as a more capable boy should be rated higher on achieve- ment motivation. Low achievement motivation - This boy shows little motivation to do well in his school work. He does not seem to be very concerned about his performance and does not put forth his best effort. He shows little per- sistence, giving up easily on a Job when difficulties are encountered. His poor motivation does not, however, keep him from being active in class. *There may be boys in the class who are so afraid of failing that they withdraw from competitive activities and from.the class in general. Please rate these boys in a group separate from those who have poor motivation but who are not generally withdrawn. High achievement motivation - This boy is highly motivated to do well in his school work. He often shows concern about his performance and tries to do his best. He is persistent, sticking to a Job until it is completed, even though he encounters difficulties. He does not appear to be afraid of failing, entering actively into competitive situations. 167 .mmsonw swoop qH mumpasq pamHoHMHv hapflmwam m>w£ hoe mwwao hook .um>oko: “30H aw m “sou azaooz odd nwfim adfiuoz on» a“ m Kmsopw swam map aw Gondflwgo w haopronummd mp dado: many .8 H 9...? a Mom .mawom =quHEEo .2? was. mo 8% 23 Samson 98 £03: 3:on a on» fi 33 finding 93 .8 ,pmou use oome nmnp “macaw Mach no Snow pom .madom mamas pooh mo moss map on HHHS swoop “hon amoomfig 950% can hop vmm3oa ASOh mo Mqfina .oawummflgp no n3oo no no on has pH “wasp Sonm vacuummfio on hue mmmHo ndflooapndm hoow mb¢ mHmB ho mwom HA< nmE nmflusswamz 33-5% 33 mmflgo mDOM ho mwom ,- .-.._ .W“- ~—.v-~. . ‘ homomoa momma llllllllll'l HOHUCOU'MHmm QPQQ mCZHBw£ has mmwao Hook «um>okon “on.nH m «30H endow: dad nwfim asadoz mop ma 0 «msoaw zmflm mop ma nondaano m haopmsfixonmmm on oaooz wasp “on wo anode m Mom .mamom znmpoafioo Ham: mop mo swam esp hanmdop was flown? masonw # mop ma omao ompoawno mop mo .pmou one 839 n93 «madam 55% no Eon» pg .395 $.30 .30» .Ho mono one on, H33 @339 309 ”~onan .90» one hop pwmsoa Hook mo ManB .oaoommfinp no asoo no no on has pH “wasp Bonk pompommflo on has mmwao poaooflwnmfl moo» é %Q< mHmB ho mfiom HA¢ 1168 . swam amE-§%m2 . 28.5% a 33 H A I.. .I... } ItI‘- /, / . / .. , n (x K / .. M ,/ x , /.. f. /./ /../. .. T // r. J //..x// . F, pm /. ./. .. _ “ -1HUVIILVm IL ,. ,‘ .. m .._ ... /,., ,, . I I .. w mmaqo moo» mo mwom . l . . . M . .. . .. . , _ w _ _ I. _ .u. _ , ||. noncooe compo Nocmaoflmsmnmnmm 8.8 mUZHBm£ has mwmao hock «ho>o:on Nana ma m «Soq adfiooz was gwfim adadoz map ma 0 .mdoam QMflm mop ma amndafino m haopmsfixoummo on dado: mane Kon Ho mmaao a mom .onom zmohdeSQ Haw: map mo ouflm one hammock mum gownx mmdonm : mop QH Oman cwnoaflno man %o ,meu 039 nomad soap “mason use» no EoSp pom .omaom mamas nook mo mono map on HHHS macaw when umoomfia Ado» dam hop pmokoa nook mo adage .oawommfiop no nzoo no mo on has we “wasp Scum pnohommfio on has mmwao paddoflpnmg noow mo< mHmB ho mwom QA< . ammo nmE-§..amz ._.S-§om: :3 n I I . I, I q z / u , .1 . I a . / / . / /_ fl / ,g,_ . / J / .. . z . / _ . .. , . _ / xx. 1. all; . / x. . //. x/ M a l m z/ I. .//.,.., . I -lezlllx..|_ um .. ..I _ mmiuo mDOH ho mwom - . ._.... .~-‘-——..._.- |||||||l honomoa momma cofium>fluoz ucwEw>mH£U¢ opflH moZHB