NCY OF ENTERACTEON IN NORMAL AND CUNIC FAMMES CONSSTE Thesis for the Degree of Ph. D. M!CH&GAN STATE umvmsm Marv Mame 1966 lflhSl. 3 1293 00012 HUN/III!IIIIHUIHW Iii/IIIHIIHII”(HIM/Lilli!!! This is to certify that the thesis entitled y CONSISTEI‘ICY OF IIJ'TERACTIOI‘I IN NOIIAL AND CLINIC FM .1 I L I E S presented by Marv Moore has been accepted towards fulfillment of the requirements for _lih'_D°degree inmul Psychology LJ‘J‘W Q M01 “422K ior professor 1* 11-25-66 Date 0-169 t_.-‘ . ,V—q . 0.- !P I v). 0“. l ‘V . ' . l 9. _._‘ -~. _- an“ c . .A' 'V .‘e ‘v ABSTRACT CONSISTENCY OF INTERACTION IN NORMAL AND CLINIC FAMILIES by Marv Moore This study was undertaken to explore two questions of family interaction: (1) Can we distinguish between "normal" and "abnormal" families in terms of interaction processes? and (2) How stable are family interaction processes over time? Eight normal and seven abnormal families (four or ' five members each) participated in two similar semi-structured interviews with an 8-10 week interval between sessions; two trained observers rated each family on an instrument con— sisting of fifty items, mostly five-point Likert scales. Normal families had no known history of psychiatric disorder or treatment and were obtained through labor union and church groups. Abnormal families were all drawn from the treatment waiting list of the Michigan State University Psychological Clinic. It was hypothesized that (1) judges' ratings of family interaction observed in the standard interview would reveal differences between the normal and clinic families, and (2) that these interaction ratings would be relatively similar in the first and second sessions. Nineteen items from the Family Rating Scale yielded , inter-rater reliability coefficients that were adequate; Marv Moore 'thirty-four from interview 2. The criterion of adequacy was that an inter-rater reliability coefficient be signifi- cantly different from zero at the .05 level (i.e.,en1r of .43 or greater for interview 1; .AA or greater for interview 2). Statistical analysis to test Hypothesis I issued from the reSponses to these items. Hypothesis II was tested by analysis of judges' reSponses to the seventeen items with adequate inter-rater reliability for both interviews. Regarding the first hypothesis normal families were differentiated from the clinic families by over-all pathology scores and a number of individual Family Rating Scale items. A picture of the normal family in this study emerged from the data. In comparison to the clinic families the normals are characterized by more inter-member agreement, more capacity for reachmg common decisions in an equalitarian fashion, less over-all anger but more tolerance of individual independence in thought and action, more interpersonal warmth and cohe- siveness, less manifest tension, ani a greater degree of happiness. A most significant complement to this description is that both parents in the normal families diSplayed more over-all satisfaction and effectiveness within their various family roles. The second hypothesis was tested in two ways. First, the correlations between interview I and 2 for each family ranged from .12 to .74 with an over-all mean correlation of .AI. Although these measures of temporal stability are not Marv Moore as high as pre-study expectations, they suggest that there exists a core of interaction consistency over time for both experimental groups. Smfliafinding lends support to the premise, frequently assumed in family research projects but seldom tested, that typical interactions are relatively permanent structures. Second, t tests between interview 1 and 2 group mean F J E : pathology scores revealed that normal families were rated i as significantly less pathological after the second inter- k view; no such differences obtained for clinic families. This interesting finding may reveal a basic difference between normal families and those which seek some form of psychiatric help, a differential capacity to profit from practice and perfect smoother ways of carrying out the experimental tasks when encountered a second time. Finally, evaluation of this study's findings led the author to suggest directions for further family interactional research. LMUl A) Q/iq‘wliu V491,): "Q g) ICIC, (A CONSISTENCY OF INTERACTION IN NORMAL AND CLINIC FAMILIES By I.“\'i \“ . .,/, MarV'Moore A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1966 DEDICATION To my wife Joyce ii ACKNOWLEDGMENTS I wish first to express my gratitude to Dr. Charles Hanley, Dr. C. L. Winder, and Dr. John Hurley for serving as members of my dissertation committee. A special word of appreciation is due to Dr. Lucy Ferguson, committee chair- man and friend, for helping me through the loneliest parts of this project. In addition I would like to thank Marv Conan, Myrna Friedman Lane, Tom Rowland, and Bob Tinker for the countless hours they spent rating families; without them this study could not have been accomplished. iii .» -"<~. U” “m r' j. —-v TABLE OF CONTENTS ACKNOWLEDGMENTS LIST OF TABLES. LIST OF APPENDICES Chapter I. II. III. IV. V. INTRODUCTION. Statement of the Problem. Relevant Literature . . Methodological Problems Connected with Measuring Family Interaction METHOD. Subjects The Interviewing Procedure The Rating Scale The Raters RESULTS Inter-rater Reliability Test of Hypothesis I . Test of Hypothesis II. DISCUSSION Inter- rater Reliability Hypothesis I. . . . Hypothesis II . . . . SUMMARY REFERENCES APPENDICES iv Page iii vi 24 28 28 30 35 36 36 37 42 45 45 so 55 58 64 “.45: J M's. saw-V... a. Table LIST OF TABLES Comparison of Normal and Clinic Family Groups on Several Composition Criteria Inter-Rater Reliability Coefficients for Likert Scale Items on the Family Rating Scale . . . . . . . . . . Inter-Rater Reliability Coefficients Within Each Family for Both Interviews . Over-all Pathology Scores for Individual Families for Both Interviews T Ratios Between Group Means on Consistently Rated Items for Both Interviews Temporal Stability Correlations Between Interview 1 and 2 for Each Family on 1? Consistently Rated Items. T Test Comparisons Between Interview 1 and 2 on Each Item for Both Experimental Groups. . . . . . . . . . Page 29 37 38 39 4O 43 44 .__.‘-..‘ LIST OF APPENDICES Appendix A. Specification of Families Sampled in this Study B. Rating Scale Used in the Pilot Study. C. Family Rating Scale (FRS) Used in Final Study . . . D. Intercorrelations for Consistently Rated Items During Interview l. E. Intercorrelations for Consistently Rated Items During Interview 2. vi Page 65 68 73 79 81 CHAPTER I INTRODUCTION Statement of the Problem Recently the emphasis in psychiatry and psychology has shifted from the study of intra-psychic processes to the study of interpersonal processes, especially as they occur naturally between individuals. In this context, there are increasing attempts to classify and understand the inter- action of married couples and members of families as well as ongoing groups in industry, group psychotherapy, and other "groups with a history." (Framo, 1965, p. 265). Study of ongoing relationships provides an opportunity for a new type of experiment with human behavior focused on variables which are often eliminated from studies of individual be- havior. This paper will investigate two hypothesis con- cerning family interaction. Basic to this investigation of the family is the assumption that "normal" and "abnormal" families differ on measurable interaction variables. The present study attempts to add support to recently increasing evidence for this prOposition. Measures of interaction consist of ratings by trained judges on several variables of "normal" and "abnormal" families as they participated in a structured l interview; More specifically, HYPOTHESIS l was: Ratings of family interaction observed in the standardized inter- view will reveal differences between "normal" and "abnormal" families. Another assumption, central to understanding psycho- pathology, at least within the framework of most major personality theories, is that interpersonal patterns of behavior persist in families for years and influence members' expectations of, and behavior with non-family individuals. To optimally study this assumption one needs to observe families logitudinally. Such expensive and painstaking studies are absent from the family research literature; but less excusably, so are studies assessing consistency of inter- action measures over very short time spans, say a few weeks to a year. This study attempted to measure the temporal stability of the interaction measures employed. Each family was engaged in the structured interview twice with an inter— vening interval of eight to ten weeks. HYPOTHSIS 2 was: Ratings of family interaction observed in the standardized interview are relatively similar in a first and a later session. Relevant Literature The Contributions of Small—Group Research In an attempt to establish an objective science of the nature of small—group interaction social scientists have contributed to a growing, voluminous literature; Hare (I962) - - '=W~;iw.:s'8u=e Cites over 1300 relatively recent studies pertinent to the small group. Without attempting to provide a comprehensive overview of the small-group literature, selected issues and findings that appear relevant to the family dynamics are reported here. In reading the small—group literature a most ironic fact emerges. A11 workers in this field readily agree that the family is the most basic of primary groups, but a study utilizing the family as its subject remains a rarity. Some of the research on groups focuses on the develop- ment of conceptual theories of groups and the inherent dimen- sions of group dynamics (e.g., Borgatta, Cottrell, and Meyer, 1956). Other research seeks solutions to social problems such as diminishing intragroup conflict, enhancing the learning process, changing attitudes, etc. As more refined techniques of measurement evolved some researchers have pro- duced more precise examinations of empirical problems such as decision making, conflict vs. cooperation, group pressures on the individual, interaction patterns, etc. Practically all systematic studies obtain their data from ad hgg groups, groups without an interaction history and no expectations of future relationships after the experiment. Without denying the obvious contributions of ad hgg group research, Framo (1965) appropriately points out that crucial differ- ences exist between ad hgg groups and the family. Nothing can change the fact that one’s parents and siblings will always be such. Although feelings of guilt, anger, jealousy, and warmth probably develop in all groups, they have different significance in the family. Subjects in an dd hdd_group do not have to live with each other after the experiment; possible consequences or after—effects of an experiment are radically dissimilar for temporary and for family groups (e.g., a family group may Openly discuss the interpersonal episodes of the experimental session or effect recriminations against family members who divulged family secrets to the experimenter.) A tOpic of continual interest in the small-group literature is that of leadership. Cartwright and Zander (1962, p. 496) state that: It is unfortunate that most of the careflxfly controlled studies of leader behavior have been conducted with temporarily organized groups where, almost of necessity, members are not concerned with the preservation of the group. In discussing leadership they state that "nearly every con- ception of leadership contains the notion that a true leader exerts more influence on the group and its activities than does the average member” (p. 493). Influence can be positive or negative but most leadership studies deal only with positive influence; therefore, they yield little that is directly applicable to pathological leadership in a disturbed gfanxily. In short, the concept of leadership applied to the fwiauily becomes immediately more elusive. For example, if an €3Xperimenter asked members of American families who is tm>sss at home, the majority would hastily reply, "father.” BUt Tlhose working professionally with families know the calllbtral stereotype of father "wearing the pants" often does :_e :fitt't'i- ' ' not apply. Less widely realized is the phenomenon of the children sometimes taking over leadership aSpects of the parental role (Schmideberg, 1948). Parsons (1955, Ch. 2) suggests that the structure of the nuclear family be differ- entiated on two dimensions, one of power, and the other of instrumental vs. eXpressive functions (after Bales, 1953) i.e. goal directed, working behavior vs. integrative, social— emotional behaviors. When mother assumes the instrumental role and father the expressive role definite implications .I'M—"l. exist for the sexual identity of the children, and when children are parentified into either role the consequences are far reaching (Framo, 1965, p. 414). Another leadership pattern relevant to the family, and not yet adequately ex- plored in the small-group literature, is passive mastery. A passive leader controls by quiet means, exerting his influence by barely noticeable, often subtle, signals-- signals largely undetected by any existing global observa- tional technique. A concept closely related to leadership is that of power. A classical eXperiment in this area done by Mills (1953) revealed that a three-person group tends to break irltCD a dyad of coalition and a third party in structuring tdfie ‘power relations; often the third party becomes a scape- goait; for anxieties of the dyadic relationship. As Mills Suggested, Strodtbeck (1954) repeated the experiment with famjgl3; as the experimental group. Strodtbeck employed the Revealed Differences Techniques (1951) WhiCh requires a 6 family to reach common consensus on problems about which they independently disagree; he did not conform Mills' main findings and suggested that the split into dyad and monad is less pronounced in families as compared to.§§.§9§ groups. In contrast to Mills‘ finding are Wynne (1961) and Haley's (1962) assertions that coalitions and Splits are the very essence of maintaining family equilibrium. Coalitions in families are intricate maneuvers; a family member allies himself with all other family members at one time or another, depending on the satisfactions in- volved or being sought. Discovery of central intrafamilial alliances is difficult because such relationships occur surreptitiously, usually behind defensive Operations, such as more superficial alliances, as any family therapist will attest. Strodtbeck's (1954) conclusion that alliances are less frequent in families than in 28.292 groups simply con- tradicts clinical eXperience. Understanding the discrepancy resides in examining Strodtbeck's measure of a coalition; he used the Bales system of interaction analysis (1950) whiCh catagorizes the manifest content of subjects' state- nmnits, irregardless of deeper levels of motivation. Sup- }301“ting this assertion are Mills' (1954) findings in a seaccond study of coalitions in three-person groups where he Sbugégested that even in dd hoc groups subjective, affective allhiéances are of major importance. He summarily states: 111 a role structure of some stability, the structure CXf personal, emotional attachments (positive or 11€3gative) is stronger than the structure of common values and beliefs that are relevant to the purpose of the group, and these structures are stronger than the pattern of manifest interaction between members (p. 667). Another concept of import to the small-group researchers is ”group cohesiveness." This term usually includes the con— ditions under which groups have appeal, command loyalty, strive toward common goals, and are united against extra- group attack. Several aspects of the concept "cohesiveness" have been studied in groups of superficially related members (e.g., Festinger, Schachter, and Back, 1951). Cartwright and Zander (1962) postulated that a group's attractiveness rests on the extent to which it is need satisfying to individ- ual members; on the other hand, Fouriezos, Hutt, and Guetzkow (1950) found that interaction in 2E,DEE groups characterized by high self-oriented need produces more conflict, less cohesion, and less member satisfaction. Gross (1956) reported that groups characterized by "symbiosis" (defined sociolog- ically as peOple cohering as a group when each has something needed to give the other) as contrasted with "consensus" (relationships cememted primarily by agreement to a set of values) exhibit greater cohesion. Gross said: As long as those needs persist, and so long as each has no easy alternative of satisfying those needs, then the two will be linked. This does not mean they will necessarily like each other; it does mean that they will remain united whether they like each other or not. And herein lies the strength of the symbiotic tie (p. 177). Although Gross does not mention the family it suffices to say his conceptualization describes the family better, per— haps, than any other group. Festinger, Pepitone, and N€wmmfl>(l952) present evidence that when ”de—individuation" occurs in a group (i.e., members failing to notice or validate<fiflwr members' ideas or feelings) inner restraints are reduced, members feel free to indulge in behavior from which they usually refrain, and they are more satisfied with the group. In contrast to Festigner et al.‘s finding is the author's suspicion that de-individua— tion in the family fails to make family members more satisfied with such pathological interactions; however, an experimenter will not immediately find family members reporting their diSpleasure since to do so would bare anxiety assumed to be intolerable. French (1941) indicates that organized groups with a longer history developed more ”We-feeling” and are less likely to break up than organized groups. Simple application of this principle to family living means viewing the family as a cohesive force of great power. In more formal groups (job, clubs, etc.) when experiences become cumulatively un- bearable, one can withdraw his membership without earth- shaking consequences. Such interpersonal anxieties in the family may actually effect the Opposite; perhaps the more frustrating the process of growing up, the more difficult it is to leave the family psychologically. Sherif and Cantril (1947), Asch (1951), and Crutch— field (1955) investigated the persuasive effects upon an individual of group pressures to conform to standards of the group even when the individual perceives the standards to be Obviously contrary to fact. ”Group pressure" in these experiments was the effect Of staged majority Opinion in forcing an unsuspecting subject to question his own diver- gent percepts. In all the conformity experiments large per- centages of the subject pool consistently yielded to group pressure, even when the bogus position was manifestly wrong. Such pressures and forces, in the guise Of overt demands and just as Often subtle, covert directives, similarly exist in families and are even more compelling. Family persuasive techniques range all the way from physical punishment to subtle and sOmetimes manipulative threats of love withdrawal. Picture a small child who is pressed to take sides in a parental argument-—with the implication Of rejection by the parent he doesn't support. A family containing a schizo- phrenic child represents a most extreme state of affairs where the parents consistently force the Offspring to deny and distort reality in a flagrant manner (Haley, 1959). Studies of formal communication patterns have stressed the power-and—status dimension as being crucial in develOping communication channels and networks (Bavelas, 1950; Kelley, 1951, Leavit, 1951; Ruesch, Block, and Bennett, 1953). Communication networks in families demand a more complicated conceptualization. Every family poses its own special rules, channels, and styles of communication (Epstein and Westley, 1959). Families exist as intricate, interrelated systems * ~r'.s- .w..,:~;-afir..w,ro' . -.~ '. A" q'IWmfib _. 1O with years of practice communicating in their particular fashions; an outsider can find it difficult indeed to trace meaningful messages between family members who themselves reap sheaves Of affective content from an unobtrusive gesture. Therapists working with families also find that certain areas of information are reserved for Special mem- bers on Special occasions, and that attempts to induce airing Of family secrets raise stout resistance from all involved. Framo (1965) Observed that as far as he knows the family treatment setting is the only one where all family members arernnflnd to discuss their intra-family secrets in front of the other members. Related to the prevalence of family members monumental resistance to attempts to examine communi- cation channels is the recurring finding that meaningful communication between husband and wife in pathogenic families is almost non-existent, both partners sending messages to each other through a selected child. Related to such growth stifling communication patterns is Bateson's formulation of the "double-bind” (1956 and 1963), now a concept commonly heard in the working psychiatric vocabulary. The present review of small—group literature relevant to family interaction reflects the interests Of the author; selection of pertinent articles was greatly aided by Framo's (1965) review of current research on family dynamics. Where- as the aim of small-group research is often the evaluation of methods for increasing group interaction efficiency (measured byixmrmuch SS learned, how fast problems get ll solved, etc.) the goal in family interaction research is more encompassing, including not only task efficiency but understanding emotional integration at various levels in terms of satisfying individual affective needs. However, though the clinical family researcher Observes more levels and intensities of interaction for a longer time, his cur- rent methodologies lack the precision of the small-group approach. Such methodological precision can prove immensely valuable to family researchers if it doesn't substitute for exploring relevant behaviors. Review of Family Interaction Studies This section will review and evaluate the few sys- tematic studies based on direct observation of family inter- action. Evaluative emphasis is on method as well as results. The earliest family interaction research consisted of direct observation of the mother-child dyad, usually in a free play situation. Representative Of this approach are papers by BiShOp (1951) and Moustakas, Sigel, and Schalock (1956). These studies, even though contributing carefully thought out schedules for categorizing facets of mother-child interactions, failed to capture the transactional meaning of the relationship. Finally, these authors essentially discounted the effects of other family members on the dyad studied. Drechsler and Shapiro (1961) suggest a technique for direct Observation of family interaction in a child-guidance 12 clinic. After identifying the family's presenting psy- chiatric problem, they gave the family a questionnaire of twenty items containing factual and fantasy material (e.g., "What sort of things does your family argue about? If you could change one thing about yourself and the other members of your family, what would it be?"). The authors asked each family to discuss the items together, hoping to glean insights about covert interaction patterns. Drechsler and Shapior (1963) Since reported how they sampled and analyzed the data yielded from the aforementioned procedure. They illustrated their method of comparing clinical and statis- tical analysis of the same data to test interactional hypotheses. Strodtbeck (1951) pioneered the study Of husband— wife interaction, develOping in the process the "Revealed Differences Technique," and thereby provided the impetus for several family interaction papers. Strodtbeck originally asked each couple to choose three families with whom they were familiar; he next separated the couple and had them Specify which of the three families best fit a series of twenty—Six conditions: for example, which family has the happiest children. Finally, he brought the couple back to- gether and required that they reconcile their differences, and decide on a jointly derived choice. Using Bales (1950) interaction categories, Strodtbeck found that the spouSe who said most tended to win more final decisions, and also that the most talkative partner more frequently asked 13 questions, gave Opinions and analyses, and made rewarding comments, whereas the least talkative partner more often passively agreed but tended towards more overt signs of frustration and aggression. Based on their eXperience with marital fighting styles and games Bach (1967) and Berne (1964) would certainly ask: does the loudest, most talkative individual in an argument necessarily win the decision? Tharanswer negatively and the author Concurs. Investigators employing the Revealed Differences Tech- nique subsequently studied a variety of problems. March (1953) examined husband-wife interaction around political issues. Kenkel and Hoffman (1956) asked each mate to pre- dict how the other would behave in a session where they would decide jointly how to spend $300.00; they discovered that not only did the marital pairs judge poorly the roles they would play, but that even after concluding the session they failed, by and large, to recognize the parts each played in the previous interaction. Strodtbeck (1954), in a study mentioned earlier, provided the forerunner to later systematic family inter— action studies when he applied his Revealed Differences Technique to three person family groups, including adoles- cent Sons as well as parents. The experimenter gathered disagreements from each member's independent reSponses to a questionnaire of alternative solutions to parent-son conflicts. The three-member families (48 in all) dealt with nine disagreements (three with mother and son against father, 14 three with father and son against mother, and three with mother and father contrary to the son) and selected one alternative in each that best represented the family's thinking. Family discussions occurred without the experi- menter present but were tape recorded. Using Bales (1950) interaction analysis Strodtbeck compared family power relations with those in 28.222 groups. Later he attempted to relate his findings to cultural factors and achievement, noting that families, even when in an Obvious state of dis- agreement attempted to give the experimenter the impression that they "never really disagreed" in the first place (1958, p. 176). This finding and Wynne's (1958) concept of "pseudo-mutuality," based on clinical experience, are con- gruent with each other. I The theme Of power stimulated other indirect studies of family interaction. An early interactional study by Garmezy, Farina, and Rodnick (1961) utilized families with a schiZOphrenic member. Farina (1960) used the structured Situation test from the above study to test experimentally Rodnick and Garmezy's (1957) finding that maternal dominance was associated with poor premorbid adjustment of the schiZOphrenic son, and father dominance with good premorbid adjustment. He studied 36 pairs Of parents, divided into 3 groups of 12 pairs each; one group had sons with pre-morbid adjustment characterized by isolation and asexuality; another group's sons had mostly been married and had numerous friends; and the controls' sons were hOSpitilized for 15 tuberculosis. After interviewing the parents individually and giving them each the Parental Attitude Research Inven— tory (PARI), Farina directed them to tell how they would handle a series of problem situations between parent and child. Finally, he requested each pair of parents to reach agreement about how they would handle the problem as a team-- in the presence Of the investigator. Interaction analysis derived from indiceS of dominance and conflict (e.g. who Spoke first and last, length of Speaking time, "yielding" behavior, interruptions, disagreements, and aggressions). Farina confirmed statistically that good premorbid patients had more assertive fathers than did poor premorbids; and that the parents of poor premorbids interacted more conflic- tually. The author also found that analysis of parents' reports (PARI and direct questions) about family dominance patterns failed to discriminate between the groups; all the groups tended to report the cultural stereotype of paternal dominance. Farina and Dunham (1963) later replicated this study (with the same results) and included the schizophrenic son as a third party in the Observed interaction. Caputo (1963), investigating the dominant—mother, passive—father notion in families with a schiZOphrenic mem— ber, also demonstrated the superiority of direct observation of interaction over pencil and paper tests. After filling out Osgood's Semantic Differential (1957), the parents of twenty male schizophrenic patients and the parents Of twenty "normal” males discussed ten items on the Parent Attitude 16 Inventory which they previously answered divergently. Caputo derived relatively benign pictures of interaction between parents of the pathological families, whereas analysis of ongoing interaction with the Bales method yielded consider— able antagonism and mutual hostility between the same parents. The results questioned the veracity Of the maternal-dominance, paternal-passivity theory. Whether parental authority was Shared by mother and father without conflict discriminated between "normal" and schiZOphrenic groups, but indices of the locus Of authority did not. Bachove and Zubaly (1959) studied role differentiation in nineteen normal families by comparing interaction patterns (scored by the Bales technique) with self-perceptual data (Leary's Interpersonal Check List, 1957). All three-member families (parents and one male Sixth grader)carried out the standardized tasks of problem Situations and joint TAT stories, on which they conversed until conjoint agreement. Comparing Bale's category findings from their families to those established by Bales from twenty-four different peer groups (Bales, 1958), Bachove and Zubaly found that the peer groups tended to agree and disagree far more frequently than the families. They explained this discrepancy by suggesting that family disagreement scores funneled into the tension category and that low agreement scores meant a greater tendency Openly to stick to one's Opinion in the family than in the peer group. A task leader (usually fa- ther) and a social emotional leader (mother) typically a..— 51.3.," ' "“- ""‘J-"m l7 evolved from the interaction, and the children expressed most Of the negative behavior. Levinger (1959), using the same design and data as Bachove and Zubaly (1959), with clinic and normal families found that mothers in clinic families participated most often and exhibited significantly more emotionally negative be— havior. This study supported the contention that reversal of male-female roles hampers emotional growth in the chil- dren. Levinger also found that marital satisfaction, defined as a relatively low discrepancy score between each Spouse's "real" and "ideal" Interpersonal Check List descriptions of his partner, correlated positively with the partner's satisfaction with himself. Ferreira (1963) executed a carefully controlled study comparing decision—making in normal and pathological families. Twenty-five normal and twenty-five abnormal (containing a psychiatric patient) families participated. A family was defined as a mother, father, and a child of either sex over age ten. Two phases comprised the test Situation. First, each family member reached an individual decision about three emotionally neutral items (e.g. "If you were going to take a trip to Alaska next month, would you rather go by train, car or boat?") and rank ordered reSponseS in terms of pref— erence. Second, the whole family attempted to reach a deci- sion on the same items, jointly considering the preferences Of individual members. Several kinds of family decisions emerged: unanimous decisions (where the family choice l8 correSponded with the individual choices of every member), majority decisions (where the family choice corresponded to the individual choices of two members), dictatorial decisions (where the family choice equaled only one member's preference), and chaotic decisions (where the family choice correSponded to no individual preferences). The data revealed that all families Spontaneously agree more often than chance expec— tations, but also that significantly greater agreement be- -.ar— fl..- tween individual and family preferences existed in the normal family group. In regard to the dictatorial decisions made, sex differences existed in the normal group but not for the abnormals. Father exceeded mother if the child was a boy and mother exceeded father if the child was a girl. In a later study Ferreira (1965) enlarged the preceding study on decision making in the family. Employing the Same eXperi- mental procedure with a much Eager sample (125 families: 50 normal, and 75 abnormal) they investigated two new variables, decision making time and decision appropriateness, in relation to expressed individual need and preferences, as well as the variable studied earlier, amount Of Spontaneous agreement. Ferreira reported significant differences for several hypotheses: First, he replicated his previous finding concerning differences of Spontaneous agreement between normal and abnormal families. Second, abnormal families took more time to reach a joint decision. Third, also as predicted, apprOpriateness of family decisions was less for the abnormal families. 19 In another inventive study Ferreira (1963) investigated patterns of rejection and eXpectancy of rejection in 55 family triads (25 normal and 30 abnormal). First, all fam- ily members colored ll flags with crayons; next, each member judged the other's productions and "threw away those he dis— liked for any reason whatsoever" (p. 237). Finally, each member of the triad guessed, based on his knowledge of the others, how many of his flags the other members threw away. The data confirmed the hypothesis of more rejection and ex- pected rejection in the pathological family group. On the basis of these results Ferreira postulated that in healthy families an individual tends to expect rejection commensurate with the amount he diSplays. In the abnormal families the attitudes of "two eyes for an eye" or "no tooth for a tooth" tend to replace the "eye for an eye" principle (p. 244). Winter, Ferreira, and Olson (1965) asked 125 three- member families to produce conjointly three TAT stories, each story based on a Specified three-card sequence. Data analysis derived from the Arnold system of Story Sequence Analysis (1962) on which judges rate sequential themes on a five-point scale of imputed emotional maturity. The scale purports to measure "the degree to which S's motivation, behaviors, and habitual ways of viewing the world are con- gruent with good mental health" (P. 394). Analysis of results revealed that the Arnold score successfully differ- entiated normal from abnormal families, but the three abnormal sub-groups (Delinquent, Maladjusted, and 20 SchiZOphrenic) did not differ from each other. Utilizing the same sample of family triads and their TAT stories, Ferreira, Winter, and Poindexter (1966) examined seVeral new variables. Data from sessions taped but not observed yielded the following results. Abnormal families required significantly more time to complete their joint TAT stories. With one exception no differences Obtained between or with- '1'! in groups in regard to who talked most; children in schizo— phrenic families talked least of anyone. Examining the mun-p natured-1.4.0 .- . »- -. . possible sequences of ”who speaks after whom,’ contrary to Haley's (1964) earlier finding, the authors Observed no sig- nificant variation from randomness. Although predicted, no differences between groups in amounts of conversation overlap (times when two or three voices simultaneously occurred on tape) obtained. However, significantly greater amounts Of silence existed in the abnormal families as hypothesized. Finally, the authors observed that the Spokes- man who reported the family's stories usually talked most during the process although the family Often never overtly agreed who would later be the Spokesman. In a valuable experiment Fisher, Boyd, Walker, and Sheer (1959) compared the interaction approach with the individual approach. These investigators compared the parents of 20 normal, 20 neurotic, and 20 schizophrenic men on a battery of measures assessing individual function- ing and patterns of Spouse interaction. AS hypothesized the parents of normal men were individually less disturbed 21 than the parents of neurotics and schizophrenics. However, an interactional measure, derived from couple's formula- tions of a joint TAT story, proved to be the only measure discriminating parents of neurotics from parents of schizo—I phrenics. Parents of neurotics, in creating their joint TAT stories, interacted with less disagreement, less ambi- guity in exchange of Opinion, and with more total volume of communication than the parents of schiZOphrenics. 0n the basis of their findings Fisher 33.3.1- suggested that schiZOphrenia results from the combined maladjustment of both parents as they interact with the child. They implied from flwfir data that if a husband and wife combined forces in a relatively congruent manner, they compensated to some degree for their individual pathologies. Framo (1965) made the relevant statement, in discussing this study, that the "lack of clarity" of communication between parents and between each parent and the child proves more pertinent to the develOpment of schizophrenia in the child than Open parental disagreement and conflict (p. 429). Haley (1964) analyzed conjoint TAT stories Of 80 three— member families (40 normal and 40 abnormal) borrowed from Ferreira and Winter's (1965) study. The 40 abnormal families all included one member suffering from schizophrenia. Haley's approach was simple but revealing. He tabulated the fre- quencies of all possible sequences of "who Speaks after whom." Armed with these data he attempted to answer three questions "basic to family research" (p. 42). The first question, can 22 we demonstrate that the family is an organization following repetitive interaction patterns?, Haley answered affirma— tively by demonstrating that in all families the variations of conversation sequences differ Significantly from random eXpectations. Secondly, he found that the 80 families dis- tributed themselves on a scale of interaction randomness in an approximately normal fashion. Haley investigated the third question, "Can we on some scale, differentiate a dis— turbed family from a normal one?"(p. 42) by hypothesizing that his normal family group interacted in a fashion more closely approximating random sequences than the abnormal group where existing pathology produces more rigid relation- ships. A strikingly significant difference (p : .00003) between the normal and abnormal families occurred on Haley's scale Of interaction randomness, but not without overlap between the groups. Haley also presented pilot study data lending support to the future possibility of measuring thera- peutic change via interactional sequence analysis with high reliability. Lennard, Beaulieu, and Embrey (1965) also studied communication sequence in 20 three-member families (10 normal and 10 families with a schizophrenic son). Data were tran— scribed accounts of each family discussing three tOpics for fifteen minutes each (e.g., "When a boy needs a helping hand with homework, is it better for mother or father to help out?")- Sequence analysis of the three party interactions found that in the abnormal families significantly less 3’5." 9: 23 communication flowed from son to father and vice versa, and mutually between mother and father. Viewing the family as a communication control system, the authors examined amount of intrusion (when a third member forces himself into an ongoing interaction between the remaining two parties) in their samples. SchiZOphrenic sons and their mothers exhibited Significantly fewer intrusions than the same member in the normal families, and fathers in the same family; and success- ful intrusions (where the intruder succeeded in changing the conversation topic of the other two parties) were sig— nificantly less in the schizophrenic families. These findings support the dominant—mother, passive-father theory of inter- action in schiZOphrenic families. Two recent studies present new techniques of family interaction analysis; both assess transcriptions of families reSponding to a semi-structured interview including tasks such as "Plan an activity that you could all do together as a family." Terrill and Terrill (1965) categorized individual Speeches with a re—worked version of the Leary Interpersonal Check List (1957) and presented interpersonal profiles of family members which this complicated procedure yields. Riskin (1963 and 1964) utilizes Skilled clinicians, exper- ienced with families, to micrOSCOpically rate family members' Speeches on several dimensions: communication clarity, topic shifts, agreement with previous Speeches, commitment to stated positions, and affective intensity. An experienced Clinician listened several times to a six minute tape that 24 judges independently rated on the above variables; and according to Riskin, he missed a good deal of significant detail gleaned from the microscopic analysis. Quantitative studies with both these promising techniques are yet to be done. Methodological Problems Connected With I Measuring Family Interaction This section deals with selected methodological issues evolving out of the previous review of family interaction studies. Sampling,Situations If the goal of presenting a family with tasks to per- form or issues to resolve is revelation Of the family's inter- action style and deeper motivational systems, then clinical researchers presumably must employ some method that circum- vents natural family defensiveness and the strong need to present the family as normal. All family interaction research to data apparently assumes the implication in the projective hypothesis; that is, that a family when presented an ambiguous stimulus or a difference to resolve can do naught but _ reSpond with its most typical patterns Of behavior. Several factors lead one to question the validity of this assumption. To begin with, the presence of the eXperimenter during the experimental session (in person or represented by a tape recorder) presents a powerful stimulus which awaits systematic investigation by family researchers. Assuredly, “W 25 the experimenter's presence introduces a new variable into the family's interaction behavior not dealt with outside the laboratory setting. Besides the artifacts Of the experi— mental situation remains the aforementioned fact that family members have to live with each other after the interview concludes. Certainly their public responses are tempered by knowledge or expectations of possible consequences or retal- iations once they return home. Even if families could be counted on to consistently remove their masks and interact as they do at home, another question challenges the family researchers. DO the commonly employed experimental tasks accurately sample the Significant issues of family living? NO one yet has attempted to Specify a domain Of tasks and Situations which family studies might sample. Most issues presented experimentally to families only accidentally represent conflicts the family is encoun- tering at the time. In this connection Framo (1965) suggests the blueprints for two possible experiments: A meaningful experiment would require that each family be presented with the controversies it is in- herently struggline with, not with abstract contro- versies which result in polite play—acting. Prelimi- nary study of the family Should reveal its Achilles's heels (p. 433). Investigation of symptom-free families in a series of exploratory sessions, dealing with the normal crises every family has to deal with, is one research project worth doing (p. 455). 26 Interaction Units NO matter what family life issues one samples there remains the difficult task of appropriate selection and measurement Of the interactional behavior. Most attempts to measure family interaction have yet to go beyond the categorizing Of interaction units and assess the underlying motivational processes. For example, the Bales system Of interaction analysis, probably the most comprehensive non- content system for quantifying face to face interaction, categorizes largely the manifest level Of interaction; un— conscious determin ntS Of behavior are not tapped. Further- more, the Bales Observer considers only the preceding act in classifying a present one--therefy exclusing emotional overtones highlighting an interrelated sequence of behaviors. Whatever method the family investigator employs in assessing interactional variables, he faces the central problem Of time sampling; he cannot just observe and quantify everything in signs. The careful researcher needs :3 ('1 to select his variables according to a consist~nt theory, try not to lose too much in translating from corcepts to Operations, and attempt to discern how often and when an instance of any given variable occurs. The selection of observation intervals must be done, over—all, SO as to give a reliable measure Of the pOpulation Of events. Complicating this procedure is the fact that significant behaviors do not necessarily occur repetitively. Some Of the most 27 emotionally significant moments of interaction in psycho- therapy occur rarely, maybe once. Selection of Supjects Sampling procedures in the selection Of subject fam- ilies continues to pose a problem. We just do not know enough yet about the universe of family dynamics to draw' definitive and representative samples. Sampling accuracy assumes consistency; the question of stability of family interaction over time, so crucial to our theories of psycho- pathology, remains essentially an empirical unknown, urgently awaiting systematic study. The fact emerges, then, that we are presently very far away from experimental manipulation of significant family life variables; however, certain avenues Of approach stand clear on the horizon. Such transactional phenomena as Kell and Miller's (1966) "eliciting behaviors," Bach's (1967) and Bateson's (1962) "double—hind” situation are concepts awaiting operational translation; furthermore, all these concepts are assumed to be relatively permanent inter- actional processes. Such assumptionscell for empirical validation. CHAPTER II METHOD Subjects All families in this study consisted Of four or five member units (both parents and two or three children) meeting the following criteria for acceptance: (1) every family lived together for at least four uninterrupted years previous to participation in the research; (2) children ranged from ages 7 to 17; (3) all families included at least one male child between 8 and 13; and (4) they met the criteria below for inclusion in one of the two samples studied. The experimenter defined his two samples as follows: Normal group, composed of 8 families where no member ever received, or was recommended to receive, any type of psychi- atric treatment for an emotional or nervous disorder. Normal families were Obtained from two sources. Three units volun- teered as a result Of a call for subjects at local labor union meetings and five volunteered after being recommended by their minister as representing the "most emotionally mature" families in his congregation. For their COOperation all normal families received $10.00 for each Of the two interviews. Clinic grodp, composed Of 8 families waiting for psychotherapy at the Michigan State University 28 29 Psychological Clinic, involving both parents and sometimes one or more children; no family received any treatment during the course Of this experiment. All clinic families initially contacted the clinic because a male child between age 8 and 13 had been referred for underachievement and/or lack Of behavior control in school. Clinic families received no remuneration for their part in the project; the treatment agreement at the Michigan State University Psychological Clinic is that families participate in some ongoing research project Since no fees are charged for psychotherapy. A total Of 16 families was tested in this project, 8 normal and 8 clinic. One clinic family, however, refused to return for the second experimental session. InSpection Of Table 1 reveals that the two groups are essentially simi— lar in composition except for mean level Of fathers'education which is 1.2 years higher in the Clinic sample. This differ- ence, however, is not Significantly different from chance expectations for this small a sample. Corresponding statis- tics for each family appear in Appendix A. TABLE 1.—-Comparison of Normal and Clinic family groups on several composition criteria. Mean years Of com leted Mean Mean educagion number Of age of children children Father Mother per family per family ’— Normal Family Group 12.8 12.2 2.3 10.6 Clinic Family Group 14.0 12.1 2.6 10.4 f «I; ass-s: ire-3-5%.: - w Innate" 30 The Interviewing Procedure After greetings and preliminary remarks to place every- one at ease the interviewer (who saw all families for both sessions) introduced the family to the first conjoint task. Preliminary remarks included reiteration of the fact that two raters would be viewing them through the one-way window and that the purpose Of the project was to provide informa- tion which would hOpefully increase our Skills in helping families. The tasks on the interview schedule occurred as follows: Task 1: The interviewer saw each family member just long enough to ask him the question, "At this point in time what changes would you like to see made in your family, as a whole or in any particular members?" While the clinician got this information remaining family members waited in an adjoining room with the instructions not to discuss the question among themselves. After the whole family had been queried it met again conjointly and carried out these in— structions: "Discuss among yourselves the question I have just asked each of you separately; you may discuss any aSpect of the question you wish. The only Specific request I wish to make is that at some point you talk about Specific steps you might take as a family to bring about any of your desired changes. You will have about four minutes, or more if you need it. I will not take part in your family discussion, but will remain quietly in the room." ’ ' ' “ __ w..." ~...., 3.! _. “M "-0. P "-."Jv:1-‘ nun-v, ."r’q' ~‘w .. ._ , _ . , _ _ ~ . 31 Task 2: The whole family was instructed: "Plan an activity you could all do together; it Should be something you might actually do. I will leave the room for four or five minutes; choose one person to summarize your plans for me when I return." In succession, father and the children excluding mother, mother and the children excluding father, and husband and wife excluding the children all planned activities with the experimenter present. Task 3: The parents received a proverb (First inter- view: While the cat's away the mice will play. Second interview: A rolling stone gathers no moss.). The inves— tigator asked them to discuss between themselves the meaning of the proverb and then to plan how they would teach it to their children. Upon the parent's request the interviewer retrieved the children and they were taught the proverb. Task 4: First the family arranged its seating so it ordered father, mother, and oldest to youngest child. .EEEE A; Each person secretly wrote on a blank card what he con- sidered the major fault of the person on his left. The clinician collected the cards, added two (1. "Gets mad too easily.” 2. "TOO bossy.“), and then read each of them aloud. After each statement all family members voted out- loud singularly for "the one person in family the fault best describes." Finally, each family member reSponded to the question, "What do you see as your major fault?" Part 2: Cards were distributed, marked, and voted on in the above manner; but this time in reSponse to the question, "What 32 do you like best about the person on your right?” The ex— perimenter again added two cards (1. ”Fun to talk to." 2. "Fun to be with."). After voting all family members announced what they most admired about themselves. Task 5: Each child was asked to name the parent he considered most like himself and to ennumerate one or more similarities; then, differences. Next, the children repeated comparisons and contrasts with parents least like themselves. Subsequently, each parent announced the child most and least Similar to himself and Specified Similarities and differences for both. Finally, both spouses stated felt differences and similarities between each other. Task 6: The interviewer introduced this task with, "In every family the members diSplay several feelings towards each other at various times. For the next few minutes I'm going to ask dddh_of you a couple of questions about three feelingS—-angry feelings, sad feelings, and happy feelings." Part 1. TO each child: "What are you most likely to do when you get very mad at your brother(s) and/or sister(s)? How about when you're really mad at your mother or father? Who makes you most angry, most Often, in your family? About what?” To both parents: What are you most likely to do when you get very angry with the children? Your spouse? Who makes you maddest most frequently?" Part 2. To the whole family: "What are you most likely to do when you feel sad or hurt inside? What's the saddest time you can remem- ber in your family?” Part 3. To every family member: 33 "What are you most likely to do when you're really feeling happy about something? What is the happiest time you can remember in this family?" Task 7: Both parents were queried: "Some parents expect their children to obey immediately when they tell them to be quiet or pick up something and so on; others do not think it is terribly important for a child to Obey right away. How do each of you feel about this point? Now sup- pose one Of the children doesn't react right away to a request from you to do something; what is most likely to happen? And if this is not sufficiently effective what fol- lows then?" Task 8: The whole family was asked to discuss the following situational question: "Suppose two of the children (used actual names) are very angry with each other; how should they let each other know how angry they are?" Empha- sis on including all family members in the discussion was stressed. After observing the conversation for two or three minutes the interviewer requested that the family compare their ideas to what actually happens in this situation-~if they had not already done so. Task 9: The experimenter issued identical instructions to Task 8 for the following situation: "Sometimes a child will get very angry at his parents and feel like hitting or shouting angry things at them; how should the child let his _parents know how angry he feels?" 34 All families except one abnormal family experienced the interview twice with an eight to ten week interval be- tween sessions. The whole interview took about an hour and a half to complete. All sessions were tape recorded and observed by two trained raters through a one—way window. The Rating Scale The Family Rating Scale (FRS) used derived essentially from a measuring instrument develOped by the Michigan State University Psychological Clinic during the year previous to collection of data for this study. The Michigan State Uni- versity Clinic intended the parent scale to measure variables observed in standardized family intake interviews. The items were constructed on the basis of their purported importance in the process Of family interaction, as implied in several theoretical approaches to family therapy (Ackerman, 1958; Satir, 1964; etc.). Before proceeding with this study ten pilot families (all either in treatment at the Michigan State University Clinic or on the psychotherapy waning list) participated in the standardized interview and were rated by two Observers on the original scale items. The reader will find the origi— nal scale in Appendix B. Thirty Likert Scale items yielded a mean inter-rater reliability coefficient Of .63 per item. AS a result of the pilot study (ten families) the author created the final version of the Family Rating Scale (FRS). FRS consists primarily Of items from the parent M \ [(1111 l. is! .. Ill AI 35 scale. Some parent scale items were discarded because Of irrelevancy or redundancy; several new items were added to better tailor the FRS to the behaviors elicited by the standard interview. Immediately after each interview Ob- servers independently rated the family on the FRS as it appears in Appendix C. The Raters Four raters COOperated to make this research a reality. All were graduate students in clinical psychology at Michi- gan State University; two at the first year level, two at the second year level. Only the second year students had done some previous family diagnostic work, so interviews with the ten pilot families served to train the raters. 'The research design called for three raters, employed two at a time; the fourth rater became an alternate when a regular missed a session. The experimenter assigned raters so that by the conclusion of the study each family had been Observed by three different raters (two each interview). Families were interviewed in a random order and observers did not know to which group any given family belonged. The mean correlations Of agreement for each pair of raters in inter- view 1 are as follows: .64 for rater combination 1 and 2; .50 for combination 2 and 3; and .74 for combination 1 and 3. Mean correlations for interview 2 are: .58 for rater combination 1 and 2; .57 for combination 2 and 3; and .65 for combination 1 and 3. "A m --_ CHAPTER III RESULTS Inter-Rater Reliability The Family Rating Scale (FRS) included 45 five-point Likert scales (1 2 most, and 5 - least pathologica1)' measuring variables assumed to be continuous. To begin with, inter—rater reliability coefficients (Person r's) were cal- culated for each Likert scale item on the FRS (see Table 2); the number of paired Observations for each reliability co- efficient consisted of all families in the study (16 for interview 1; 15 for interview 2). Final data analysis derived from pooled judgment scores (rater l + rater 2) on those continuous items where the inter-rater reliability coeffi— cients were found to be significantly different from zero when p : .05 (i.e. an r of .43 or greater for interview 1; .44 or greater for interview 2). Nineteen consistently rated items remained for interview 1 after this procedure; 34 for interview 2. Far more items were judged consistently for interview 2. Using the pooled judgment scores for only the consist— ently rated Likert items, over-all inter-rater reliability coefficients were then determined within each family for both groups. Table 3 presents these correlation coefficients 36 37 and the means for each group. Judges tended to rate the clinic families with higher agreement on both interviews. TABLE 2.--Inter-rater reliability coefficients for Likert scale items of the Family Rating Scale. L. Item Interview Item Interview Designation l I PIE Designation .1 2 IA .42 .64 IIIG .26* .67 IE .58 .79 IVA .25* .59 IC Father) .21* .87 IVB .19* .68 IC Mother .32* .7o IVC .25* .49 IE .50 -.36* IVE Father .26* .26* IF .5o .74 IVE Mother .03* -.23* II .17* .76 IVF Father .31* .08* IIA .50 .64 IVF Mother .14* .33* IIB .oo* .34 IVG Father .56 .83 IID .85 .84 IVG Mother .50 .79 IIH Father .62 .33* IVI Father .78' .91 IIH Mother) .46 .16* IVI Mother .40* .56 III .31* .73 IVI Oldest .72 .66 IIJ .36* .80 IVI Youngest) .52 .69 IIF Father .51 .83 IVI Middle) .69 .78 IIF Mother .61 .71 IVJ .44 .89 IIIA .64 .67 IVK .2o* .71 IIIB .2o* .69 IVL .26* .25* IIIC .56 .39* IVM .23* .57 IIID .46 .34* IVN .65 .96 IIIE .16* .77 IVO .2o* .23* IIIF .38* .79 *Not significantly different from zero correlation when p = .05. Test Of Hypothesis I Hypothesis I, which stated that ratings of family interaction Observed in the standardized interview will reveal differences between normal and clinic families, was tested in several steps. -n' m— o... ‘~ “were": I. d 38 TABLE 3.--Inter-rater reliability coeffi- cients within each family for both inter- views. mil—m m Families Interview 1* Interview 2**' __.~_ v Normal Group 1 .62 .o8*** 2 .44 .26*** 3 .45 .72 4 .l9*** .52 5 .42 .51 6 .35*** .69 7 .62 .42 8 .70 .51 7 = .47 3(- = .46 Clinic Group 9 .63 .79 10 .76 .84 ll .71 .84 12 .46 .82 13 .74 .77 14 .80 .89 15 .73 .76 16 .82 Absent I‘ = .71 'Y = .82 * n of 19 items ** n of 34 items *** r not significantly different from zero when p = .05. Summing only the consistently rated Likert scale items provided an over—all pathology score for each family; these scores and the means for each group appear in Table 4. The t test, applied to the difference between normal and clinic group mean pathology scores, yielded highly significant dif— ferences between the two groups for both interviews. The t 39 ratio was 3.34 for interview 1 (p = .005 with 14 d.f. and 19 items); 4.19 for interview 2 (p - .005 with 13 d.f. and 34 items). AS hypothesized, the normal and clinic families differed significantly on several measurable variables. TABLE 4.--Over-all pathology scores for individual families for both interviews. Families Interview 1* Interview 2** 1 I - , 5 Normal Group 1 1 105 311 b 2 136 307 3 89 ' 309 4 117 293 5 109 308 6 122 325 7 100 - 276 8 91 318 I’ a 108.62 '2 = 305.88 Clinic Group “8‘ PI 9 134 249 10 144 284 11 132 191 12 104 291 13 126 224 14 149 177 15 140 209 16 130 Absent I = 132.38 I’ = 232.14 * n of 19 items ** n Of 34 items Specification of Likert scale items which singularly discriminated between normal and abnormal groups appears in 40 'Table 5. In all cases the normal group obtained less patho— logical judgment scores than did the abnormal group. (Appen- dices E and D present the inter-correlations for consistently rated items during interviews 1 and 2 reSpectively.) For detailed inspection of any item listed in Table 5 consult Appendix C where the eXperimental version of FRS is found. TABLE 5.--T ratios between group means on consistently rated for both interviews. W FRS Item t ratios t ratios Designation FRS Item Label Interview 1 Interview 2 IA Overt Agreement ‘ 2.56** 1.40 IB Covert Agreement 1.66 1.64 IC(Father) Communication Clarity: Prevalence of Double Messages , ~- 2.02* IC(Mother) Communication Clarity: Prevalence of Double Messages -- 4.26*** IE Degree of Consensus with . Family Spokesman 1.90* 3.05***- IF(Father) Communication. Discrepency .63 .93 IF(Mother) Communication Discrepency -— .94 IF(Oldest) Communication Discrepency 1.68 1.43 IF(Youngest)Communication Discrepency —— .63 II Freedom of Speech -- 1.69 IIA Decision Making . Potential 1.68 2.57** IIB Stability of Decisions —- l.86* IID Balance of Power —.40 .74 III Degree of Insight into Alignments -- 2.21** IIJ Over—all Comfort with ' Role Relationships -- 3.45*** IIK(Father) Over-all Emotional Maturity 1.82* 5.70*** IIK(Mother) Over~all Emotional - ' Maturity .93 5-29*** 41 'TABLE 5.--Continued. FRS Item t ratios t ratios Designation FRS Item Label Interview 1 Interview 2 IIIA COOperation —- 2.34** IIIB Cohesiveness 2.00* 4.14*** IIIC Closeness 3.60*** -— IIID Tolerance of Differences l.82* —- IIIE Manifest Tension -— 2.66*** IIIF Openness of Family System -— 2.36** 1110 Participation in Session -~ .82 IVA Spontaneity of Affect -~ 1.74 IVB Amount of Anger -- 4.l8*** IVC Control Of Anger —- l.92* IVG(Father) Severity for Negative Sanctions .51 3.85*** IVG(Mother) Severity for Negative Sanctions 1.96* 6.00*** IVI Father Positive Self Regard 1.20 2.34** IVI Mother Positive Self Regard —- 1.04 IVI Oldest Positive Self Regard 1.06 2.17** IVI Youngest)Positive Self Regard .58 2.38** IVJ Warmth 1.83* 2.59** IVK Empathy -— 2.32** IVN Happiness 2.54** 2.15** Totals 19 items 34 items * = .05 ** p = .025 *** p = .01 A handful of FRS items were not Likert scales, but rather called for discrete, categorical judgments, e.g., ”From the following Six categories check the top most used techniques of behavior training for each parent." All such items except one failed to provide sensible bases for judg— ments, either because Of faulty item construction or absence of apprOpriate behaviors to rate in the standard interview J \———/I-.-=—L——-—-—_ 42 Situation. Consequently, items IIE, IIF, IIG, IVD, IVH, IVM, and IVO were discarded at the stage of data analysis. Only item IIC, Decision Making Process, was analyzed with the X2 test for independence between two samples. Judges chose one of three categories (1. Laissez—faire, 2. Authoritarian, or 3. Equalitarian) that best described the family being rated. Category consensus between judges was satisfactory (75% agreement, interview 1; 93%, inter- view 2). Categories 1 and 2 were combined to form one cell 2 in order to increase the expected frequencies for the X calculation, and the Yates correction for X2 with small samples was also employed (Walker and Lev, 1953, p. 106). NO differences between groups obtained for interview 1 (X2 = .06, p = .20); however, abnormal families were judged to be either laissez-faire or authoritarian for interview 2, while the normal families were all rated in the equali— tarian category (X2 = 11.25, p = .001). Test Of Hypothesis II Hypothesis II, which stated that ratings of family interaction Observed during interview 1 are relatively similar to ratings made during interview 2, was tested in two ways. First, correlations were calculated between interview 1 and 2 for each family with an n of 17 items, those items rated consistently during both interviews. Table 6 contains those measures of temporal similarity. InSpection of 43 rI‘able 6 reveals weak confirmation of Hypothesis II. Although all correlation coefficients are in the predicted direction only 7 of the 15 families reach the .05 level of significance i.e., an r of .40 or higher. On the other hand, the mean correlation for each group just reaches this level. NO dif- ferences between normal and abnormal family groups stand out in this family by family analysis of temporal Similarity. TABLE 6.—-Temporal stability correlations between interview 1 and 2 for each family on 17 consistently rated items. ~—v Normal Family Clinic Family Family Group Family Group 1 .29 9 .54* 2 .23 10 .48* 3 .28 11 .39 4 .65* 12 .12 5 .31 13 .14 6 .41 14 .55* 7 .74* 15 .59* 8 .57* 16 Absent 2nd Interview 'I = .41 'I = .40 * Significantly different from zero correlation when p 2 .05. Second, the pooled judgment scores on the 17 items rated consistently during both interviews, were summed to yield a Single pathology score for each family. The t ratio tested for significant differences between interview 1 and 2 group mean pathology scores. Apparently the judges consis- tently rated the normal families as slightly less pathological 44 (during the second interview (t ratio = -2.48, p = .025 with 14 d.f.); no such differences Obtained between interview 1 and 2 for the clinic family group (t ratio = —.04, p = greater than .25 with 13 d.f.). Table 7 contains t test comparisons between interview“ 1 and 2 mean scores for each item, thereby indicating which items were individually rated with least stability over time. Only five items (marked by an asterisk) singularly discrimi- nated between the two interviews for the normal family group. TABLE 7.——T test comparisons between interview 1 and 2 on each item for both experimental groups. FRS Item FRS Item Label t ratios t ratios Designation Normal Group Clinic Group IA Overt Agreement 0.00 -0.73 IB Covert Agreement -l.2l -0.02 IE Consensus with Spokesman -3.32* -0.07 IF(Father) Communication Discrepency —1.65 —0.49 IF(Oldest) Communication Discrepency -0.97 —0.99 IIA Decision Making Potential -l.l3 0.47 IID Balance Of Power -l.39 0 05 IIK(Father) Over—all Emotional Maturity —1.73 1.02 IIK(Mother) Over—all Emotional Maturity —l.56 1.29 IIIB Cohesiveness —0.29 0.43 IVG(Father) Severity for Negative Sanctions -2.45* 1.08 IVG(Mother) Severity for Negative Sanctions -2.51* 0.10 IVI Father Positive Self Regard —0.78 0.11 IVI Oldest Positive Self Regard -l.48 -0-63 IVI Youngest)Positive Self Regard —0.93 0.60 IVJ Warmth —1.84* 0.10 IVN Happiness -l.92* -0.84 .__.__— * p 2 .05 or less. CHAPTER IV DISCUSSION Inter-Rater Reliability Two findings in regard to inter-rater agreement bear discussing. First, and rather thought provoking, is the finding that judges rated the clinic families with consistently more mutual agreement than the normal families (see Table 3). One plausible explanation for this discrepency derives from noting the participation sets which each experimental group probably possessed. The clinic families, because they were on the Michigan State University Psychological Clinic treat- ment waiting list, previously experienced the Clinic's intake procedure. Typically, the intake procedure includes indi- vidual conferences with each family member, a family inter- view, and sometimes diagnostic testing with the identified patient. By nature Of that experience and their personal commitment to future psychotherapy, one could easily main— tain that the Clinic families were unintentionally primed for participation in this research project. In short, the clinic family group practiced, before the study, revelation of intra-familial difficulties as part of their request for extra-familial help. Furthermore, the clinic families may 45 46 have considered, although informed otherwise, that the two project interviews were somehow connected to their treatment. The normal families, on the other hand, knew they were being compared to a clinic group of families. Naturally, their participation set would be to Show the experimenter just how "normal" and "adjusted” they really function. Where- aS the clinic family group's set for more honest self revela- tion might increase inter-judge agreement, the normal group's set to demonstrate their "normality" might very well restrict their range of typical behaviors and thereby depress inter- judge agreement scores. Another possibility is that the normal family group is more homogeneous and therefore more restricted in the range of behaviors rated in this study. InSpection of the range Of over-all pathology scores for each family (see Table 4) reveals that the normal family group is, in fact, more re- stricted. Thus, the differential consistency with which the judges rated the two experimental groups is, no doubt, related tO the two groups inherent difference in variability. Second iS the finding that judges rated far more items consistently during interview 2 than during interview 1: 35 compared to 19 respectively. Although he has no way Of directly proving the following assertion, the author main- tains that this difference results from the raters increased skill and experience with the measuring instrument. Essen- tially, this is a desirable practice effect. Hypothesis I Experimental Hypothesis I was confirmed; that is, normal and clinic families differed significantly on several inter- action variables. The reader will remember that the judges rated family interaction immediately after the standard interview, and that they therefore made their judgments on the basis Of global impressions of the families' behavior. Consequently, one cannot purport that judgment Scores reflect 100% the content of every FRS item. Examination Of Appen- dix D suggests that the notorious "halo effect" exerted con- siderable influence on the interview 1 results. The mean intercorrelation between only discriminating items was .52 as compared to .29 for the non-discriminating items. This pattern, although not as clearcut for interview 2 seems again to be Operating (see Appendix E). The mean intercorrelation between discriminating items was .53 as compared to .39 for the non-discriminating items. In short, the raters clearly discriminated between the experimental groups in a sort Of gross, over-all fashion; but which of the Significant dif- ferences between the two groups on specific items are real, a result of the halo effect, or both is not at once dis— cernible. With this in mind the author will Specify the variables (FRS items) which singularly discriminated between the normal and clinic samples. Study of Table 5 reveals that the judges' pooled ratings for interview 1 differentiated normal and clinic families on ten separate items. All interview behavior 48 considered, normal families agreed with each other more Often than clinic families (item IA), as well as exhibiting more consensus with individual members who from time to time acted as family Spokesman (item IF). At the same time, normal families diSplayed more tolerance for intra—family diversity of Opinions and actions (item IIID). Normal families were judged as more openly expressive of warmth between members (item IVJ), as more comfortable in each other's company (item IIIC), and as possessing more Spirit Of unity, or identifi— cation with the family unit (item IIIB). The mothers, but not the fathers, in the normal group were assessed to be less severe in administering negative sanctions in child rearing. Both parents, however, demonstrated significantly more over-all satisfaction with their various family roles i.e., Spouse, parent, provider, individual, etc. (item IIK). Finally, normal families appeared to be happier than clinic families (item IVN). Only one item (IA: Overt Agreement) which differen- tiated between the groups in interview 1 failed to do so in interview 2; two other discriminating items (IIIB: Closeness, IIID: Tolerance Of Differences) from interview 1 eluded analysis in interview 2 because of insufficient inter-rater reliability. Eleven additional items, not rated consistently enough to merit analysis for interview 1, emerged as adequate discriminators during interview 2. The eleven discriminating items peculiar to interview 2 complement empirical confirmation Of Hypothesis I. Normal 49 and clinic families differed markedly in the manner that decisions were reached; the Observers assessed the normals' decision-making process as equalitarian and shared, the clinics' as either authoritarian or laissez-faire (item IID). Furthermore, the normals Obtained higher judgment scores on potentiality for reaching rational family decisions (item IIIA), as well as the capacity to stick with decisions once they were finalized (item IIIB). Normal families displayed less manifest tension (item IIIE); more interpersonal empathy (item IVK); and less overt anger, either towards each other or extra—familial Objects (items IVB). Anger also tended to be more controlled, even over—controlled, in the normal family group (item IVC). Judges perceived both parents in the normal group as communicating fewer double messages to each other and to their children than in the clinic group (item 10), and as less severe in administering negative sanctions as child rearing practices (item IVC). Normal families exhibited more insight about inter—member align- ments than did clinic families (item III). Finally, normal family members were noted to feel more positively towards themselves, to possess more positive self regard than the clinic families (item IVI). Some of the above differences between normal and clinic families are consonant with previous findings in the family interaction literature. Two earlier studies (Fisher §t_dl., 1959 and Ferreira, 1965) found as did this research that normal families exhibit more inter-member agreement than do SO clinic families. Fisher §E_2l- (1959) studied conjoint family TAT stories; Ferreira (1965), the family decision- making process. The finding that parents Of normal families are both judged emotionally healthier than parents of clinic families was also reported by Winter E£_El- (1965); in con- trast to the present study Winter rated parents' joint TAT stories. Differences between normal and clinic families that were found on the variables measured by the remaining items represent this project's addition to existing family inter- action literature. However, the author finds it somewhat difficult to explicate these welcomed results, primarily because the item content is not always associated with a Specific behavior incidence occurring during the interview. This is a major shortcoming of the global rating procedure employed herein. It is Significant, indeed, to discover that two samples SO small as those in this study can be reliably differentiated with this technique, but family inter- action research must proceed beyond this point if more SOphisticated classification systems Of family types are to evolve. The distinct contribution Of this study is that its results suggest fruitful variables for further, detailed research. Hypothesis II Two findings concerning the hypothesis of predicted temporal similarity of interaction will be discussed. 51 First, the correlations between interviews 1 and 2 for each family on the 17 consistently rated items ranged from .12 to .74 with an over-all mean correlation Of .41. This finding suggests that there exists a core of interaction consistency over time for both experimental groups which exhibits itself in various degrees for each individual fam- ily. Although the measures Of temporal similarity are not as high as pre-study expectations,they stand almost alone in contrast to the majority Of family interaction studies lacking such measures of reliability over time. Such a finding lends welcome support to the assumption underlying family research that typical interactions are relatively permanent structures. Another reason for attempting to assess the temporal stability of an interaction rating scale is to ascertain its potential for measuring change which occurs as a product Of the psychotherapeutic relationship. The level Of temporal stability yielded by the FRS as it now exists will need to be improved somewhat to render an adequate instru— ment for gauging psychotherapeutic change. Such improvement will probably take the shape of better Operationalizing the variable represented by the items that together distinguish between the normal and clinic groups. At this stage of the family research enterprise, how— ever, expecting higher temporal stability than found in the project is, perhaps, a bit unrealistic. Heightened re- liability in assessing family interaction over time will 52 not be forthcoming until measuring instruments reflect changes in conceptual schemes which include genotypes as well as phenotypes Of family interaction. Instruments such as the rating scale herein primarily reflect temporal changes in phenotypic rather than genotypic interactions. Conse— quently, low temporal reliability scores inaccurately reveal genotypic stability which is assumed to be more unchanging over time. To punctuate these assertions, visualize a family characterized by a particular type of inappropriate sexual seductiveness; that is, a family marked by repetitive, covert sexual interaction between the parents and children. In light of the above argument, assessing this family with FRS on two different occasions could readily yield valid ratings so far as each item is concerned, but low scores on the similarity measures. How might this occur? Observers might witness the family's seductiveness during interview 1 via a heated argument between the father and teenage daughter while the mother seems bored and lost in a day—dream. In- terview 2 might reveal a different example Of seductiveness altogether, an incident featuring both parents, but es- pecially the mother, depreciating their pre-adolescent son's young and attractive female teacher with whom he is persis- tently negativistic. Taking these different instances of inapprOpriate seductiveness, Observers might very likely judge this family higher on Amount of Overt Anger (item IVE) than on amount of Warmth and Empathy (items IVJ and IVK 53 reSpectively) after interview 1, but vice versa after inter- view 2. In short the ratings accurately reflect the pheno- typic interaction fluctuation (anger, empathy, and warmth), but reveal little about the genotypic stability (degree Of inapprOpriate sexual seductiveness). It appears to this author that a salient goal for fur— ther research is the delineation of central genotypic inter- actions in the complex process of family living. This endeavor will most likely evolve from a wedding of relevant interpersonal theory and the discovery of inter—relationships between empirically known phenotypical interactions. Earlier in this paper such concepts as Bach's (1967) ”marital fighting styles,” Berne's (1964) "transactional games," Kell and Meuller's (1966) "eliciting behaviors," Bateson's (1962) "double bind," and Wynne's "pseudomutuality," were suggested as relevant theoretical sources. Delineation of genotypes in family interaction may be expedited by employ- ment of a non-linear model such as Crego's (1966) pattern analytic approach. Second, t tests for significant differences between interview 1 and 2 mean group pathology scores yielded an interesting finding. The normal family group was rated sig— nificantly less pathological after the second interview; no such differences obtained for the clinic group. Interviews 1 and 2 were essentially identical in format and families obviously anticipated this fact early in the second session. 54 'Why then did the normal families change in amount Of judged pathology, but not the clinic ones? One plausible eXplana- tion may represent an important difference between the two experimental groups. The normal families apparently profit from practice and perfect smoother ways of carrying out the experimental tasks when encountered a second time. Such flexibility is less characteristic of families like the clinic sample seeking psychiatric help. The items which singularly reflect this flexibility most are presented in Table 7 (i.e., items marked by an asterisk). During interview 2, as compared to interview 1, the normal families were judged as a group to be signif- icangly different from itself on five items, in the direction of exhibiting less pathology. During interview 2 normal families displayed more consensus with individual members who from time to time acted as family spokesman, less severity for negative sanctions on the part Of both parents, more warmth between members, and more happiness. CHAPTER V SUMMARY This study was undertaken to eXplore two questions of family interaction: (1) Can we distinguish between "normal” and "abnormal" families in terms of interaction processes? and (2) How stable are family interaction processes over time? Eight normal and seven abnormal fam— ilies (four or five members each) participated in two similar semi-structured interviews with an 8-10 week interval between sessions; two trained observers rated each family on an in— strument consisting of 50 items, mostly five-point Likert scales. Normal families had no known history of psychiatric disorder or treatment and were obtained through labor union and church groups. Abnormal families were all drawn from the treatment waiting list of the Michigan State University Psychological Clinic. It was hypothesized that (1) judges' ratings of family interaction Observed in the standard interview would reveal differences between the normal and clinic families, and (2) that these interaction ratings would be relatively Similar in the first and second sessions. Nineteen items from the Family Rating Scale yielded inter-rater reliability coefficients that were adequate; 55 56 34 from interview 2. The criterion of adequacy was that an inter-rater reliability coefficient be significantly different from zero at the .05 level (i.e., an r of .43 or greater for interview 1; .44 or greater for interview 2). Statistical analysis to test Hypothesis I issued from the reSponseS to these items. Hypothesis II was tested by analysis of judges' responses to the 17 items with adequate interwrater reliability for both interviews. Regarding the first hypothesis normal families were differentiated from the clinic families by over-all pathology scores and a number of individual Family Rating Scale items. A picture of the normal family in this study emerged from the data. In comparison to the clinic families the normals are characterized by more interwmember agreement, more capac- ity for reaching common decisions in a equalitarian fashion, less over-all anger but more tolerance of individual independ- ence in thought and action, more interpersonal warmth and co- hesiveness, less manifest tension, and a greater degree of happiness. A most significant complement to this description is that both parents in the normal families displayed more} over-all satisfaction and effectiveness within their various family roles. The second hypothesis was tested in two ways. First, the correlations between interview 1 and 2 for each family ranges from .12 to .74 with an over-all mean correlation of .41. Although these measures of temporal stability are 57 not as high as pre—study expectations, they suggest that there exists a core of interaction consistency over time for both experimental groups. Such finding lends support to the premise, frequently assumed in family research projects but seldom tested, that typical interactions are relatively permanent structures. Second, t tests between interview 1 and 2 53222 mean pathology scores revealed that normal families were rated as Significantly less pathological after the second inter- view; no such differences Obtained for clinic families. This interesting finding may reveal a basic difference be- tween normal families and those which seek some form Of psychiatric help, a differential capacity to profit from practice and perfect smoother ways of carrying out the ex- perimental tasks when encountered a second time. Finally, evaluation of this study's findings led the author to suggest directions for further family interactional research. REFERENCES 58 REFERENCES Arnold, M. 1962. Story Sequence Analysis. New York: Columbia University Press. Asch, s. E. 1951. "Effects of Group Pressure Upon the Modification and Distortion of Judgments,” Groups, Leadership and Men. Ed. H. Guetzkow. Pittsburgh: Carnegie Press, pp. 177-190. Bach, 0.6 §967. Intimate Enemies. (To be published January, 19 7. Bach, G., Bacon, D. L., Searles, H. F., Warkentin, J., and Leland, T. 1966. "Conference by Telephone Discussing Hate and Aggression in Psychotherapy,’ Voices, 1, 18-31. Bachove. C., and Zubaly, B. M. 1965. "Exploratory Study of Family Inter-relationships in a Small Sample of Non-clinic Families,” Masters thesis, Bryn Mawr, described in Chapter 11, Intensive Family Therapy. Ed. Boszormenyi- Nazy and Framo. New York: Hoeber Medical Division, Harper and Row. Baldwin, A. L. 1949. "Effects of Home Environment on Nur— gury School Behavior," Child Development, 20, 49- 2. Bales, R. F. 1952. "Some Uniformities Of Behavior in Small Social Systems," Readings in Social Psychology. Ed. G. E. Swanson. T. M. Newcomb, and E. L. Hartley, New York: Holt. jBatescwn C», Jackson, D. D., Haley, J., and Weakland, J. H. 1956. "Towards a Theory of Schizophrenia,” Behavioral Science, 1, 251-264. . 1962. "A Note on the Double Bind," Family Process, 2, 154-164. Bavelias, A. 1950. ”Communication Patterns in Task-Oriented C}roups," J. Acoust. Soc. Amer., 22, 725-730. -B€rnea E. 1964. Games Pepple Play. New York: Grove Press, IEnc. 59 ”.75 ‘3‘. 1 60 Bishop, B. M. 1951. "Mother-Child Interaction and the Social Behavior Of Children," Psychol. Monogr., NO. 11, 65, 1-43. Borgatta, E. F., Cotrell, L. S., and Mayer, H. J. 1956. "On the Dimensions Of Group Behavior," Sociomet., 19, 223—240. Caputo, D. V. 1963. "Parents Of SchiZOphrenics,” Family Process, 2, 339—356. Cartwright, D., and Zander, A. 1962. Group Dynamics: Research and Theory. (2nd edition). Evanston, 111.: Row, Peterson. Champney, H. 1941. "The Measurement of Parent Behavior,” Child Develop., 12, 131-166. Crutchfield, R. S. 1955. "Conformity and Character," Amer. Psychol., 10, 191-198. Drechsler, R. J., and Shapiro, M. I. 1961. "A procedure for Direct Observation Of Family Interaction in a Child Guidance Clinic," Psychiatry, 24, 163-170. . 1963. "Two Methods of Analysis of Family Diagnostic Data,” Family Process, 2, 367-379. Epstein, N. B., and Westley, W. A. 1959. ”Patterns of Intra-Familial Communication," Psychiatric Research Reports. Ed. D. E. Cameron and M. Greenblatt, Vol. 11, pp. 1—12. Farina, A.' 1960. ”Patterns of Role Dominance and Conflict in Parents Of SchiZOphrenics,” J. Abnorm. Soc. Psycholg 61, 31—38. Farina, A., and Dunham, R. N. 1963. "Measurement Of Family Relationships and Their Effects," Arch. Gen. Psychiat., 9) 64‘73' .Ferreira, A. J. 1963. "Decision—making in Normal and Pathological Families," Arch. Gen. Psychiat, 8, 68- 73. 1963. ”Rejection and Expectancy of Rejection .______‘~_. .in Families," Family Process, 2, 235-244. Ferreigra, A. J., and Winter, W. D. 1965. ”Family Interaction Eind Decision Making," Arch. Gen. Psychiat., 13, 214— 2223. 61 Ferreira, A. J., Winter, W. D., and Poindexter, E. J. 1966. "Some Interactional Variables in Normal and Abnormal Families," Family Process, 5, 60-75. Festinger, L., Schachter, S., and Back, K. W. 1950. Social Pressures in Informal Groups. New York: Harper. Festinger, L., Pepitone, A., and Newcomb, T. M. 1952. ”Some Consequences Of De—individuation in a Group,” J. Abnormal Soc. Psychol., 46, 382-389. Fisher, 8., Boyd, I., Walker, D., and Sheer, D. 1959. ”Parents Of Schizophrenics, Neurotics, and Normals,” Arch. Gen. Psychiat., 1, 149-166. Fouriezos, N. T., Hutt, M. L., and Guetzkow, H. 1950. ”Measurement of Self-Oriented Needs in Discussion Groups,” J. Abnorm. Soc. Psychol., 45, 682-690. Framo, J. L. 1965. "Systematic Research on Family Dy— namics,” Intensive Family Therapy. Ed. Boszormenyi- Nagy, I.,and Framo, J. L. New York: Hoeber Medical Division, Harper and Row. Chapter 11. French, J. R. P. 1941. "The Disruption and Cohesion of Groups,” J. Abnorm. Soc. Psychol., 36, 361—377. Garmezy, N., Farnia, A., and Rodnick, E. H. 1960. "The Structured Situation Test: A Method for Studying Family Interaction in Schizophrenia,” Amer. J. Orthopsychiat., 30, 445—452. (3rOSS, E. 1956. "Symbiosis and Consensus asn Integrative Factors in Small Groups,” Amer. SociOl. Rev., 21, 174-179. .Haley, J. 1962. "Family Experiments: A New Type Of Ex— perimentation," Family Process, 1, 265-293. 1964. ”Research on Family Patterns; An Instru— ment Measurement," Family Process, 3, 41—65. Halwa, A. P. 1962. Handbook of Small Grddp Research. Glencoe, 111.: Free Press. .Kell, LB., and Mueller, W. 1966. Impact and Change: A fitudy of Counseling Relationships. New York: Ikppleton-Century—Crofts, 1966. Kelleifl, H; H. 1951. "Communication in Experimentally (3reated Hierarchies," Hum. Rel., 4, 39-56. 62 Kenkel, W. F., and Hoffman, D.K. 1956. "Real and Con- ceived Roles in Family Decision Making,” Marr. & Fam. Liv., 19, 323-343. Leary, T. 1957. Interpersonal Diagnosis of Personality. New York: Ronald Press. Leavitt, H. J. 1951. "Some Effects of Certain Communica— tion Patterns on Group Performance,” J. Abnorm: Soc. Psychol., 46, 38—50. Lennard, H. L., Beaulieu, M. R., and Embrey, N. G. 1965. "Interaction in Families with a SchiZOphrenic Child," Arch. Gen. Psychiat., 12, 166-183.. Levinger, G. 1965. "The Assessment of Family Relationships: A Progress Report," Western Reserve University, Cleveland, Ohio, (mimeo), described in Chapter 11, Intensive Family Therapy. Ed. Boszormenyi-Nagy and Framo. New York: Hoeber Medical Division, Harper and Row. March, J. G. 1953. "Husband-Wife Interaction over Political Issues," Pdpl. Opin. Quar., 17, 461-470. Mills, T. M. 1953. "Power Relations in Three-Person Groups,” Amer. Sociol. Rev., 18, 351—357. - 1954. "Power Relations in Three-Person Groups," Amer. Sociol. Rev., 19, 657-667, ldoustakas, C. E., Sigel, I. E., and Schalock, H. D. 1956. "An Objective Method for the Measurement and Analysis of Child-Adult Interaction," Child Develqp., 27, 265—279. (Dsgood, C. E., Suci, G. L., and Tannebaum, P. H. 1957. The Measurement Of Meaning. Urbana: University Of Illinois Press. Paiwmons, T. 1955. "Family Structure and the Socialization Of the Child," Family, Socialization, and Interaction Process. Ed. T. Parsons and R. F. Bales. GlenCOe, Ill.: Free Press, pp. 35-131. rRiSKiJl, J. 1963. "Methodology for Studying Family Inter- action,” Arch. Gen. Psychidt., 8, 343-348. . 1964. "Family Interaction Scales: A Preliminary IReport,” Arch. Gen. Psychiat., 11, 484-494. l 1 ill ' ll .Illl. Ill Al 63 Rodnick, E. H. and Garmezy, N. 1957. ”An Experimental Approach to the Study of Motivation in SchiZOphrenia,” Nebraska Symposium on Motivation. Ed. M. R. Jones. Lincoln: University Press, pp. 109-184. Ruesch, J., Block, J., and Bennett, L. 1953. ”The Assess— ment Of Communication: A Method on the Analysis of Communications," J. Psychol., 35, 59-80. Schmideberg, M. 1948. "Parents as Children,” Psychiat. Quart., 22, 207-218. Strodtbeck, F. L. 1951. "Husband-Wife Interaction Over Revealed Differences,” Amer. Sociol. Rev., 16, 468- 473. - 1954. ”The Family as a Three-Person Group," Amer. Sociol. Rev., 19, 23-29. Terrill, J. M., and Terrill, R. E. 1965. ”A Method for Studying Family Communication,” Family Process, 4, 259—290. Walker, H. M., and Lev, J. 1953. Statistical Inference. New York: Holt, Rinehart, and Winston, Inc. Weakland, J. 1962. "Family Therapy as a Research Arena,” Family Process, 1, 63—69. ; W. D., Ferreira, A. J., and Olson, J. L. 1965. 'Story Sequence Analysis of Family TAT'S," J. Proj. Techn. and Person. Assess., 29, 392-397. Winter ‘Wynne, L. C., Ryckoff, I. M., Day, J., and Hirsch, S. I. 1958. "Pseudo-Mutuality in the Family Relations Of Schizophrenics," Psychiat., 21, 205—220. 'Wynrmg I“ C. 1961. "The Study of Intrafamilial Alignments and Splits in Exploratory Family Therapy,” Exploridg the Base for Family Therapy. Ed. N. W. Ackerman, F. L. Beatman, and S. N. Sherman. New York: Fam. Serv. Asc. of America, pp. 95—115. APPENDICES 64 APPENDIX A SPECIFICATION OF FAMILIES SAMPLED IN THIS STUDY 66 Specification of families sampled in this study; families 1-8 are the normal group and families 9-16 are the abnormal group. Family Father MOther Children Number . . . . ‘77 Occupation Education Occupatlon Educatlon Sex Age 1 Tool and Die 12 Housewife 12 M 11 Maker M 14 2 IBM Programmer 14 Housewife 14 M 9 M 5 3 Personnel 12 Housewife 12 M 7 Director, Mich. M 7 Dept. of Social F 12 Service 4 Labor Union 12 Housewife 12 M 8 Leader M 9 F 11 5 Bricklayer l2 Housewife 12 M 15 F 10 6 Accountant 16 Housewife 12 M 17 M 12 7 Postal Clerk 12 Secretary 12 M 12 F 16 8 Insurance 12 Housewife 12 F 7 Salesman M 9 9 Machinist 12 Housewife 12 M 7 F 9 M 12 10 Graduate 17 Housewife 13 M 10 Student F 13 11 Pet Store 15 Housewife 12 M 8 Manager M 12 £12 Mechanical 17 Housewife 14 M 8 Engineer M 12 lg3 Cartographer l3 Housewife 11 M 7 F 12 M 14 14- Graduate 16 Secretary 12 M 8 Student M 12 M 15 67 Specification of families sampled in this study, continued. Family Father Mother Children Number Occupation Education Occupation Education Sex Age l5 Factory 12 Housewife 12 M 7 Worker F 8 M 9 16 Construction 10 Secretary 11 M 8 Worker M 11 M 16 APPENDIX B RATING SCALE LSED IN THE PILOT STUDY 68 ' mm: soars as Revision) .. mums... :2: ~ '1 ; :3: ~‘ -' A. Overt: Agreement: To what degree {Idenfamlly members agree with each other? 1. No agreement: with what: is said; {aon'stantly arguing. 3. Seem to agree and disagree about edual amounts with what: others say. 5. Members seem to articulate with each other so well that listeners in the family agree virtually all the time with the speaker". B. Covert _A_greement:: To what: degree do family members agree on a non-verbal, more hidden level (nods, facial expressions, sounds)? 1. Practically no agreement with what is said; constant evidences .of difference. .3. Seem to agree and disagree about: equal amounts with what others say. 5. Other members of the family seem to .agree with each other most: of the time. 0. Consistency: To what: degree does the family maintain a stated position or view? Infra-family - To what degree do they shift or not shift in response to pressures from one another? ' 1.Vi~:tua11y no consistency of comunication; constant change of mind, contradictions within statements, shifts in position. 3. Moderate and equal amounts of consistency and inconsistency. 4. Consistent communication; mqnbers maintain positions and change them only after considerable persuasion by others. 5. Rigid - party-line conformity. D. Home the. most: vocal member of the family: B. Family Spokesman: The degree. to which a particular member does the talking. I. No one speaks for anyone else; a collection of individuals speaking only for themselves. 3. One spokesman who dominates communication and speaks for everyone almost: all the time. 5. Shared spokesmen; one or two members" may eXpress family positions, but: the spokesman varies according to the tapic -' members feel free to ' speak for all when they feel they can do so. P. berg ree Spokesman Represents Family Consensus: Assuming that: there is a person, or persons, who seems to be speaking for the entire family unit, to what degree does the speaker really seem to represent the family opinion? 1. Not at all; although the speaker seems to be speaking for everyone, he or she is clearly representing only his or her own opinion and . no one else seems to agree. V (o'ver) Rating Scale '2' (24 rave)- - . ‘ . . A 2 .j ‘ Rating 3. The spokesman seema'to be speaking for hbmself but the rest of the I family agrees with the spokesman to a.moderate degree. . 5. The spokesman seems to truly be expressing for all the family, which is in agreement with what is being said. F G. Freedom of Speech: How free and comfortable do members feel about speak- ing and taking positions? 1. Members feel reluctant to speak meaningfully to each other or for themselves. 3. Some inhibitions about talking to each other but can communicate at some times about some things. 5. Virtually all members feel free and comfortable to speak when they want to. G H. Rank order each family member on frequency and amount of actual speaking, ram who spoke the most to who spoke the least in total time. (#1 is most talkative.) II. ROLE RELATIONSHIPS A. gglance of Power: Is there a boss? 1. No one seems able to take over leadership in the family, even though attempts may be made by one or more members to do so. 3. While there is no consistent family leader, at different times various members take over the "boss" position, but cannot hold it for long. 5. There is definitely a family leader who has control of the family §§E_§ whenever he or she decides to exert his power. B. Who is the Boss - in the clinician's judgment? E Rank order each family member according to their leadership position in the family - who is the boss, the most powerful, etc., and who is next in line? (#1 is the strongest leader.) C. Who does the family perceive as the boss? D. Decision-Making: Can the family arrive at a common decision? (Rate if decision attempted during hour.) 1. The family is so paralyzed with confusion and conflict that it seems incapable of reaching any decision. 1 3. Arriving at a decision is difficult, but the family does so with only moderate conflict and decisions are often rational ones. 5. The family seems capable of working tagether smoothly in reaching a 'rational decision. E. Stability of Decisions: Once reached, does the family sustain its = decisiofia? (Infer frm available information.) J 1. Decisions made are extremely unstable and are constantly being altered, reversed, or criticized. ‘ '5 . . Rating- Scale -3- . (2d rev.) G. 3. Decisions made are heldto about as often as they are changed. 5. Once a decision is made, it "is held' to firmly and consistently. Decision-Maki Process: How 'are decisionsreached? (Use scale if decision is made.) 1. The process is unstable, various members being involved, and decisions often seem to emerge out of general confusion. 3. The process is generally a one-sided, authoritarian one. 5. Decisions are reached through an equalitarian, shared process in which all involved members play a role in the solution. Alignments: Degree of Insight into Alignments: How aware, at all levels, is the family of the alliances and subgroupings which go on within the total unit? I. Family seems to be completely unaware of alignments. 2. Family denies any knowledge of alignments, but covertly seems to be aware of their existence. 3. Family seems to be aware of some of the alignments but unaware of others. 4. Family aware of most of the alignments but unaware of at least one alignment. 5. Family seems to be completely aware of all alignments. Use B if can obtain information. Must ask specifically about this to rates J. ' alliances'whenever possible. Starting with a sketch of the seatg Ali nments° Stabilit : How stable and long standing are the alignments within this family, o¥ whatever kind? ' 1. There appears to be little permanence or stability to any of the alignments made; they are constantly shifting and in flux. 3. There is a moderate degree of stability to at least some of the alignments in the family; while they do shift, once made they are likely to exist for some time. A 5. The present alignments in the family, or general lack of them, have existed this way for a very long time and seem quite firm and permanent. Rate the amount of ambivalence in the family about the status of family aligmnents. (Infer from available information.) 1. None. 2. Little. 3. Moderately ambivalent. 4. Quite. 5. Extremely conflicted. 'Aliggnt Chart: . 'Attempt to sketch out a sociogramatic diagram of the main fami y a ignments, including subgroupings and movements between over) arr_e_ngements may help. Rating Scale - ; . 24- (2d rev.) K. Sexual Roles: How apprOpriate are the overt roles assumed by the male and female adults in relation to the cultural stereotype?' l. 2. Generally reversed. w ,....-. :‘B -" 154132" u Unclear, unstable, constantly changing. 3. ApprOpriate. D L. Consistency of Sexual Roles: 1. Completely erratic; always shifting; constantly struggling to maintain them. 3. Moderately stable; some shifting, but not in all aspects or for very long. . . 5. Quite consistently firm and stable. M. gyerall Comfort with Role Relationships in Family: How comfortable does ‘ E. this family feel with their present role relationships? 1. Not at all happy; no one seems to be satisfied with his or her role. 2. Generally unhappy, but at least one member seems reasonably comfortable and satisfied with the way things are. 3. Generally comfortable, but at least one member reasonably dissatisfied with the way things are. - F 5. All quite satisfied with present role relationships. III. FAMILY COHESIVENESS. A. Cooperation: Can this family work cOOperatively on a task or problem? 1. 3. 5. Little or no c00peration within the family; considerable squabbling or arguing; everyone goes his own way. Moderate degree of c00peration, up to a point, but members insist on having their way, too. Considerable degree of c00perativeness, of working together and submerging individual goals and feelings for the common goal. B. Cohesiveness: Identification with the family unit; a feeling of "we" rather than "I." l. 3. 5. None; just a collection of seemingly unrelated individuals. Mbdest degree of identification with the family unit. Pleasure and pride in family unit and real feeling for it. C. Closeness: Is this family a close unit; do they like to do things together? '1. No closeness; members are quite isolated, separated, independent of each other. I Wm Ratiflg Scale -5-_ _ (2d rev.) \& m: a; 3. Moderate degree of closeness, especially if something interests a member. " 5. Quite close, tend to do things frequently as a group; seem to like and feel comfortable in each other' s company. D. Tolerance of Differences - degree of real acceptance and understanding regardless of what they say: 1. No tolerance for differences of Opinions or actions - everything has to be done in one way only. 3. Moderate tolerance for differences in certain areas and up tO‘a certain point. 5. Considerable flexibility and tolerance for a diversity of opinions and behavior by members of the family. 5 E. Tension: Is the family a tense or a relaxed unit? 1. Extremely tense family; no relaxation at all. 3. Moderate amount of tension, but can relax and be casual virtually to same degree. 5. Relaxed and comfortable. F. Openness: Is the family unit a very tight and closed system, or is it receptive and Open to others. 1. Family tight, closed, self-contained, suspicious and rejecting of others attempting to understand or entree the system. 3. Family reserved, but also able to Open up for another person, especially after they get to know the person. i“ 5. Family very Open, accepting of others, bids them welcome, likes others to join in with them, friendly. I” G. Participation.1n.Family Session: Does the family participate in the family session and to what degree? 1. Very reluctant to get involved in the interview; no real'willingness to participate. " - 2. Entire family reluctant to get involved, but gradually at least one member begins to participate. 3. Mbst of family reluctant to participate, but at least onejperson does (may act as family spokesman). 4. MOst of the family willing to participate, but at least one person reluctant to do so, prefers to stay uninvolved. tfié 5. Family very willing to participate and be involved in family session. =oi~ ' (over) Rating Scale -6- (2d rev.) IV. FAMILY Pastimes A. Spontaneity: How free is the family in eXpressing affect? 1. Greatly inhibited, over-controlled, or very impulsive; members express little feeling and may be quite discomfited if this happens; reserved, deliberate. 3. Moderate degree of inhibition, but in controlled manner can express affect and occasionally can become spontaneous. 5. Quite spontaneous and free with feelings. B. Agggg: How angry is this family as a whole? 1. Intensely angry; furious; in a rage. 3. Moderately angry. 5. Little or no anger. G. Control of Anger: No matter how angry the family is, how well is this controlled? 1. Poorly controlled, family has little or no control over its expression. 3. Moderately controlled. 5. Well controlled, perhaps even to degree of over-control. D. Direction of Ange : At whom is the family angry? (Indicate with whom the family is most angry.) I. At everyone; everyone is mad at everyone else. 2. Usually at one person at a time, but the target is constantly shifting. 3. At one person, usually a child. 4. Parents angry at siblings and/or vice versa. 5. Siblings angry at each other. 6. Parents angry at each other. 7. At persons outside the family group - neighbors, boss, teachers, etc. 8. No one=really; anger, when expressed, seems appropriate in terms of degree and Object. E. Whrmth: Is this a warm family? 1. No warmth, members very cold and distant from each other. 2. Slight warmth, but members generally reserved and cool. 3. Moderate degree of warmth, but somewhat inconsistent. Ratit l \M . a. r ."5 'Ratixug Scale -7- (2d rev.) 4. Generally warm, but somewhat reserved in showing this. 5. Quite warm and open in expressing this to each other. . hpathy: . Is this a family which can feel for others, including members 0 the unit? 1. Little feeling for anyone but oneself. 2. Some feeling for others, but the self is still paramount. 3. Equal feeling for self and others. 5. Considerable feeling and caring for other people, both in- and outside of the family unit. ' Nurturangg: ‘Is this a giving family in which dependent behavior is accepted? ' 1. Very withholding, ungiving, discouraging of dependency, no time for others. 2. Somewhat giving and nurturant, but mainly around specific things (e.g., an accident) or time of life (e.g., to an infant). 3. Nurturant, give comfortably to each other, not only in response to needs of each but in response to the nurturer's needs as well. 4. Somewhat overnurturant - overprotective. 5. Extreme overnurturance, to point of symbiosis. - Concern about Sexuality; Is the family concerned about sexual feelings or behavior? (Infer from available information.) -1. Little or no concern, almost a-denial of problem or area. 3. Moderate and/or appropriate concern. 5. Overconcern which seems an inappropriate magnifying of this area. Comfort with Sexuality: How comfortable is this family in discussing sexual concerns or behavior? (Rate if discussed during interview;) 1. Most uncomfortable, anxious, unwilling to discuss this area, guilty. 3.‘MOderate1y comfortable, but with some anxiety and some reticence at discussing this area. 5. Quite comfortable and relaxed in discussing this area; Open.among themmelves about this area. Independence: Does this family encourage independent behavior by its members. 1. Very punishing and discouraging of independent behavior, fearful of it, rewarding dependent behavior. (over) Rating Scale -8- (2d rev.) ' 3. Ambivalent about this area, encouraging and discouraging both dependent and independent behavior. 4. Quite encouraging of it, but in an age-apprOpriate and realistic manner. 5. Push family members to be independent at all costs. K. Happiness: Is this a happy family? 1. No - very unhappy, depressed, sad. 3. Moderately so, but mainly about specific events or happenings. Do not seem as happy as they should be. 5. Happy with_each other and themselves - pleased with the way things are going. L. Likeability: Do you like this family? 1. No, very unpleasant. 2. Generally no, only occasionally so. 3. Ambivalent, sometimes likeable, about as often not. 4. Generally yes, only occasionally no. 5. Yes, this is a most pleasant family. V. IMPRESSIONS Please describe anything else about this family which seems important but which does not seem to be covered adequately in the existing rating scales. This will not only aid in the family description, but will help in the generation of better or additional ways of conceptualizing family interaction. As revised Oct. 1965 MA:mc m I ‘81.. FAMILY RATING SCALES_£3rd,Revision) I. IQOMMUNICATION A. C. D. E. F. Overt Agreement: To what degree do family members agree with each other? 1. No agreement with what is said; constantly arguing. 3. Seem to agree and disagree about equal amounts with what others say. 5. Members seem to agree with each other practically all the time. Covert Agreement: To what degree do family members agree on a non-verbal, more hidden level (nods, facial expressions, sounds)? l. Practically no agreement with what is said; constant evidences of difference. 3. Seem to agree and disagree about equal amounts with what others say. 5. Other members of the family seem to agree with each other most of the time. Clarity: How prevalent are double messages in the interactions between parents and between parents and children? Rate each parent separately. 1. High percentage of double messages; parent says one thing overtly while communicating simultaneously a contrasting covert (nonaverbal) message most of the time, e.g. Parent telling child: "It's o.k. for you to say what you don't like about me, but be honest about it." 3. Moderate prevalence of double messages; such communication is present about as often as not. 5. Low percentage of double messages; almost no evidence of double bind communication; most communications lack contrasting overt and covert meanings. Name the most vocal member of the family: Family Spokesman: The degree to which a particular member does the talking. 1. No one Speaks for anyone else; a collection of individuals speaking only for themselves. 3. One spokesman who dominates communication and speaks for everyone almost all the time. 5. Shared Spokesman; one or two members may express family positions, but the spokesman varies according to the topic - members feel free to speak for all when they feel they can do so. Degree Spokesman Represents Family Consensus: Assuming that there is a person, or persons, who seems to be speaking for the entire family unit, to what degree does the speaker really seem to represent the family opinion? 1. Not at all; although the speaker seemsto be speaking for everyone, he or she is clearly representing only his or her own opinion and no one else seems to agree. (over) Rating Scale -2- (3d rev.) 3. The spokesman seems to be speaking for himself but the rest-of the family agrees with the spokesman to a moderate degree. 5. The spokesman seems to truly be expressing for all the family, which is in agreement with what is being said. Communication Breakdown: When communication breaks down what is most likely to happen? A communication breakdown is any instance where one family mem- ber fails to understand the meaning of another's message to him. Disagree- ment is not considered a communication breakdown unless there is obvious misunderstanding. For example, if one of the children fails to understand the meaning of the proverb as the parents are teaching it, what happens? Rate for all family members if possible, but only if such breakdowns are observed. More than one category may be applicable. 1. Communicator forces the receiver to profess understanding in an authoritarian manner; communicator may or may not ask for clarification from the receiver. 2. Communicator belittles the receiver for not understanding but does not force receiver to profess understanding. 3. Communicator and receiver reach an impasse, mainly because misunder- standing is seen by both as a disagreement and agreeing with the other is viewed as a loss of self esteem. Neither comes to understand the other‘s intended meaning. h. Communicator either ignores or simply fails to see that there is a misunderstanding and consequently does not have to deal with it. 5. Communicator typically acquiesces to receiver's perception of his message to avoid disharmony. 6. Communicator asks for clarification on the part of the receiver and makes an attempt to understand the breakdown. Restructuring of the message so that the receiver can make better sense of it usually follows. 7 o Other Discrepency: How much discrepency is there between what family members tell the clinician alone and what they tell each other in the family session? This rating is specifically of the interview task nWhat would you like to see changed in you or the rest of your family at the present time?" Rate each family member separately. 1. High degree of discrepency; individual tells clinician something but will not discuss it with rest of the family. 3. Moderate degree of discrepency; some of what is told the clinician is discussed with the rest of the family, and some is not. 5. Low level of discrepency; what is related to the clinician is also_ discussed with the rest of the family. RaJr'fl-lrlg Scale -3- (3d rev.) I. Freedom of Speech: How free and comfortable do members feel about speaking and taking positions? 1. Members feel reluctant to speak meaningfully to each other or for themselves. 3. Some inhibitions about talking to each other but can communicate at some times about some things. 5. Virtually all members feel free and comfortable to speak when they want to. II. BQLE RELATIONSHIPS A. C. D. Decision-Making: Can the family arrive at a common decision? (Rate if decision attempted during hour.) . w l. The family is so paralyzed with confusion and conflict that it seems incapable of reaching any decision. 3. Arriving at a decision is difficult, but the family does so with only moderate conflict and decisions are often rational ones. 5. The family seems capable of working together smoothly in reaching a rational decision. Stability of Decisions: Once reached, does the family sustain its decisions? (Infer from available information.) 1. Decisions made are extremely unstable and are constantly being altered, reversed, or criticized. 3. Decisions made are held to about as often as they are changed. 5. Once a decision is made, it is held to firmly and consistently. Decision—Making Process: How are decisions reached? Check the one category below that best describes the family being rated. l. The process is unstable, various members being involved, and decisions often seem to emerge out of a general confusion if they arise at all -- laissez faire. 2. The process is generally a one-sided, authoritarian one. 3. Decisions are reached through an equalitarian, shared process in which all involved members play a role in the solution. Balance of Power: Is there a boss? 1. No one seems able to take over leadership in the family, even though attempts may be made by one or more members to.do so. 3. While there is no consistent family leader, at different times various Inembers take over the "boss" position, but cannot hold it for long. 5. There is definitely a family leader who has control of the family whenever he or she decides to exert his power. (over) Rating Scale - -h- ‘ (3d rev.) E. F. G. H. I. J. Who is the Boss ~jin the clinician's judgment? _____ ‘ Who does the familygperceive as the boss? Sociogramatice alignment charts: Make sociograms on the basis of the following questions: 1. Towards whom does each family member show (overtly and covertly) the most antagonism? 2. To whom does each family member show the most warmth‘and affection (overtly or covertly)? Sexual roles: How appropriate are the overt sexual roles assumed by the male and female adults and children in this family? Rate once for adults and once for children. 1. Very inappropriate and reversed. 3. Moderately appropriate. 5. Very appropriate; stereotypically overdone. Alignments: Degree of Insight into Alignments: How aware, at all levels, is the family of the alliances and subgroupings which go on within the total unit? 1. Family seems to be completely unaware of alignments. 2. Family denies any knowledge of alignments, but covertly seems to be aware of their existence. 3. Family seems to be aware of some of the alignments but unaware of others. A. Family aware of most of the alignments but unaware of at least one alignment. 5. Family seems to be completely aware of all alignments. Overall Comfort with Role Relationships in Family: How comfortable does this family feel with their present role relationships? 1., Not at all happy; no one seems to be satisfied with his or her role. 2. Generally unhappy, but at least one member seems reasonably comfor- table and satisfied with the way things are. h. Generally comfortable, but at least one member reasonably dissatisfied with the way things are. ' - 5. All quite satisfied with present role relationships. Overall maturity: How effectively are the parents able to function in all ii}?}mf§E-afforded by the family context, i.e. spouse, parent, individual, prcvjde;. Rate each parent on this global dimension of psychological :25uru7f. a mature parent would seem to be one who gets satisfaction from €if€thf§ functioning in all his or her various roles. Rating Scale 1. 3. 5. -5- (3d rev.) Mature functioning in almost all family roles. Mature role functioning about as often as immature functioning. Immature functioning in almost all family roles. III. FAMILY CHHESIVENESS A. (n F: 1}. E. Cooperation: Can this family work cooperatively on a task or problem? 1. Little or no cooperation within the family; considerable squabbling or arguing; everyone goes his own way. 3. Moderate degree of cooperation, up to a point, but members insist on having their way, too. 5. Considerable degree of cooperativeness, of working together and sub- merging individual goals and feelings for the common goal. Cohesiveness: Identification with the family unit; a feeling of "we" rather than "I". Tolerance of Differences: 1. None; just a collection of seemingly unrelated individuals. 3. Modest degree of identification with the family unit. 5. Pleasure and pride in family unit E23 real feeling for it. Closeness: Is this family a close unit; do they like to do things together? 1. No closeness; members are quite isolated, separated, independent of each other. 3. Moderate degree of closeness, especially if something interests a member. 5. Quite close, tend to do things frequently as a group; seem to like and feel comfortable in each other's company. degree of real acceptance and understanding regardless of what they say: 1. No tolerance for differences of opinions or actions — everything has to be done in one way only. 3. Moderate tolerance for differences in certain areas and up to a certain point. 5. Considerable flexibility and tolerance for a diversity of Opinions and behavior by members of the family. Tengion: Is the family a tense or a relaxed unit? 1. Extremely tense family; no relaxation at all. 3. Mcderate amount of tension, but can relax and be casual virtually to same degree. (over) Rating Scale ' -6- (3d rev.) F. G. S. Relaxed and comfortable. Openness: Is the family unit a very tight and closed system, or is it receptive and open to others? 1. Family tight, closed, self-contained, suspicious and rejecting of others attempting to understand or entree the system. 3. Family reserved, but also able to open up for another person, especially after they get to know the person. 5. Family very open, accepting of others, bids them welcome, likes others to join in with them; friendly. Participation in Family Session: Does the family participate in the family session and to what degree? 1. Very reluctant to get involved in the interview; no real willingness to participate. 2. Entire family reluctant to get involved, but gradually at least one member begins to participate. 3 - 3. Most of family reluctant to participate, but at least one person does (may act as family spokesman). b. Most of the family willing to participate, but at least one person reluctant to do so, prefers to stay uninvolved. S. Family very willing to participate and be involved in family session. IV. FAMILY FEELINGS A. B. Spontaneity: How free is the family in eXpressing affect? l. Greatly inhibited, over-controlled ' ; members express little feeling and may be quite discomfited if this happens; reserved, deliberate. 3. Moderate degree of inhibition, but in controlled manner can express affect and occasionally can become spontaneous. S. Quite spontaneous and free with feelings. Anger: How angry is this family as a whole? 1. Intensely angry; furious; in a rage 3. Mederately angry. 5. Little or no anger. Control of Anger: No matter how angry the family is, how well is this controlled? Rate fer each family member separately and then fbr the family as a whole. l. Poorly controlled, family has little or no control over its ergression. ' - Eat W ‘‘‘‘‘‘ orhi mg C: ..n-.~‘ ‘ E. .- .. 5:- - F. Rating Scale 3. S. Direction of Anger: -7- (3d rev.) Mederately controlled. well controlled, perhaps even to degree of over—control. At whom is the family angry? (Indicate with whom the family is most angry.) More than one category may be applicable. 1. 2. Permissiveness for aggression towards parents: At everyone; everyone is mad at everyone else. Usually at one person at a time, but the target is constantly shifting. At one person, usually a child. Parents angry at Siblings and/or vice versa. Siblings angry at each other. Parents angry at each other. At persons outside the family group - neighbors, boss, teachers, etc. No one really; anger, when expressed, seems appropriate in terms of degree and object. Rate each parent separately. 1. Not at all permissive. Believes this is something one should not per- mit under any circumstances. Always attempts to stop child immediately; neither verbal nor physical aggression permitted. Slightly permissive. Feels that one must expect a certain amount of May’permit Moderately permissive. this, but that it should be discouraged rather firmly. some "sassing" but no hitting. Quite permissive. Completely permissive. Does not attempt to stop child from hitting parent or Shouting angrily at him. May express belief that child has right to hit parent if parent has right to hit child. Permissiveness for agression among Siblings; rate parent separately. l. \"1 Not at all permissive. Parents try to stOp quarreling and fighting immediately. Punish severely. Slightly permissive. Stop if somebody getting hurt; may allow Scolding given but not Moderately permissive. verbal battles if they don't go on too long. severe punishment. Quite permissive Entirely permissive. Parent never interferes in children's quarrels; they are allowed to fight it out. Parents do not try to stop or prevent things. (over) Rating Scale -8- (3d rev.) G. H. I. J. K. Severity for negative sanctions: Rate each parent on a scale of severity (harshness) for negative sanctions (physical punishment, privilege depri- vation, love withholding, and isolation) used in behavior training. 1. Low level of severity; negative sanctions in question are not admini- stered with hostility other than that which is a natural and appropriaua response to the child's error. 3. Moderate level of severity. 5. High level of severity; negative sanction used are very overdone and tinged with a great deal of hostility not merited by the child's mis- take. Techniques of Behavior Training: What kind of training techniques are most prevalent for each parent? From the following categories check the two most used techniques for each parent. I. Use of praise and approval. 2. Use of tangible rewards like money or candy. 3. Physical punishment. h. Deprivation of privileges. S. Withdrawal of love e.g. "Mommy doesn't like little boys who swear." 6. Isolation e.g. "Go to your room until you stop crying." Positive self regard: To what extent do family members see themselves as worthwhile individuals? Rate each individual and then the family as a whole on this dimension. 1. High positive self regard; able to state with pride and confidence one's own attributes and value. 3. MOderate self regard. 5. Low self regard; unable to make statements of self worth and even self depreciatory when asked to do so. M: Is this a warm family? 1. No warmth, members very cold and distant from each other. 2. Slight warmth, but members generally reserved and coOl. 3. MOderate degree of warmth, but somewhat inconsistent. h. Generally warm, but somewhat reserved in showing this. 5. Quite warm and open in expressing this to each other. Empathy: Is this a family which can feel for others, including members of the unit? 1. Little feeling for anyone but oneself. ’5‘: 3 Rating Scale ~ ‘ \F “‘,\ L. 3 N. 2. 3. S. -9- ' (3d rev.) Some feeling for others, but the self is still paramount. Equal feeling for self and others. Considerable feeling and caring fbr other peOple, both in- and out- side of the family unit. Nurturance: IS this a giving family in which dependent behavior is accepted? 1. 2. 3. h. 5.- Very withholding, ungiving, discouraging of dependency, no time for others. Somewhat giving and nurturant, but mainly around specific things (e.g. an accident) or time of life (e.g., to an infant). Nurturant, give comfortably to each other, not only in response to needs of each but in response to the nurturer's needs as well. Somewhat overnurturant - overprotective. Extreme overnurturance, to point of symbiosis. Independence: Ebes this family encourage independent behavior by its members? Check the one category below that best describes the family. 1. 2. 3. h. l. h. S. 7. Very punishing and discouraging of independent behavior, fearful of it, rewarding dependent behavior. Ambivalent about this area, encouraging and discouraging both depen- dent and independent behavior. Quite encouraging of it, but in an age-appropriate and realistic manner. Push family members to be independent at all costs. Happiness: How happy is this family? Very unhappy with each other and themselves; not at all pleased with the way things are going. Very happy with each other and themselves; pleased with the way things are going. Likeability: Do you like this family? 1. 2. 3. t. No, very unpleasant. Generally no, only occasionally so. Ambivalent, sometimes likeable, about as often not. Generally yes, only occasionally no. (over) Rating Scale U_ ”-10- (3d rev.) 5. Yes, this is a most pleasant family. v. ' MISCELLANEOUS A. Appropriateness of proverb interpretation: Check the category which is most appropriate for each parent's interpretation. Rate for each child if he or she makes an interpretation. 1. Complete literalness, e.g. "When the cat's out of the house the mice can play without getting caught." 2. Generalization from the literal. Subject subsumes a specific instance Cited in the proverb under a general statement -- a kind of psuedo abstraction, e.g. "Small animals can be more active when their natural . enemies are not present." ' 3. Concrete interpretation. A low level of abstraction. Meaning is given fairly well but tinged with literalness, e.g. "You mean like parents are cats and children are mice and the children play instead.of behave when their parents are away?" A. Moderate degree of abstraction, e.g.-"werkers are more likely to take advantage of their boss when he's not watching. 5. High level of abstraction, e.g. '*People are more likely to take advantage of authority figures having some power over them when those figures are not immediately present." 6. False abstraction. Faulty meaning attributed to the proverb, e.g. "Cats and mice are nice pets but you shouldn't keep them together." 7. Refusal to interpret the meaning. 8. Other B. Fantasy detachment: To what extent can family members imagine themselves performing an activity which they may never do in actuality? Rate each family member and then the family as a whole. Check the one category that best fits. 1. Unable to imagine or fantasy such an activity; Show confusion about the instructions to do so. 2. May not be unable to do so but refuse to do so by depreciating the instructions Or making excuses. 3. Can imagine themselves doing such an activity but it is an unrealistic one (in terms of the instructions to plan something they could actually do if they wanted to) e.g. a low income family planning a trip to Europe. W h. Are able to imagine themselves doing such an activity and plan one they have actually done, are going to do, or one they have never done before. MM:cc APPENDIX D INTERCORRELATICNS FOR CONSISTENTLY RATED ITEMS DURING INTERVIEW l '51 I\,U .sosos mo. ore em pesos es mmflaflsmm secede use anamoc Cmozpop ommeflEHaomflo aapcwoflwflsmflm Loans EmpH* as Aw H>H os.-xx o H>H on. me. we a H>H as. mm. oo. as a o>H em. sm.-ao. so. as QHH oo. we.-ws. sm. so. as «HH mo. om. as. os. me. so. as Mowae Ho. as. so. mm. me. as. sa. as a iH ss.-ea.-mm. mo. om. mm. mm. om. as me mm. mm. mm. mm. mm. ss. me. me. ms. as *z>H o oo. ms. or. as. so. ss. so. mo. mm. ms. as ao>H no sm. sm. oo. mm. oo. we. me. mo. am. om. ms. xx *szo>H mm. ms. sm. mm. oo. mo. om. am. we. mm. mm. mm. as *oHHH om. ss. mm. so. mo. mm. as. om. mo. ms. mm. as. so. as *oHHH mo. mo. sm. om. mo. ow. mo. mm. s». ss. oo. os. we. we. as *mHHH em. om. os.-ss. mm. ss. mm. mm. ms. ow. sm. am. sm. os. ms. as *szeHH Hm. ms. ms.-ms. mm. mm. sm. mm. om. ms. mm. oo. ss. om. ss. me. xx *Ie mHH oo. sm. mm. mm. am. mm. ma. ss. mm. mm. mm. mm. as. em. to. ss. ms. as see me. me. mo. am. am. om. mm. mm. mm. mm. ms. m . mm. as. ss. mm. oo. am. as aae I I I I I I I I I I I I I I I I I I I A A A A I I E .d a A A A I I I I I 3 WV Tl II II no G V \/ \J M” f 0 Ti Ti Tl w“ H * * ) \|/ ) ) O Tum * or ) G 0 HQ ) \I/ .A O a .a /\ /\ W * r * W .a or * * .fim>oa mo. one pm show some pcmaowwflo aspcmoflmflsmflm we as. we pCOflOflwwooo soapmfioamoo m ma 3oe>aoQCH mafiaoo nampfl oopmm a pcopmflmmoo Mom mCOflpmHmaaoopepCH APPENDIX E INTERCORRELATIONS FOR CONSISTENTLY RATED ITEMS DURING INTERVIEW 2 81 ‘2' ll, r—‘.1.IHJJI.‘II .1 I‘ .xm>ox mo. vcp pm umwma pm mmxxxfisx pxcxx. ram xaExo: :mwzpmn UquCHExuomxv AxucwomecmHm soap: amth xx . 3.ng om. xx . . <>H mm. me. xx . UHHH mm. mx. 00. xx . _ . nxx mm. mm. mm. mm. xx . xx. xo. mx. mo. mm. co. xx . . x xx xm. mo. ox. mx. xm. om. xx 0 xx mx. mm. xm. xx. mw. om. mm. xx x xx mm. mo. xo. om. ox. mm. xo. mx. xx . m an em. xm. xx. om. mx. cm. xo. ox. om. xx . . . mu xx. mx. ox. om. mm. mo. Ox. mo. xo. mx. xx H. mm. ox. m . mm. xx. 00. om. mo. xo. xm. om. mm. xx . tx>H mm. mm. m . xm. mx. mo. mm. xo.-mx. mm. mm. m . mm. xx . *n>x ox.-xm. mx. om. mmd mo. mo. mo. mo. mx. xx. m . ox. xm. xx x x H>H mx. mm. mm. mx. m. mx. mm. om. mm. ox. xx. xx. m. mm. mx. xx , . o H>H xm. mx. mm. xm. m. ox. xx. mo.-xx. ox. xm. xm. x. om. mm. xm. xx . t m H>H mm. mm. xx. xm. xo. xx. mx. xx. mm. ox. om. ox. om. om. xw. mm. mm. xx . x : u>H mx. mm. ox. ox. mm. mm. mm. mo. mx. xo. mm. xx. me. xx. ox. mx. mx. om. xx x m u>H xm. mx.-mx. mo.-mx.-mo.-mo. mm. xx. mo. mx. mo. xo. mo. om. mm. ox. xx. 2.. xx . xo>H ma mm. xo. mm. mo. mo. 00. xx. mm. mo. mx. mm. ow. we. mx. me. xx. me. xx. xx. xx. xx xm>x nmv x. mx. we. xm. xm. mm. mm. mm. mm. mx. mm. mm. mx. om. mm. mm. xx. an. ex. ex. mo. xx . *mxxx x. om. mm. mm. xx. mx. ox. mo. xm. xm. x. mm. mm. om. wx. wx. xo. ox. 7x. ox. xx. xx. xx *mxxx xx. mm. on. mx. ox. xm. xm. mo. xx. mx. 0. mm. ow..mm. mx. mx. am. on. xx. “a. ox. mm. mm. xx . :mxxx xm. ox. mm. xm. xx. mm. mx. mx. om. mm. mm. mm. om. mx. ow. xx. xx. m-._mc. ,2. xo. om. xm. xx. xx xmx m9. mpg on oxmu Sega pcmxmuuxu Axucwoxchwxm.mx xx. x: Cqumxwxxou w xx zox>xmpcx wcxxso maovx cmuuu Axucwumxmcoo how mcoxudfiuuuoouopCH , -.‘ -‘7 ~- _ -~- ‘__. «.- _ _ -. 7- , . _ _ “'1'“. ‘- "‘fiifiriurinlgwmmmw“