THE COMMUNTCATTON 0F AFFECTS WITHIN; _ cums AND NORMAL TAM'TLTEsi 'i'hesis for the Degree of M. A, MTCHSGAN STATE .UNTVERSTTY “ MICHAEL ARTHUR wass 1968 SSSSSS IILI] HITTTZIITTTIITTTTTT Ufllfliflflflfl ‘ J 2 L I B R A R Y Michigan 7 ta tc Universjr 461‘“! ABE TRACT '13:; CHM-PUNK»;CEIIQN UF 33.1"}?! 1.3138 'v’II’i‘Hlfl CLINIC Iii-3D MRMAL FAMILIES by MiCAael Arthur Heias This study waa unfiertakon in order to explore a question pcaad by family interaction: Can we differentiate betweca normal and clinic families with regard to the communi- cation of negative and positive affects? *ight normal and seven annortal families (four or five members each) partici- pated in a semi-structured interview during which they inter- acted as a family in three task activities. Normal families had no known history of ycychiatric disturbance or treatment and were Obtained through labor union and Church groups. The source of the abnormal families was the treatment waiting list of the Michigan State University Paycnclogical Clinic. It was predicted that tha devised Rating scale for the Communication of Affect (RbCA) would be able to cis- tinguish between normal and clinic families, with the clinic families exhibiting lofiar scores on the positive affects and higher ones on the negative affects of the rating scale. Nine out of the thirteen items yielded inter-rater reliability coefficients that were significantly different from zero at Michael Arthur £6195 - «~- ~ ‘ -' - ,n’ ~'- - .. _ . MK??- L. r-v. fi-.».-..~.1-L .4. tut? 1.1" 9'} 5".‘3' 1:0 I .- attrihztm‘i to Hie vmfz’ 10w ivviflmcw cf twee 83':‘€:f;‘t£ i. .1 the {is 311‘; #21. recorded companicat QT“. c . ’ - In ‘: g e. “0.1-. -I.I.~o 1. I. P xiv-NH! ‘9- “:5": [IQE‘R‘E‘L fr. 16113.5.“ 5 V'E'rfi'.‘ ‘2} till“; 1,1151%: f...L'...1‘. t}; 7‘2" tie indivifiual itsma. Th6 clinic families exéiLZtsfl si3* nificantly were urgetive effectb, sucL 59 deyxtnsisn. hsxtili- t 3 (5 I} ‘_| 21. g) D I ty. azcicty, CLnfusicn and dcunle mcazagss, wccrsrw r? clinica shQW£fi lore gciitive cornaiicationml foif falsifieas :35 éialayue, than éii tha clinics. Ewttxczrore, the clinic ranilies exhiciteé mars extrema affiéctivc re- ars th41 did 1‘: I; w, w ’1‘. with a much meta limitéd range of h. 9" IV. (\ :7) I 'r‘ a: cfio ”- th" mun~clinic famil Evaluaticn of tha study‘s {indimga grmr ts tié ewm garimantcr to suggest further research in the daveicgvent of reliatlc rating scales as well as interactiamel tank! suited to the fitudy of familie;. THE COMMUNICATION 0? EFFECTS WITBIH CLIHIC LSD NORMAL FAMILIES BY Michael Arthur Weiss A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MAsTER OF ARTS Department of Psychology 1968 (35/ ”77/ z .’ ’f' 7 I! _ __.l' - / '. I 4.." PC KNOWLEDG RENT S I would like to express my appreciation to Dr. Gary E. Stollak and Dr. Dozier w. Thornton for serving as members of my thesis committee. I would especially like to express my gratitude to Dr. Lucy R. Ferguson, my committee chairman, who provided me with invaluable aid during the course of this undertaking. In addition I would like to thank Lennard Leighton for his assistance in devising the Rating Scale for the Communication of Affect (RSCA) as well as for the many hours of rating tapes and tabulating data. Finally, I would like to thank Dr. Marv Moore for the use of the taped family interactional interviews, without which the study could not have been executed. ii ACKSOWLEDGMEHTS s . TABLE OF CONTENTS 0 O O O C O O C 0 LIST OF TABLES . C I O O C ‘ I O O O 0 LIST OF APPEED CBS 0 Chapter I. II- III- IV- V. IETRDDUCTIOH s . s . . . . . . Statement of the Problem . Relevant Literature . . . mmon I O O 0 O O O O I O O D Subjects . . . . . . . . . The Interviewing Procedure The Rating Scale . . . . . RE SULT S O 0 O O O O O O I O O Inter-Rater Reliability . Test of Hypothesis . . . . DESWSSICS I O O O C O O O C . Inter-Rater Reliability . HYWthQBiU s s e s s s s e SUMEARY s . . . s . . . . . . REFEREECES s . . . s e s . s . . . . . APPEmICES 0 O O C O O O O O O O O O 0 iii Page ii iv 18 18 24 29 29 32 38 38 40 47 49 52 LIST OF TABLES Table Page 1. Comparison of normal and clinic family groups on several compositional criteria . . 21 2. Inter-rater reliability for items of the Rating Scale for the Communication of Aff'flCtsssssssssesssssssss 3O 3. Intercorrelationa of Positive RSCA Items . . . 31 33. Intercorrelationa of Negative RSCA Items . . . 31 4. i's and t ratios between groups and within taek categories for positive and nega~ tive itemfl Of the RSCA s e e e s s o e s s s 33 5. i's and t ratios between groups on each item OfthGR-SCA-ssssassssssssas 35 6. Total scores, means, standard deviations and 1 ratios of the clinic and non-clinic families on the 13 items of the RtCA . . . . 36 iv LI 51’ OF AP m‘mzcas Apponfiix Page A specifications of families aamyled in this study: families 1-8 ate the non—clinic group and iamilies 9-15 are the clinic groug.........o......... 53 B Total scores for RSCA Items: Clinic versus Eon-Clinic families . . . . . . . . . . . . 55 CHAPTER I INTRODUCTION Statement of the Problgg Recently there has been a large amount or research on the small group. Unfortunately very few of these studies have used the family as their focus and even fewer have studied the family group in any systematic manner. There- fore, there is a need to deal systematically with this most significant group~—a group which, due to its prior common experiences and expectations of continued relations. can yield much more information of a heuristic nature to re- searchers than can any ad hoc group. or the relatively few studies in the past few years dealing with the family, most used time-saving techniques such as questionnaires or impressions of the observers. Al- though this may have been efficacious with regard to time, money and ease of experimental design, such devices allow such unwanted variables as interviewer and respondent biases (which subsume a whole host of other undesired variables) to intrude upon the scene. Direct observation of behavior of an interacting family when coupled with some objective rating scheme will uncover its dynamics and will prove to be closest to the natural real-life situation. The present study sought to study normal and clinic family interactions through the device of a rating scale de- veloped to measure various affective modes of communication. It predicted that the rating scale would be able to dis- tinguish between normal and clinic families. More specifi- cally, the clinic families would exhibit lower scores on the “positive” effects and higher scores on the 'negative' ef- fects on the Rating Scale for the Communication of Affect (RSCA). ‘elevsnt Qitergture The stugyngg the Family as a Small Group It is only very recently that psychology has joined psychiatry, anthrOpology, sociology and social work in the area of family interaction. Handel (1965) traces the notion of the “interaction concept“ from a 1926 paper by Burgess in his plea for more systematic studies by psychological re- searchers in this all-important area. He decries the fact that psychology has been content to study the individual as if he functions in a vacuum. While it is true that there have been many well- executed experiments done in the area of the small group, there arises the question of the heuristic value of many of these studies. Strodtbeck (1954) asks how far we can generalize from the results obtained by many of these studies. He was seeking to test the apprOpriateness of certain propositions concerning sd-hoc three-person groups in a study dealing with the father, mother and son as a three- person group. Etrodtbeck favored the family over sd-hoc groups because of its prior common experience and expec- tations of continued relations--factora which make the ex- perimental situation more like that of real life. In com- parison with Mills (1953) strodtbeck found that when the two most active members are in conflict, the stability is not as low for families as the former had found with ad hoc groups. Thus, it appears that many of the previous notions of the family which have been generalized from experiments on ad hoc groups may have to be modified accordingly. To date the field of research on family interaction is in a rather amorphous and chaotic state. Frame (1965) states. ”Despite the prodigious literature on the family, there is no body of formalised literature on systematic re- search on family dynamics with clearcut stands taken on is— sues and specific limits from which departures can be made“ (p. 409). He also points to the unfortunate fact that there has been an extensive reliance by behavioral scientists upon questionnaires in family research. In addition the members of the family are usually seen separately rather than in the presence of other family members. Their conscious reports are subject to memory distortions and conscious and un— conscious biases and falsifications. Framo foresees much heuristic value of family re— search as he quotes from Hare (1955) “the study of small groups is a microscopic study of small cultures which has implications for the study of social systems, of cultures at large and of personality. The virtual ommiseion of the family as a subject of systematic small-group investigation is singular indeed“ (p. 412). Frame points out to us that there is a great wealth of knowledge to be gleaned by obser- vations of interactional behavior within the family. He makes reference to the studies of Birdwhistell and Scheflen, who made very detailed microscopic analyses of interactional behavior by drastically slowing down film recordings. In a study encompassing supplementary methods in family research Levinger (1963) provides an overview of the many techniques which have been utilized by past researchers in the area. He quotes Bill (1958) who examined marriage and family behavioral study techniques of 1945 to 1956. Of 422 studies, almost half used interviews, questionnaires or tests: a fifth utilized literary, legal or historical docu- ments: one tenth of those studies were based on statistical summaries: one tenth were reliant upon the impressions of ob- servers or upon various other unspecified methods: one tenth were ethnographic behavioral observation. Levinger found that each of the above techniques had its merits. The questionnaire or interview is economic and focused. He states that competent judges avoid the 5 distortion of self-report. Behavioral observation yields a first-hand sample of interaction, although we receive no rich introspective data, Levinger points out. This method avoids distortions and gains insight into the functioning of the whole group. The observer can record action as it oc- curs and predict more directly future action. Problems of the respondents' unawareness or unwillingness to report the critical behaviors are greatly reduced. Levinger notes that it is also possible, unfortunately, for biases stemming from the subject’s reactions to being observed and for the un- representativenesa of the observed situation to,distort the picture we obtain of the family in question. Levinger sees the situational study technique of family research as being the best. It brings the family re- search closest to the usual mode of the scientific method while still remaining as close as possible to the real life situation. He adds that this is the most costly method in terms of time per subject. Also, without introspective re- port it may give misleading or unrepresentative impressions of family patterns. By virtue of the fact that it may ins fluenca the very process which is being studied it may pro- duce artificial behavior on the part of the family members. As soon as researchers began to realize that many of the older methods utilized in family studies would have to be supplemented by other techniques, a wealth of studies be- gan to come into existence with regard to new devices for family interactional exyeriments. Blood (1958) now has the observer taking notes on the spot. He believes that this will keep the observer active as he watches the family inter- acting in the home setting. This also enables the family to ignore the busy, non-participating reporter after a while. The observers concluded that volunteer families may be ob- served in this manner without causing a distortion in their behavior. Blood concludes that observation is a necessary adjunct to interviews and questionnaires. Blood calls for the develOpment and testing of obser- vational categories to capture family interaction sequences. He decries the use of questionnaires in that they all too often leave unstudied little things that go on in family interactions which are more important than such classic questions as "When did your parents first explain to you about menstruation?“ In order to glean as much as possible with regard to the above-mentioned “little things that go on in family nteractions” Strodtbeck (1351) made tape recordings in ex- perimentally induced decision—making situations in order to be able to trace the role of each family member. He was studying mother-father interaction within three cultures (go- ing from a mother-dominant Eavajo to a tether-dominant Hermon). The balance of power was revealed in the families who were instructed to reach a common consensus on problems about which they independently disagreed. The dominant partner was judged to be the one who won the greater number of decisions. Caputo (1963) has demonstrated the superiority of direct observation over a paper and pencil test in a very graphic fashion. arents were instructed to discuss ten items of the Parent Attitude Inventory which they had pre- viously answered in divergent fashion. He had derived relatively benign pictures of the parents. relationship from Osgood Semantic Differential data. However, analysis of the overt interaction of the parents revealed considerable an- tagonism and bilateral hostility. In another early study attempting to demonstrate a significant relationship between family interaction and mental illness Jackson, Riskin and Satir (1961) sought to escape the trap of proving this in a retrospective manner. In their at- tempt to identify different patterns or family interaction and to relate these to the form of emotional disorder in the referred ”patient,” the authors utilized a communication theory ayproach. It was shown that the more disturbed fami- lies uttered a greater number of incongruent messages as compared to sequential disqualifications (i.e. “Yes" and then changing it to “so') than do normal families. Riskin (1953) continued to seek out variables which could distinguish normal from pathologic families. He dis- cusses covert and overt messages, with the former, an extreme form, indicating pathology. The experiment involved a series of tasks, such as “Plan something you'd like to do as a family, etc.“ The specific variables on.which the family was rated were: "clarity,“ "content," "agreement,” "com- mitment,“ 'congruency,” “intensity,” etc. It was found that in coalitions of any two members to the exclusion of a third, in families with schizOphrenic children, the two are really not talking about the same thing although they seem to be. Riskin states that the healthier families exhibit what he terms ”teamwork“ or meeting in the middle of the road in coming to decisions. As many other researchers in the area, Riskin concludes by stating: 'we do not have yet available adequate tools for evaluating the assumed relationships be- tween family interaction and personality formation‘ (p. 348). More recently a method was devised by Drechslsr and Shapiro (1963) which compared clinical and statistical anal- yses of the same data in order to test hypotheses about a given family and to compare different families. The EXpeti~ mentors arbitrarily selected twenty-minute clinical samples to present in capsule form the family's characteristic path- ology. In addition, from each family session, twenty one- minuts segments were extracted at equal intervals and were scored independently for the number of times each person spoke to each other person. The same authors had previously done research in the same area in a study which provided the family members with tasks to be completed without the presence of the investigators (1961). They were to discuss together a family questionnaire of 20 items, containing factual as well as fantasy item: (3.9. "Draw a diagram of the home with all the rooms.“ “What sort of thing does the family argue about?“ “Describe how one special holiday is celebrated by the family.“ “What was the worst nightmare that each of you ever had?” “If each of you could change one thing about yourself and the other members of the family, what would you change?") Although this latter piece of research did not involve the saphisticated statistical methods of their later work, Drechsler and Shapiro had provided a setting where the family could act out its characteristic relationships and thereby reveal interactive patterns of which they were unaware. Other investigators have decried the great diffi- culties encountered by researchers in this area, such as the lack of suitable methods for investigation and the paucity of coherent and specific theories which identify the criti- cal variables and describe their presumed function. rarina, Storrs and Dunham (1963) used a “structured situational test“ for assessing the effect of the relationships within the family on the behavior of the patient who is a member of that group. Each parent separately came to a solution of a hypothesized Problem and then the family, consisting of the father, mother and son, arrived at a joint solution.of that same problem. The patients consisted of two groups; ”good 10 and "poor" premorbioa. It was found that there was more con— flict in the family interaction of poor premorbida and that the fathers were more cominant in the good premorbid group. In another study involving abnormal and normal family groups interacting in task activities, Ferreira (1963) had each family member separately make his decision about three comparable neutral items (9.9. ”It you were to take a trip to Alaska next month would you rather go by train, car or by boat") and to then order them in term: of his own preference. The second phase of the study involved the whole family ce~ ciding upon the same items with the awareness that the new ordering of the items would have to take into account the wishes of the other members. Ferreira states that four broad categories or group decision: could be distinguished. They were “Unanimous,“ "Majority,“ ”Dictatorial“ and “Chaotic“ decisions. In the Clinic Pamiliee it in often the child who éecioee what the family‘ggggg 5 want, whereas in the Eormal Families Ferreira noted that he found the greatest agreement between the preferences of the individual and those of the family. Replicating his previous (1363e) finfling concerning differences of eycntaneoua agreerent between normal and ab- normal families, and that abnormal families took more time to reach a joint decision which was often quite inappropriate with regard to the wiehee of the individual family members, Ferreira (1963b) investigated some further phenomena cropping 11 up when normal families are compared with those in the clinic. In a family interectional activity, Perreira states that the individual member in the Normal Families expects rejection in an amount that is commensurate with the amount that he tends to display. Perreira likens this to the tali- onio, oye—for-en—eye principle. However, he notes that in the Abnormal Families, the principle can be stated more on the basis of "two eyes for an eye” or as “no tooth for a tooth.“ _ One solution to the problem of the chaos in the field of family research is reported by Hill and Eaneen (1960) at the University of Minnesota's Family Study Center. At this institution, there exists an inventory to codify substantive findings, research procedures employed and theoretical propo- eitione derived from findings in family exteriments. The authors have nicely categorized the various approaches to the study of the family and identify the discipline in which the conceptual framework was derived. The "International" approach was developed in sociology and social psychology: the ”Structure-Function" in sociology and social anthro- pology; the ”Situational“ in sociology: the “Institutional" in sociology and historical psychology: the “Developmental” in sociology and borrowing from rural sociology, child pey- chology and human development. Another critic of family interaction atudiea, Haley (1964) states that the methode which have been used to date, 12 namely statistical, anthropological, individual and intox— national are inadequate due to the fact that they must mace-- sarily rely on the subjective reports of observers, inferw ontial catagoriea and the possibility of the raters possoss- ing some common bias. Goa of Haley's notions is that it is not important what the individual does, but to whom he reapoafia. In hll aaalyais of recorded speech, to discovered that tho order of Speaking and the patterns of order in abnormals showed non- randomoesa, wherean ranfiomuoas of speaking order and patterns or order we: exhibited by normal families. In normal fami- lies the interchange between mother and child wen highest; that of father and child was least. In the abnornals the mothor-father interchange was highest and that between father and child was least. Another unusual method of studying family inter- action sequences was devised by Riakin (1964). he had fami- 1165 plan something tagather which thoy could do with each other. His thesis was that significant incongruoncies be— tween body movements and vocal behavior can be fouoa which will possess correlates in incongrucncima hetwean tonal and verbal behavior. Ee concluded that hy just listening to family interactions. as does the average clinician, much valuable information on the family being stufiiod is missed. Riskin believes that it is possible to make clinically meaningful and accurate discriptiona of whole familios and 13 the various members through the use of scales in blind in- terpretations of tapes. The problem of studying the family remains a compli- cated one. For, as Ackormnn (1957) states, “the moment we get absorbed in rigid statistical study we find ourselves counting the number of freckled mothers and fathers" (p. 74). The whole family should ideally be treated as an individual case study and should then be compared with others. It is neceseary to have categories which can be applied universally. Referring to an early study executed by Moustakas, Sigal and Schalock (1956) he states it is also essential that the cate- gories be comprehensive and objective. Situations should be as close to real life as possible. Ackerman (1957) cautions us on the problem of apecifaction of bases for judgment in analysis of the data, i.e. the evaluation of actual life per- formance: against a psychiatric model of "ideal" family and individual mental health. we must be aware that value judg- ments enter into the picture as well as do cultural biases. A good deal of the research done in the area of family interaction has been executed so that we may gain in- sight and understanding into the ways in which families have come to be disturbed and how this disturbance in perpetuated within the family. Much of human response: to the behavior presented to them by their fellow humans in contingent upon the talion principle of "an eye for an eye a tooth for a tooth“ as mentioned by Perreira (1963b). In his study, the 14 payment was found to be two-fold in schizophrenic families in the case of negative responses and only half—fold in that of positive behavior. Thus it is not unexpected when two cisturbea adults marry, that their progenty should help to maintain the vicious circle. Rckermen (1933) noted that boys possessing problems with regard to conduct very often come from families where the parents are hostile or rejecting. He found that the child is not just a passive recipient of this hostility but in rather an active participant in the reinforcement of hostility within the family environment. Raueh, Pittman anfl Taylor (1959) showed that in groups of aggressive and normal boys that aggressive behavior generally begot aggression from others, whereas friendly be- havior generally begot the same in return. This tends to be in line with the principle of reciprocal relationn of Leary (1957) which holds that "interpersonal reflexes tend with a probability greater than chance to initiate or invite re- ciprocal interpersonal responses from the other person in the interaction that leads to a repetition of the original reflex" (p. 123). The unfortunate problem with disturbed familieo is, accorfiing to Leary, that they poueeee a narrow range of interperlonel behavior. Iney tend to repeat the same behavior which is for them a way of avoiding .niiety. It aids in minimizing conflict and provides for them the sa- curity of continuity and sameness. however, the price which 15 they must pay for this security is a restricted 503131 33» vizonmont with little Opportunity for growth or change. In a st'fi; of fiio hwsgitali 3311163, with each p35333"i -3 a 33:15: 3 2d; 3 to by schizovbsmo Lo, £53dy'3 (1353) fi 3L Mi :3 not or y corroEQrate Lear; but they Canada stxato even n-Jro cont lo: wizaly the 333335 biais 31133 zbu :1 0 n 1": {ma familie$ tenfl to «at thomboluua 1323. [3335 ohfi3rved hat the ctc:o-t;L£L r3133 in the digtuxbod families wo;o like thoaa in a medLeval morality play whe the “actors ta 9 all 03.3 r1351 roles £33 positiona that are ster3otv§oj anfi confirfi i--oao 13 632-3, another Evil and a third Temgtation" (p. 390). A very earlyv study in group behavior done by Kurt Lewi a (1938) “amonstra tea very clearly some of tla ante~ cefient variables resgoxsibla for n3gative and positive 32- teots existing within the group. Lewin set up two groups of children--onm autocratic and the otb er den :3: atic—-to work together in the profluction of hanR-crafted articles. A much higher atato of tension was $3333 to exist in tho autocratic group along with about thirty-one times as much hostility as Ln tie democratic 9:0 39 On the other hand he found more praise 3.3 .d frienfiliaess in the democratic group which led to 3 hi; her dagre a of cogstruoti mass as juggofl by s oporicr group products. Mora recent studies, such as that of Bishop (1951) 335 Mackenzie (1968) point to the oorralation between the 16 rigidity of roles in the group, extreme degree of negative effects and consequent lack of possitivs affects with psycho- pathology of that group. MacKenris demonstrates that in a normal group of families more friendly responses were sent than in a matched clinic group. The behavior in the abnormal group also tended to be more extreme. thua pulling more ex- treme responses. The normals tend to show a wider repertoire of responses than do the clinic families. Bishop round that children reflected directly in their own.behavior the mother's use of directing-interfering-criticism. strong stimulation and suggesting types of control and also the tendency toward nonacceptancs ot stimulations. She also noted that as the child becomes mors tsmiliar‘with a “neutral adult“ in the experiment, his behavior approached that displayed to- ward the mother. Thus, we see that the disturbed families tend to re- strict their behaviors to a much greater extent than do the normal families. Unfortunately when certain types or be- havior are negatively reinforced between mother and father, the children do not always play the game well. or perhaps they play it too well, for sometimes the parents may not even realize which behaviors in the children they are rein- tercing. we all know of the mother who believes that she is negatively reinforcing aggressive behavior in her child when she slaps him in the face and says, “I don't ever want to see you hitting anybody again, Johnnyt'I Unfortunately for 17 the disturbed family, Johnny adopts his mother's behavior and contributes to the general pool of negative, destructive behaviors within the family. It is not very often in a die- turbed family that parents transmit positive behaviors to the children since they themselves show little such behavior. CHAPTER I I METHOD The data for this study were gathered by Marv Moore for his doctoral dissertation, carried out in 1966 at Michi— gan State University. Dr. Moore was kind enough to provide this experimenter with a series or taped family interactions. The families in the study were comprised of tour- or five-member unite (both parents and two or three children) meeting the following criteria for acceptance: (1) all families had lived together for at least four uninterrupted years previous to participation in Moore's project: (2) the families had at least one male child between 8 and 13: (3) children ranged in age from 5 to 13 years old: and (4) they met the criteria listed below for inclusion in one of the two samples studied. The sample groups were defined by Moore in the following manner: Korma; group, consisting of 8 families where none of the members had ever received, or had ever been recommended to receive, any type of psychiatric treat- ment for an emotional or nervous disturbance. Mbore obtained the sample of normal families from two sources. He states 18 19 that three of the families volunteered as a result of his solicitation of subjects at labor union meetings in the Lansing area. The five other family units volunteered after they had been recommended by their minister as representing the “most emotionally mature” families in his congregation. Each of the normal families received $20.00 as payment for their coOperation in the two interviews. Cl nic rou , con- sisted of 8 families waiting for psychotherapy at the Michi- gan State University Psychological Clinic, the treatment to involve both parents and sometimes one or more children: none of the families received any treatment during the course of Moore's study. All clinic families selected for Moore‘s experiment had arranged for psychotherapy with the clinic as a result of a male child between age 8 and 13 having been re- ferred for underachievement and/or lack of behavior control in school. The clinic families, unlike the normals, were not paid for their cooperation due to the fact that the treat- ment agreement at the clinic at Michigan State University is that families participate in some ongoing research project: there is no clinic fee. The present study had been initiated with the hepe of utilizing the first interview for each of the sixteen families. ‘Unfortunately some of the tapes of both the first and the second interviews were missing with no possible hope of their retrieval. Therefore a method had to be devised to provide a reasonable sample size as well as a sample that 20 would retain the original normal-clinic differences as found in Moore's study. It was decided to combine data from first and second interviews where both were available, otherwise to use whichever was available. The problem was that while Moore had found no di fer- ences between the behaviors exhibited by the clinic families in the first and second interviews he had found a signifi- cant difference between the normal families' first and second interviews with regard to the behavior which he was studying. The latter finding Moore attributed to a favor- able practice effect: the normal families were able to profit from their first interview experience and hence perfect smoother ways of carrying out the second interview. There- fore while the present study utilised a completely different rating scale for effect, the experimenter had to decide whether or not he would be justified in combining first and second interviews in.order to provide a reasonable sample size for himself. In essence the problem was one of devising a legiti- mate method of combining the two interviews. After all of the first interviews had been listened to and rated (there were thirteen out of Mbore's sixteen remaining, 7 normal and 6 clinic) the experimenter rated the second interviews (there were eight out of the fifteen remaining. 4 normal and 4 clinic: of Moore's original sixteen families who had ap- peared for the first interview one had refused to show up 21 for the second interview). Finally, a sample of fifteen families emerged-—eight normals and seven clinics. Scores received by the families were pro—rated so that a composite score was arrived at for each of the fifteen families. no significant differences were found with regard to the ef- fects measured in the first and second interviews. Where the first and second interviews for a family still existed on tape the scores for each of the thirteen items on the scale were averaged. Where only the second interview ex- isted (in.one normal and one clinic family) the score for the second interview is that which was reported. Thus, there were a total of 15 families included in this project, 8 normal and 7 clinic. Inspection of Table 1 shows that the two groups are essentially similar with re- spect to composition, except that the mean level of the fathers' education is 1.4 years higher in the Clinic sample. For this small a sample, the difference is not significantly different from chance expectation. Table 1. Comparison of normal and clinic family groups on eeveral composition criteria. .Mean years Mean Mean oidgggiizz°d number of age of children children Father Mother per family per family Normal Family Group 12.8 12.3 2.3 10.2 Clinic Family Group 14.2 12.3 2.6 10.6 Moore began the family interviews by greeting the families and obtaining some very global facts about them. This aided in establishing rapport between himself and his subjects. He then introduced the family to the first con- joint task. Preliminary remarks made by Moore include re- iterating that there would be two raters who wouldbe viewing them through the one-way mirror. and also that the purpose of the study was to provide information which would hopefully increase our skills in aiding families. While Moore's experiment involved the family's par- ticipation in nine tasks, taking approximately one and one- half hours to complete, the present study has utilised only three of those tasks with a total time of about one-half an hour -. The tasks on the interview schedule occurred as follows: Task ls Moore spent enough time with each family member so as 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 family members were reporting this information to the experimenter the remaining family members waited in an adjoining room with the instructions not to discuss the question among them- selves. After all members had reported to the experimenter 23 the family met again oonjointly and carried out the following instructions: “Discuss among yourselves the question I have just asked each of you separately; you may discuss any as- pect 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. Yen will have about four minutes, or more if you need it. I will not take part in your family discus- sion, but will remain quietly in the room.“ The present study utilized only that part of the task which involved the family as a whole discussing the changes they reported and how they'would like to bring them about. The reason is that here we are attempting to devise a rating scale to be used on interactions in which the en— tire femily participates. 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 Moore's study, first the father, mother and children: then the father and the children ex- cluding mother: mother and the children excluding father: and husband and wife excluding the children all planned activities (with Moore present in all but the first). Again, the present study, due to the fact that it is interested in 24 family unite, did not rate the task planning activity in which only the father and mother participated. Task 3: The parents received a proverb: while the cat’s away the mice will play (Firet interview). In the second interview the proverb was: A rolling stone gathers no moan. The experimenter asked them to discuss the meaning of the proverb between themselves with the fact in mind that they would be required to teach it to their children. When the latter wee accomplished the interviewer requested the parents to retrieve the children so that they might be taught the proverb'e meanin4. The present study rated only that family interaction which came about as the parents attempted to communicate the proverb'e meaning to their children. The rating scale was develOped by this experimenter in conjunction with Lennard Leighton, a student at Michigan State University. It grew out of a series of observations made by these two students who listened to many taped inter- views of family interactions. After having listened to the tapes and having noted a series of varieblee on which the families could be rated, the two of us examined the item: which we had selected. Many of these original item. (there were over forty of them in all) were eubeumed under more global categories. with the help of Dr. Lucy Ferguson of Michigan State University, 25 other categories were added and some of the categories which were drawing a great deal of data were broken down into eamller ones. Thus, in the end, a group of thirteen items made up the final version of the Rating Scale for the Com- munication of Affect (RSCA). Due to time limitation: the experimenter could not afford to train other raters. Therefore all families were rated by the exocrimenter himself and Lennard Leighton. This may well have introduced some biae into the results, since both ratere had knowledge of and investment in the hypotheses, and it proved difficult not to be aware of which group a family belonged to. In addition, the ratere being clinical psychology student: may have exhibited a bias toward looking for pathology in all families. The raters gave each family a check mark whenever one of the items on the RfiCA wan ex- hibited in one of the measured family interaction sequences. Thus, the date for each family could be broken down into the three tasks and each teak was scored for any of the thirteen items on the RSCA which occurred. All tapes were rated in random order and the raters presumably did not know to which group any of the families belonged. The ecorea for the familiee on the RSCA are the averages of the two raters. Pearson product-moment corre- lations were utilized in comparing the scores aeeigned to the familiar by the two raters. I. m ”Vi-1:9: allows-young: . u I! 3. etiflf‘ Ema-14: $3.: t3‘fiw “fif'KlL (_i 1'"I‘_I_'h":.‘.fi_- .05.?! Ass-.1»- first EMPEfiIY I P03 ITE‘FE ‘J " .i ‘2'... r... J~r ""hf'" °a _fiIAQQGUB: of gigoc ct {Ric [ Marked examples of total Eami ly en- thusiasm, spontaneity, genuine group laughter sné vocal inflections. This is to be contrasted with examples of laughter which is sarcastic and biting, s.g. laughing at, not with, and/or which involves only one or two family members. fianifestsd as I positive, warm feeling toward others: an aypreciation and prion of others within the family. In- volves either verbal report, such as “I like you“ and/or such behaviors and statements from which affection can be inferred. Marked examples of ono family member asking for_§§1 trulv wisf aing to hear and respect the opinions of another member of the family. Involves genuine respect and understanding of another person's feelings, wishes and opinions, s.g. “I know how you fool.” Involves the placing of positive values on the ideas of oneself or of others or, in a similar manner, upon the worth of various objects. Examples are, “I like that,‘ “That's good.“ This is in regard to how the various members of the family impart knowledge to each other. Involves a democratic moon of the communication of ideas. Involves the Socratic notion of a sharing of idoas and opinions with neither sios “right" or ”wrong.“ Each is willing to respect the loans and opinions of the other until a common ground is reached. 27 II “on?!“ cos-7- qnewfinn O :f“.“3‘ gul- ‘v 4-: fi‘ ., ask-1.. 6. E‘FEEEC C3: Often manifested as very vocal sighs or groans on the part of the members of the family. Involves also at times crying, complaining, apathy and inap— propriate silences. 7. 'ggfiggggggx Is very much the opposite of nffcction. ' Involves the utilization of critical and disrespectful commence, such as “shut up:“ Also, curses, mocking tones, sarcasm, destructive criticism, rejec- tion, ignoring other family members, bitterness and resentment. 8. .gfifilfifY: Often involves long silences which have been induced by parents as a re- sult of severe, negative criticism, especially of the children. Involves theses of fear, guilt. xenifested often by feelings of tensauess, agi- tation, nervous laughter: extreme de- fensiveness of salt and or family. 9. (Emil-(35102.3: Lack of continuity, lack of understand- ing, interruptions, ignoring other's ideas in favor of oneself, complete noise and chaos, going off on topics which are tangential to the family task to be conploted. 10 . 1‘01??? igrzgszs. Is essentially a failure or communi- cation in the extreme sense. Examples are, “I don't give you too much house- work to do Johnny, now do I?“ Saying one thing and meaning another. Leaves the recipient of the message in an ex- treme bind. ll. Involves loss of control in a situation in which anger is warranted. However, implies that there existed other, less negative ways, in which to respond to the frustrating situation. 12. 13- ..' d . v m... .....::...... = 1". ECLIIW} a 23 Involves the placing of negative values upon the ideas or oneself or of others or, in a similar manner, upon the worth of various objects. Examples are, "That stinks,“ ”That's no good,“ ”I don't like that.“ Involves an autocratic mace of the communication of beliefis and opinions. The sender is unwilling to respect the ideas of the receiver should they be contrary to his own. The speaker is like Moses, trying to away his audience in an autocratic manner. CiiI-‘J’TLR III RILEUL‘I‘S Inferffgter Rtliebilit *— —-v-.— The rating scale for the Communication of Affect (RECA) contained 13 items on which the fifteen clinic end nonrclinic families were rated as they participated in the task. The inter-rater reliability coefficients (Pearson r'e) were calculated for each of the scale item» on the RSCA using the total number of judgments (see Table 2): the number of paired observations for each reliability coefficient con- sisted of all families in the study (15). The final data analyeie was derived from pooled judgments scores (rater l + rater 2) on those items where the internreter reliability coeffiiciente were found to be significantly different from zero when p - .05 (i.e. an r of .44 or greater). The nega- tive items were judged far more coneietently than were the positive ones (i.e. a mean inter—rater reliability of .22 for the positive items as compared with .66 for the negative ones.) It will be noted in the next section that even some of the less consistently rated items differentiated between the two groups of families. The low interereter agreement on some of the items seems related to the very low occurrence 29 of these affacts. (Sea Table 6 whare total scores of each group on each item are reported.) Table 2. Inter-rater reliability for items of the Rating scale for the Communication of Affect. ;;v-;;--¢- -z.........il—.; «- -——-£ .1: *1 1 ‘3‘; ”“2" 21* 2m RSCA Item Label r Across All Task: _Ehfiitive gtemn HAPPIHESS '.59 AEPECTIOH .49 EHPAIdY .21 PO$ITIVE VALUE JUDGMEKTS «.09 DIALOGUE -.03 E . .22 21.12533 0 'v' @212 nEpnsssxos .66 HOSTILITY .87 AEKIETY .44 cannusxox .92 DOUBLE MESSAGES .39 APPROPRIATE AflSE-R . 45 ELerrxvs VALUE JUDGMERTS .30 Pagacsxms .77 E - .66 Referring to tables 3 and 3A, to test for nagetivo halo effect, an all-too-fraquenc contaminant of atudiaa, 31 Table 3. Intercorrelationa of Positive RSCA items. .. 25:: z============================================= HAP.‘ DIALOG.‘ EMPATH. AFFEC. pus. VAL- HAP.* XX .69 .25 .70 .87 DIALOG.* XX .71 .77 .34 EfiPfiTH. XX .81 .35 AFFEC. XX .42 VAL. XX 'Item which significantly discriminated between normal and clinic families at least at the .05 level. Table 3A. Intercorrelationa of Negative RSCA items. W '0. App. ate. DEP.* uoar.t ANX.‘ cour.* MESS.‘ AHG. VAL. PREACH. DEP.* XX .92 .35 '.67 .07 .32 .37 .24 HOST.* XX .47 .79 .28 .52 .82 .21 ARK.' XX .76 .07 .14 .16 .06 COEF.' xx .22 .77 .54 .49 D. MESS.* XX .05 -.06 .18 APP. ANS. xx .21 .37 - VAL. XX .01 PREACH. XX ____ *Item which significantly discriminated between normal and clinic families at least at the .05 level. 32 intercorrelations of negative and of positive items were _calculated. The mean intercorrelation within the positive items was .59 and was .35 within the negative items. Thus, no strong negative halo effect seems to have been present. zest o; Hypothesig The hypothesis stated that the RSCA ratings in the three task sequences would differentiate normal and clinic families. It was also hypothesized that the normals would score higher on the positive items and that the clinic families would score higher on the negative items of the RSCA. These hypotheses were tested in various ways. Since the normal and clinic families participated in three task activities, i.e. Changes, Planning an Activity and the Proverb, and were judged by the RSCA, which contains items with regard to both positive and negative effects. over-all patholOgy scores could be arrived at for the three tasks as a whole and within the separate tasks. The results of these comparisons are presented in Table 4. Looking across the three tasks (Table 4), it can be seen that the combined negative and positive items on the RSCA nicely differentiate the non-clinic from the clinic families. With regards to the negative items, the t ratio at 3.15 (p - .01 with 13 d.£.) shows a highly significant difference with the clinics exhibiting more negative effects. The ratio of 2.16 for the positive items (9 - .05 with 13 33 d-f.) indicates the significant difference between the fami- lies once again, with the non-clinics demonstrating more positive effects than the clinics. Table 4. X's and t ratios between groups and within task categories for positive and negative items of the RSCA. I Task Category Item Polarity normal Clinic t ratio Across Tasks Positive 8.44 2.50 2.16* Across Tasks Negative 5.06 22.07 3.15“ Changes Positive 2.87 .71 2.45. Changes Negative .69 4.50 3.46" Activity Positive 4.38 1.50 1.99 Activity megative 3.56 14.87 3.62** Proverb Positive 1.31 .29 2.17* Proverb Negative .81 2.71 1.38 0p - .05. **p - .01. within the separate tasks, the hypothesis of a sig- nificant difference between normal and clinic families with regard to positive and negative affects is further sub- stantiated. In the Changes task the t ratio of 3.46 (p - .01 with 13 d.f.) for the negative items and the t ratio of 2.45 (p - .05 with 13 d.f.) for the positive items establishes 34 the validity of that hypothesis ae doee the t ratio of 3.62 (p u .01 with 13 d.£.) for the negative iteme within the Activity tank and the t ratio of 2.17 (p - .05 with 13 d.f.) for the positive items within the Proverb task. The t'a of 1.99 for the positive iteme of the Activity task and 1.38 for the poeitive items of the Proverb task were in the predicted direction but non-significant. Hence the hypothesis of a eignificant difference between the families along the linen of positive and negative atfecte has been substantiated. Table 5 demonstrates further the differences between the non-clinic and clinic families, thie time with regard to each of the thirteen iteme.o£ the Rating Scale for the Communication of Affect (RSCA). Seven of the 13 items have distinguished the families at the .05 level or better. A greater percentage of the negative items than positive one: have dietinguiehed the two groups (respectively, 5 out of 7 as compared with 2 out o! 5). Table 6 above the rather extreme affects exhibited by the clinic families ea Opposed to the more moderate ex- tent and range of affecta manifested by the nonrclinice. More specifically, the positive effects were rare and iimited in range in the clinic families and were moderate and showed a fairly wide range in the non-clinic group. With regard to the negative affects, the clinice showed a great deal as well as a great deal of variance. The non—clinics however exhibited a moderate amount and tended to very much less than did the clinics. 35 Table 5. 7 ratio. between groupl on each item of the RECA. ——-— Normal Clinic RSCA Item Label Manna Means t ratio. _gcqitiveiltemg KAPPIHESS 1.1 3.30 2.78. AFFECTION .29 1.00 1.69 EMPATHY .06 1.00 2.02 POSITIVE VALUE JUDGMENI‘S .57 1.00 1.1.8 DIALOGUE .36 2.00 2.88. flmtiwve 11:919. DEPRESSION 3 .07 . 31 3 . 38" HObTILlTY 4.57 .19 2.43. ANXIETY 1.64 .25 3.07" COKFUSION 6.00 1.63 3.380. DOUBLE MESSAGES 1.14 .19 3.03.. APPROPRIATE ANGER 1.14 .56 .71 NEGATIVE VALUE JUDGMENTS 1.93 .63 1.71 PREACHING 2.51 1.31 1.78 .P . 005- *.p - .01. Only four items, COEPUSION, APPROPRIATE ASSER, PREASHING and POSITIVE VALUE JUDGMENTS, give us group vari- ance: which do not lead to P ratios significant at the .05 level. The other nine items have divided the clinic and non- clinio groups along two difforont continua. This was to be expected to a large extent with than. items since we are dealing with two vary dissimilar groups. A more microscopic treatment of the flute which com— pared the scores of the two group. to: each item and within 36 .aopma mo. gnu u» ucuoflunamaa naoausn o«uauuv odde01oo: can vanquo any mo noucmwuo> on» ad ouaououuwo a Invocaunnc on." v.« ~m.u m.oH c~.n hm.“ . mzHMU(mmm o a.o no. . no. o.n Nh.d mm.d QEZHXQQDH mnfld> Wauaflomm min hat" on. m.‘ MO.H finin “mafia afi w>HBHmOm Cmm.h m~.H oo.d 0.5 ed. 00. m.o ”makafim com.md om. OQ.H o.m on. mm. o.N BOHHmehfi ohm.v «N.d om.n o.wN mm. H.H m.h mmmmnmmflm mmmwmgwpauwwom Gwyn“ u .a.m one: onoom .o.m noes ouoom Henna noun Guam HouOB HOHOF o«n«~u:ooz Outwau III, I? r JIIw'l It II III-Iii! lulu .oo cuonanuu .uoooe .oouoon and .w wanna 36 .«uuaqc u«a««u.aoc can uuawuo any mo moccaaun> .~o>0H no. oz» uo macawuwdmau unoauon onu ca oucouomuwo o nouuoauaHo an.“ ..H ~m.~ m.o~ on.“ pm.“ o.m~ manmu flaufiflflflm 0..“ NH.H 6m. m.‘ no.H Qdod O.m “Wmfi‘ QBHaHmom .mn.~ m~.a so.” o.m ea. no. m.o mmaxmmm .om.m~ so. oo.~ o.m on. mN. o.u aouaowmmm .hm.v H~.~ om.n 0..“ mm. A.” m.h mmmgummem Odunfi h .Q.m 6&0! CHOOQ .D.m aflofi Ouoow Haflflfl EOHH <0wm mouoa Hauoa oqaqao.con agenda ._.EB< .mma‘ .hfiOO .xmd .Fwom .mmfi .GHQ .Afi>+ .mZh .hhfl .m¢m .mmd on 33H «92 .332: i «Hill! fit f .nodaaficm u«n«HUncoa neaum> uwcwuo unaouH 40mm How acuoom HaUOB ill .41 ““ll 16" m fingmuwunz X/E/M' HICHIGRN STRTE UNIV. LIBRARIES lllll lllllll ll Illll HI llllllllllll 9 697 312 31022 88