ABSTRACT A PILOT STUDY OF AGREEMENT ON ISSUES AND THEIR PERCEIVED IWORTANCE AMONG MARITALLY ADJUSTED AND “ADJUSTED COUPLES by Arnold S. Carson Marital adjustment within the last three decades has been researched by various behavioral sciences. However. it has been only within the last decade that modern statistics and theories of personality have been used in investigating marital adjustment. None of the previous studies on marital adjustment has been replicated. The purpose of the current study was to explore two methods of developing a marital adjustment scale inferred from two reSpective theories while deriving content from known spousal agreement correlates of mm adjustment. The first of these theories embodies the concept that marital adjustment is a function of Spousal agreement while the personal construct theory assumes that marital adjustment is a function of’ spousal agreement in interaction with mates' perceived importance of ‘ their agreeméhtso A secondary purpose of the current study was to . provide a replicated scale of marital adjustment borrowing the measurement procedures which are typical for personality inventories. An original pool of 1140 items for the experimental scale (labeled the "Issues Scale") was constructed. These items were derived from the following seven spousal agreement correlates of marital adjustment: "Handling Finances", "Recreation",' ”Dealing with Ill-Laws", ”Intimate "4 it 1‘; '13:;- 337. cu! \‘x Arnold S. Carson Relations", "Friendsi’. ”Demonstrations of Affection", "Philosophy of ‘Life". The scale was administered to a pilot group of fifteen couples to gain their reaction to the wording and ordering of items in the scale. The scale was then administered to the main sample. 47 maritally adjusted and.h8 maritally maladjusted couples. The maritally adjusted couples were operationally defined as those scoring above 109 on the Locke-Wallace Short Marital Adjustment Test.‘ The maritally maladjusted couples were either starting procedures for divorce or receiving professional services for diagnosed marital problems. The subjects were randomly divided into a validation and cross validation groupe Each item of the original pool was individually validated and cross validated in accordance with its ability to differentiate significantly between maritally adjusted and maladjusted couples of the cross! validation group. Responses were scored by two alternate methods according to the two reSpective theories described earlier. It was found flhat the scale scored according to the theory embodying only Spousal agreement produced #8 validated and 20 cross validated items. The scale scored according to the theory embodying spousal agreement in interaction with mates' perceived importance of the issues produced 38 validated and 13 cross validated items. The 48 validated items scored.for spousal agreement differentiated significantly between the maritally adjuSted and maladjusted subjects of the cross validated sample at a p«<.0005 while the 38 validated items scored for Spousal agreement in interaction with spousal.perceptions of issues differentiated significantly between the reSpective groups at a p (.00 5. The reliability of the scale was estimated for each of the groups in Arnold S . Carson the validation and cross validation samples by Hoyt's2 Analysis of Variance Method. Although the straight Spousal agreement scoring was more productive of items, the configural scoring was more reliable (straight agreement scoring: r = .4? to .77; configural scoring: r = .60 to .90). It was concluded that the traditional interpretation of compatability in marriage as a function of Spousal agreement in certain well defined areas is a misleading framework fostering only partial scaling of marital adjustment. The traditional framework warrants alterations. Compatability in marriage is a function of spousal agreement in certain well defined areas and on the perceived importance of these areas. The areas of spousal agreement which have been correlated with marital adjustment in previous studies included "Handling Finances", "In-Laws". "Recreation”, "Intimate Relations", ”Demonostrations of Affection”. "Friends", and 'PhiIOSOphy of Life”. In the current study where spousal agreement on perceived importance of issues was considered, 69.2 per cent of the items were drawn from the areas ”Dealing with In-Laws", "Handling Finances" and "Friends", and no items were drawn from the area "PhilOSOphy of Life”. ‘ H. J. Locke and K. M. Wallace, "Short Marital Adjustment and Prediction Tests: Their Reliability and Validity." Marriage and Family We XXI (1959). Pp. 251-5- 2 c. J. Hoyt. "Test. Reliability Estimated by the Analysis of Variance Method," Psychometrika. V1091”). pp. 267-87. COpyright by ARNOLD STANLEY CARSON 1962 A PILOT STUDY OF AGREEMENT ON ISSUES AND THEIR PERCEIVED IMPORTANCE AMONG MARITALLY ADJUSTED AND MALADJUSTED COUPLES 0‘ Arnold 5: Carson A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY College of Education 1961, ACKNOWLEDGEMENTS I would like to express a debt of gratitude to my advisor Dr. William Farquhar who Spent many hours guiding my research efforts from the initial preposal to the final draft. Acknowledgement‘is also given here to the members of my doctoral committee: Dr. John Krumboltz. Dr. John Essen, Tm. billism Hail and Zr. Hamehek. Much thanks goes to my wife. Lilyan. who typed and proof read all drafts of the thesis. ii. TABLE OF CONTENTS IAIST OF TABLES O O O O 0 O O O 0 O O O O O O O I 0 Chapter I III THE PROBIEM O 0 0 O O O O I O O O O O O O O 0 me Success or Failure of Marriage . . . . . Die Personal Construct Theory of Human Behavior The Psychology of Personal Constructs . . . The Similarity Theory of Marital Adjustment . ThePurposeoftheStudy . ... . . . . . . . GeneralHypotheses ............. REVIEWOFTHELITERANRE ......'.... Background Factors in Marital Adjustment The Importance of Background Factors in Marital Adjustment EWHCMCFacmrS 0000000000. The Impress of Cultural Background . . Age and Duration Factors . . . . . . e Methods of Scaling Marital Happiness . . . Psychological Factors in Marital Happiness Predicting Success or Failure in Marriage . Further Developments with Existing Scales . Predicting Adjustment in Marriage . . . . . Current Attempts to Scale Marital Adjustment The Development of the Criterion Instrument Summaryeoeeoeeoeeeeeeeeoeo MESIGNOFTHESTUDY........... C O O O O O ThemSigneeeeeeoooe The Independent Variables . . The Dependent Variables . . . Consensus of Mates on Issues . O O O O O O O O O O O O O O O O O O O O O O O Consensus of Mates on the Perceived Importance of Issues. The Configural Scoring System . . Coping with Extraneous Variables meSample ........... Magrammatic Plan of the Design . 'D’ISWPOtheseS..ooeoeooo The Statistical Analysis . . . . . Summary............. 0 O O O O O O O Page ‘OVO‘WJIC'N 11 11 11 11+ 17 21 26 A L 32 39 43 67 69 73 73 73 73 75 . 76 7? 80 80 83 86 Chapter IV Famous AND recess-tats . . . . . . Instrumentationeeeooeeeé Criteria Instrument . . . . . . The Ebtperimental Instrument . . Selection of Subjects . . . . . . The Maritally Maladjusted Couples The Maritally Adjusted Couples . The Validation and Cross Validation Groups The Sample Characteristics . . . . . . . . Administration of the Instruments Summary ANALYSIS OF THE RESULTS 'Dxe Set '1' Scale . . The F Scale 0 e e o Analysisofthelhta. .... Restatement of the Hypotheses Validation Group Data . . . . Cross Validation Group Data . Estimates of Reliabith . . . . Attribution of Scores to the Items 0 C O 0 Development of the Set '1' and F Scales O O O O O O O Eaglerimental Scale Cross Validation of the Issues Scale Summary SUMMARY AND CONCLUSIONS . . . . Sumary . . . . . . . . TheProblem . . . . . TheDesigl...... Methods and Procedures Remts O O I O O 0 0 Limitations of the Study Conclusions . . . . . . . Suggestions for Future Research 0 O 0‘. O O O O C O C O O O Bmummm O O O O O O O O O O O mmmcm O O 0 O O O O O O O O 0 Appendix A Appendix B mmdi-x c O O O O O O O O C O 0 iv. e. E C O O O O O O I O O O O O O O O O O O O O O O O O O O O O O O O O O O O I O O O O I O O O O O O O O O O O Page 88 88 89 91 92 93 94 99 100 100 102 102 102 103 106 106 107 108 109 111 112 116 119 119 119 120 123 125 128 129 131 133 137 137 11+? 150 LIST OF TABLES Table P age 2.1 Occupations Distribution of Divorces and Ehxployed Males in Iowa (Excluding Fann Occupations) . . . . . . . . . . 20 2.2 wration of Age by Separation Dates and Divorce Dates: Philadelphia County Sample. 1937-1950 . . . . . . . . . . 211 2.3 Duration of First Marriage in Desertion and Non-Support Cases, Philadelphia County. 1950 . . . . . . . . . . . . 25 2.4 Correlation between Ratings of Marital Happiness and Extentoi‘AgreementsandIlisagreements . . . . . . . . . 37 2.5 Relation of Items of Agreement amd Disagreement to Ratings of Marital Happiness Itans of Agreement and msagrement O O O O O O O O O O O O O O O O O 0 I 0 0 O 38 2.6 Per cent of Happily Married and Divorced Couples Whose Average Combined Marital Adjustment Scores Fell in Spedfic Intems O C O O 0 C C O O O O O 0 O O O O O O “7 2.? Findings of the Three Methodological Approaches in A Comparison of Five Levels of Marital Adjustment on Various MMPI Scales . . . . . . . . . . . . . . . . . . . 53 2.8 Current Attempts to Scale Marital Adjustment (In the Chronological Order of First Publication): Years 19W196000000000000eeeeoeeoeeeoe7o 3.1 MagrammaticDesignoftheSmdy ............. 81 4.1 Number of Years Education for Maritally Adjusted and Maladjusted Men and Women in the Validation and CrossValidationGroups................. 94 (-1.2 Frequency of Maritalb' Adjusted and Maladjusted Men and Women in the Validation and Cross Validation Groups Who Have Attended Trade School or College . . . . . . . . 95 4.3 D.O.T. (Part IV) Occupational Classification of Maritally Adjusted and Maladjusted Men and Women in the Validation and Cross Validation Groups . . . . . . . 96 4.11- Ages of Maritally Adjusted and Maladjusted Men and Women in the Validation and Cross Validation Groups . . . 97 V. Table l}. 5 u.6 _5.1 5.2 5-3 5.1+ 5-5 506 5-7 5.8 Figure 1 LIST OF TABLES CONTINUED Number of Years Married for the Maritally Adjusted and Maladjusted Couples of the Validation and Cross ValidationGroups................. Number of Maritally Adjusted and Maladjusted Male and Female Subjects Married More than Once . . . . . . 'lheFScaleItemsandRareResponses ... ..... Number. Accumulated Proportions. and 3 Scores of RareReSponsestotheFScale. . . . . . . . . . . Number of Items Generated by the Configural Scoring and Straight Agreement Scoring for Validation and CrossValidationGroups.............. The Hoyt's Analysis of Variance Estim3te of Reliability fortheIssuesScale ............... Mean Score and Standard Deviations 01' Maxitally Adjusted and Maladjusted Validation and Cross Validation Groups ontheIssuesScale..,................ Cross Validation Group's Mean Scores and Standard Deviations on the Validated Items of the Issues Scale . The Cross Validation Maritally Adjusted and Maladjusted Couples' Scores on the Validated Straight Agreement ScoredItems ..................... The Cross Validation Iviarz'1.talljr Adjusted and Maladjusted Couples' Scores on the Validated Configural Scored Items 0 O O O O O I O O 0 O O 0 O C O O O O O O O O 0 0 LIST OF FIGURES The Configural Scoring System for the Ewerimental Scale Page 98 98 100 105 109 110 112 113 111+ 11‘! Page CHAPTER I THE PROBLEM Marriage, one of the oldest and most traditional institutions, has become an object of study by the newest of the sciences. The conduct of modern marriage. though given token definition by its traditional counterparts, is no longer Specifically prescribed by these same counterparts but instead by the personalities and personal relationships of the mates. Burgess and Cottrell give credence to the changed regulators of marital conduct and point to their implications for adjustment in marriage Marital adjustment must, then. be defined in the modern conception of marriage. Adjustment is not insured here, as it is in flhe orient, by customs and ceremonies minutely regulating the conduct of the young married couple. If marriage has become a personal rather than social relation. adjustment is to be defined in terms of personalities, and the degree of assimilation taking place. 1 It is not unlikely that if the conduct of marriage is a function of the personalities and personal relationships of the mates that the success or failure of modern marriage may be diagnosed and prognosticated by the study of the same variables. During the period between 1929 and 1939 psychologists, sociologists, and psychiatrists spent a goodly portion of their research energy describing and forcasting adjustment in marriage. Since this period, the same disciplines have devoted their energies to refining their methodologies to give credence to the 1 E. w. Burgess and L. s. Cottrell, Jr., Predic Success or Failure in Mange (New York: Prentice Hall. 19395, p. ‘10. complexity of the variables which make an adjusted or an unadjusted marriage. Theory building and the application of already existing theory is a recent development of note in the study of marital adjustment. The current study represented the development of a marital adjustment scale in which both the items and scoring of the scale were inferred from theories. The Success or Failure of Marriage The large increases in divorce rates over and above the increase of pepulation in the United States has drawn attention to maladjustment of marriages as a serious social problem. While the pOpulation increased only 215.? per cent from 1870 to 1930, the number of divorces increased 1,607.8 per cent. The divorce rate rose from 28 per 100.000 people in 1370 to an estimated 193 per 100,000 people in 1937.2 In 1932 there was one divorce for every six marriages. Current estimates indicate that the ratio of divorces to marriages has increased further. The lack of divorce does not necessarily imply an adjusted marriage while divorce almost unanimously implies maladjustment in marriage. It is conceivable that divorce is used as a means to other ends. However, the remaining aura of sanctity surrounding the institu- tion of marriage plus the necessity of legal evidence proving a marriage intolerable has confined the use of divorce for release from an undesirable marriage. Divorce has served as a decisive and clear cut criteria of marital maladjustment in previous investigations. The criterion of adjustment in marriage is more sought and less 2 S.A. Stouffer and L. M. Spencer, "Recent Increases in Marriage and Divorce," American Journal of Sociology, XL (January 1939), p. 552. \J) 0 found than the criterion for maladjustment in marriage. It is clear. though. that a criterion of adjustment in marriage must mani est a strong negative correlation with divorce. Burgess and Cottrell3 eXplored the possibility of having 526 couples rate the happiness of their marriage on a f ve e-point scale. They found that 1) most people can estimate the happiness of their marriage on a five-point scale regardless of the ambiguity of the terms "happiness", 2) husbands and wives usually agree in their estimates of marital happiness, 3) an outsider who is fairly well acquainted with the marriage will generally agree with a member of a couple on his happiness rating, and a) two outsiders reasonably familiar with a given marriage will usually agree in their appraisals of the marriage. Moreover, people's estimates of their marital happiness do not fluctuate markedly over short periods of time, but reliable and stable ha:g mess rat: mg are only crude indices of marital adjustment. Burgess and Cottrell interpret their results, "It must be admitted, however, that the evidence does not enable us to say how much of apparent agreement in the estimation of happiness in marriage is due to a generalized unconscious tendency to rate marriages as they are supposed to be rather than they are." u Terman studied eig ht-hundred married couples and found a seven point happiness rating to be skewed for his sample in the direction of greater happiness than would be expected were his sample normally 30.46. i 3 Burgess and Cottrell. Q2. cit., pp. 0 Ibid., p. 44. u. distributed. The author attributes the skewness to the inequality of scale units, selective influences and the omnipresent "generosity" tendency in personality ratings.5 Due to the difficulties of obtaining a large sample of divorced couples who would be willing to participate in a study on marriage, both of the above studies had to rely upon the happiness ratings of relatively adjusted couples as a criterion for further correlates of marital adjustment. Burgess and Cottrell included 18 divorced couples in their sample, and Terman included 50 and 75 couples known to be maladjusted. The extent of divorced couples in both studies were not proportionate to the ratio of those divorced during the period from 1929 to 1939. However, both Terman's and Burgess and.Cottrell's studies are classical presentations of what variables comprise an adjusted marriage. They did not purport to determine any cause and effect relationships. The determination of cause and effect in the adjustment or maladjustment of marriage was a task left to a future generation of researchers, refinement in technology, and the development of applicable theories of human behavior. The Persona Construct Theory of Human Behavior A number of current personality theoreticians have constructed theories of human behavior embracing three aspects: perception, behavior, consequences. Stated in simplest terms these theories posit that the behavioral and affective consequences in human behavior are V 5L. M. Terman, Psychological Factors in Marital Happiness (New York: McGraw Hill, 19387. pp. 148-83. 5. 6 a function of a person’s perceptions. In the following seetior a :9‘ review is presented of the aSpects of Kelly's Theory of Personal 5 Constructs’ ard.their impli Ma ‘ions for scale Feveicpnent in marital adjustment. ‘ The Psyghplogy of Personal Constructs Kelly formulated his theory in terms of the following fundamental postulate and eleven corrollaries: A. A person's processes are psychologically channeled by the ways in which he anticipates events. 1. A person anticipates events by constructing their . replication. 2. Persons differ from each other in their construction of events. 3. Each person characteristically evolves, for his convenience in anticipating events, a construction system embracing ordinal relationships between constructs. 4. A person' 5 construction system is composed of a finite number of dichotomous constructs. 5. A perSon chooses for himself that alternative in a dichotomized construct through which he anticipates the greater possibility for extension and definition of his system. 6. A construct is convenient for the anticipation of a finite range of events only. 7. A person's construction system varies as he success- fully construes the replication of events. 8. The variation in a person's construction system is limited by the permeability of the constructs within whose range of convenience the variants lie. 9. A person may successively employ a variety of con- struction systems which are inferentially incompatible with each other. 10. To the extent that one person employs a construction of experience which is similar to that employed by another, his psychological processes are similar to those of the other person. 6.... ‘ , , salv1n. all and Gar ner Lind2.ey, Theories oijersonaiitv (United States: .iley, 1;,e,, chaps. 8 and 12. 7George KEllY. The Psychology of Personal Constructs (New York: fiorton, 1355,, pp. Lo- -113'~. 6. 11. To the extent that one person construes the construction processes of another, he may play a role gn a social process involving the other person. Kelly's basic postulation and.first three corollaries suggest that human behavior is a function of the manner in which individuals perceive events and anticipate the consequences of behavior. His formulations applied to marriage imply that the relative adjustment of a marriage is a function of the mates' perceptions of events in their marriage and anticipated consequences of these events. Corollaries numbered 10 and 11 are the foundations for inferring that similarity of perceptions are based on similar psychological processes. The personal construct theory has not been considered in scales of marital adjustment which are scored only for responses to the content. However, the implications of Kelly‘s theory may be studied by affording the subject with the Opportunity to make a choice regarding his perception of the item to which he is responding. Accordingly, items which subjects perceive as being important.to them should have more discriminatory power than items which are perceived as unimportant. In the context of marriage it can be inferred that items which mates perceive as important have greater implications for the consequences, marital adjustment or marital maladjustment. than items which the mates perceive as unimportant. The Similarity Theory of Marital Adjustment In both common folklore and clinica1.practice it has been assumed that the similarity of mates is associated positively with adjustment 7. in marriage. In 1938 and 1939 Terman9 and then Burgess and Cottrell1O studied similar and disimilar mates and their reSpective marriages. In both studies similarity of mates was Operationally defined as the agreement of mates on certain well defined issues thought to be relevant to marital adjustment. In each of the studies agreement of mates in areas such as demonstrations of affection. friends. and dealing with in-laws correlated with marital adjustment ranging from an r = .33 to .70. Scales of marital adjustment for the last two decades have in part-been based upon the correlations established by Terman and by Burgess and Cottrell. Continued emerience with such scales has suggested that the relationship between agreement of mates and marital adjustment is a fruitful basis for the further study of adjustment in marriage. The Purpose of the Study 'Ihe purpose of thepresent study is to construct a scale of marital adjustment from a conceptual framework of the agreement of the mates subsumed under Kelly's personal construct theory of behavior. The items of the scale will be develOped through the systematic eaqaloration of seven Spousal agreement areas which in past studies were significantly correlated with marital adjustment. The scoring system of the scale will be configural in order to incorporate the inferences from the similarity theory of marital adjustment. The items also will be scored only for straight agreement of the mates on issues in order V 9 Terman, pp. 9312., pp. 118--83. 10 . Burgess and Cottrell. 92. .c_i_t_.. pp. 30-116. 8. to determine whether or not the configural scoring system increases the reliability and validity of the scale. Previous scales of marital adjustment or marital happiness have incorporated the notion that similarity of the mates or agreement of mates is related to marital adjustment. However, the construction of marital adjustment scales has been a pragmatic venture. The variables subsumed under the concept of similarity of mates have not been derived from an integrative theory of behavior; thus, any interpretation of cause and effect relationships could not be logically inferred from the reSponses of the couples. Instead scale constructors have had to confine their interpretation of findings to statements of concomitance. i.e. low marital adjustment is related to variables X. I. 2. Another characteristic of existing scales of marital adjustment is that there has been no systematic exploration of any one of the variables known to be concomitant with marital adjustment. 'me typical approach in constructing such a scale is to weight gross items in accordance with their extent of correlation with marital adjustment as rated by friends. relatives or experts and then give scoring deference to the higher weighted items. .A third characteristic of existing marital adjustment scales is that they are not constructed to allow for individual interpretation of protocols. In order to afford interpretive material on the individual tested. a scale would have to measure in some way the individual's perception of the items to which he reSponds. Recent advances in research on marriages suggest that marital 9. adjustment may be successfully scaled by configural scoring techniques11 and that the concept of similarity of mates as it is related. to marital adjustment is subsumable under a personal construct theory of human behavior. 12 General Hypotheses T‘wo general hypotheses are stated referring to the two alternate 'scoring systems used in the study. The first hypothesis is constructed to consider the effects of mates' perceived importance of issues as well as their consensus on the truth or falsity of issues. The second hypothesis refers only to mates' consensus on issues correlated with marital adjustment. A third hypothesis is stated referring to the content of the scale. 1. Maritally adjusted couples agree more than maritally maladjusted couples on the truth or falsity of issues correlated With marital adjustment which they perceive important. 2. Marltally adjusted couples agree more than maritally maladjusmd couples on the truth or falsity of issues correlated with marital adjustment regardless of the perceived importance of the issues. 3. Items derived from spousal agreement areas known to be correlated with marital adjustment will discriminate 11 R. J. Swan, "The Application of A Couple Analysis to the IvIMPI in Marriage Counseling" (Unpublished Ph.D. dissertation, University of Minnesota, 1953). 12 a. J. Corsini, "Understanding Similarity in Marriage." Journal of Abnormal and Social Psychology, LII (1956). pp. 327-32. betxveen maritally adjusted and maladjusted couples. A more comprehensive report of the research pertinent to the current study will be presented in Chapter II. In Chapter III the design of the study will be discussed with reference to the null , hypotheses. alternative hypotheses and the statistical analysis. The selection of the subjects. administrative procedures and the instruments involved in the collection of the data for the study will be found in Chapter IV. The accumulated data will be presented and analyzed in Chapter V. CHAPTER II REVIEW OF THE LITERATURE Marital adjustment is usually studied from a sociological or psychological frame of reference. In order to develop a clear cut theoretical foundation for a study of marital adjustment, the relation- ship of sociological and psychological data to one another needs to be organized in accordance with the logic of science. The first part of Chapter II contains a discussion of the principles for relating sociological to psychological data. In the second part of the chapter the sociological data on marital adjustment as background factors are reviewed. Alflhough they are kept in mind as possible sources of extraneous variance. no attempt is made to critique the literature in this area. In the last section of the chapter attempts to scale marital adjustment are reviewed and critiqued. Unlike the research on back- ground factors, the adequacy of previous scale construction must be determined in order to weight the theoretical importance of past scales for the development of the experiment scale. Backgrounthactors in Marital Adjustment The rtance of Back re d Factors in Marital Ad ustment What is considered a background factor in marital adjustment is largely a question of discipline. The psychologist may consider economic status and dwelling units as background factors while the sociologist may focus upon the above variables placing other variables such as impulsitivity and ego strength in the background. The integrity 11. 1‘ 14. of results from psychological analysis and sociological analysis need not contradict each other since they each afford explanations at their appropriate levels. A hypothetical situation may be invented to examine the above issue, which is one of reductionism, more thoroughly. A sociologist contends from empirical investigations that a lack of cultural opportunities is the main cause of divorce. A psychologist studies the same sample and contends that hostility of the mates is the main cause of divorce. The two investigators are not contradicting each other but merely studying the same phenomenon from different frames of reference. Brodbeck1 suggests that the relationship between the two types of explanation is not one of antagonism but one of derivation: The patterns of observable individual behavior are the referents of the group concepts. The latter are therefore definable in terms of behavior of individuals, including, of course, their relations to each other. These definitions alone, however, do not permit the explanation of group behavior by means of the behavior of individuals. Or, to say the same thing differently, definition alone is not sufficient for the reduction of sociology to psychology.d Brodbeck contends that psychological concepts are not inter- changeable with sociological concepts. The basic terms referring to complexes and those referring to members are not interchangeable. The composition laws of psychology are about how the "elements" or members interact with each other. The sociological laws are about the resultant be- havior of groups. These two kinds of laws need not and in general will not have the same form.) 1 \' I a May Brodbeck, "models, Leanings, and Theories," gymposium on Sociological Theory (Evanston, Illinois: Row, Peterson, 1959), pp. 373-QO3. 2 Ibiu.,_p. 397. 3 Ibid. 13. Brodbeck's analysis of the relationship between psycholgical and sociological results implies that a phenomena may be explained on different planes and that the partialing out of the variance at one plane would entail the loosing of variance at the other plane. That is to say, 'f lack of education on the sociological level and feelings of inadequacy on the psychological level are causes of marital discord when the sociological and psychological variables are correlated, the sociological variable cannot be controlled without reducing the effects of the psychological variable and visa versa. The extent of education must be left free to vary in order not to constrict the psychological variable. On the other hand, the intrusion of unpredicted psychological variables in a psychological study are unwelcomed and must be anticipated and controlled. The background factors which will be discussed on the following pages are those factors which are related to marital adjustment on a different plane from that which is being studied in the present thesis. In accordance with the perSpective suggested by Brodbeck, they will be described and analyzed as concomitants to the proposed causes rather than variables which might contradict the integrity of the predicted causes of marital adjustment or maladjustment. By disregarding the evidence secured from a different frame of reference as confirmatory or disconfirmatory, the researcher accepts the following ethical obligations: 1. To illustrate the pervasiveness of the predicted variables among groups described by different background factors. 2. To interpret empirical disconfirmation as a result of faulty prediction, faulty instrumentation or faulty 11+. theory rather than.interference from uncontrolled back- ground factors which are not controlled because they are purported to be concomitant with the predicted results. Economic Factors Common sense suggests that when the breadwinner does not bring home the bread, marital problems might ensue. In 1938 Terman“ stated that although low income is no doubt a factor is domestic discord, the incidence of marital difficulties associated.with this factor is offset by the incidence of marital difficulties among couples having surplus money. Since more than four-fifths of Terman's couples were high school graduates and more than a third were college graduates, it is understandable that the economic factor did.not prove to be diagnostic of maladjustment among his sample. Level of income is more strongly associated with marital adjustment in Burgess and Cottrell's study5 but only in interaction'with educational status, occupational status and mobility. Burgess and Cottrell's sample is also skewed in the direction of the middle and upper socioeconomic classes. Judging from the above early studies, a foregone conclusion may be expressed: economic factors are related to marriage among those couples where income is scarce. Goode6 studied the financial status of recentLy divorced couples in Detroit, Michigan. His data affirmed a rough inverse relationship between economic factors and divorce. His statistics, which are 4 L. M. Teman. Psychological Factors g mud Happiness (New York: McGraw Hill, 1938 , p. 170. 5 E. W. Burgess and L. S. Cottrell, Jr. Predictin Success or nglure in Marriage (New Yerk: Prentice Hall, 19395, p. 157. 6‘w. J. Goode, “Economic Factors and marital Stability," American §gciological Review, XVI (1951), pp. 298-301. 15. descriptive but not inferential, suggested that the effect of economic factors in divorce interact with other factors which would suggest that their association with divorce might be expected to be lower in a direct cause-effect relationship. Williamson? hypothesized that the following three economic factors are positively associated with marital adjustment. 1. Social Status as indicated by location. type of residence, level of education and occupation. 2. Economic Security as indicated by indices of savings, regularity of employment and freedom from debts. 3. Effective Economic Imemmt as indicated by budgeting and efficient management of the home. The author randomly selected 210 couples from the white pepulation of Los Angeles, California, interviewed them separately, and gave them a marital adjustment scale adapted from scales deve10ped by Burgess and Cottrell and by Locke. Those couples earning mediocre scores on the marital adjustment scale were eliminated from the analysis of the data. The results were based on 85 men and 86 women scoring high and 66 men and 62 women scoring low on the scale. Husbands having occupations in the skilled labor, business or professional classes at the time of marriage were significantly happier than husbands in an elemental or an unskilled class of work at the time 01' marriage. The pattern was similar for the wives' former occupations. A Significantly larger percentage of the happy couples were employed as White collar or managerial workers. Couples earning less than $436 _ 7 R. c. Williamson, "Economic Factors in Marital Adjustment,” Mania 6 and F . XIV (1952). pp. 298-301. ‘ —r .. 16. per month prevailed among the unhappy couples adihough these couples did not significantly qualify as unhappy. The associations between residential area and marital.happiness were differential. Males and females living in "Low" residential areas ‘were significantly unhappy in marriage. .Males and females living in "High” residential areas were significantly happy. However. males living in "Middle" residential areas were happy while females living in this same area were unhappy in marriage. Although the percentage was not significant. there was a tendency for couples who rented.their homes to be unhappy in their marriages. A combined economic security rating discriminated significantly between happy and unhappy couples. The sub-categories of the security rating afforded data.which had not been collected in the United States befbre Williamson's study. Couples with at least $600 savings were significantly happier than couples with less than $600 savings. Husbands owing less than.$300 were significantly happier'than.husbands owing more than $300. Although insurance per se was not as predictive as other assets, wives protected by $5,000 were more often happy than those protected by less. Unemployment was significantly related to marital.happiness in a negative direction. Williamson's study was a.thorough investigation of the economic factors in marriage. However, some caution in the interpretation of his results is warranted. The author did.not state the level of Statistical significance which he used to reject the (unstated) null hypotheses, nor did he state statistical and sampling procedures used. AsSuming that the above factors could be accounted for, it may be interpreted.fromWilliamson's results that.the economic factor associ- 17. ated with marital adjustment is in fact a multitude of factors each worthy of a detailed investigation. The ress of Cultur Back ound The impress of cultural background refers to all the cultural or social influences upon a person. Burgess and Cottrell8 first applied the above terminology in order to classify a segment of their results. They found that likeness of cultural impress beWeen the mates was associated with marital happiness and marked differences in cultural impress with unhappiness in marriage. It was also discovered that the overall cultural level of the husbands' parents was more; important than the overall cultural level of the wives' parents for success in marriage. Certain differences betwaen the husband and wife popularly assumed to be inimical to a successful marriage. such as differences in religious affiliations and educational status. showed no relation to marital adjustment during the first six years of marriage. Residence in the country during childhood and adolescence was favorable, but residence in the city during the same periods was unfavorable to marital. adjust- ment. The indices of cultural impress were many. Maturity, educational opportunities and achievements, participation in religious activities. number of friends. membership in organizations. and residence in neighborhoods of single dwelling units all play some part in providing an atmosphere conducive to adjustment in marriage. Burgess and Cottrell concluded. 8 Burgess and Cottrell. Jr., 99, 2%.. p. 157. 1b. This socialized person is perhaps characterized by traits of stability, conventionality and conformity. He has been molded by and has participated in our social institutions. He is, therefore, well fitted by training and experience to make the adjustments required in marriage.y Granting the association between cultural impress and marital adjustment, the question may be raised if levels of cultural impress ‘0 o 0 ' m 4.0 ' ‘ are predictive of marital adjustment. Terman pOinted out that occupational level was not predictive of marital adjustment, although this finding was contra-indicated in later studies. he significant relationships were found between occupational classification and marital adjustment for 1,584 subjects. His findings did indicate that similar mental abilities of mates were related to marital happiness. When the husband's mental ability was significantly inferior, the wife was unhappy. When the wife's mental ability was significantly inferior, the husband was unhappy. “11 locke's 1951 study of a representative sample of divorced and happily married nevi-plea 337::33323; '7' ‘i w . 4.. -.'.~q-l-»4.A.r’ q-s"r. _.-"‘v ‘v‘v I. :(Jl. ’31. 3" ‘r I-‘._‘= Ic- -ng .uummmmmnmu 0° Hacksaw ..0aoomdn Hagan“: who? a“ 0000000. .H0 .00 .. .m. m .0osn< .330 533.338 3.5: .3503.» .Nnémm .00 . . 0H0: o g ..owaflfiz 5 figfla 0.833 05 3.0.30 53855. cm 60820 RS 05 0032005. 4.0 .3980 $.08 .00. m3— HHS .mm fifl m 0 3.0005033 60.300 0535009 30.5 .Bamuozg 33 mo 300004 H038m 0:0 330 053.33... no 330 3598a 3308 so 08582 0035030000 3400980 .3303 30 0.3023328 amm— Huuzonm no 3005-... .m 500.0605. 0.0% .00 63.3.3... . an? Ex .3 .305 00.833. $3302 0 . .2303...“ 320: 05 000.3300 .00Hauoo 60.3500 RN .5005.” $9 03030330 330?. . him £0.73” ._ .ll» :05. .3... a .345: 00 30320 602508 6.0 0309 2 7 .Bqflnsam 830 .noduaou I30 .323 .333 33335: 2 2:. 55% 3388a 8.33 Bananfidwa mm :38 93 :8 o: .593 mm? if . mm? a .m o .. 5.83; an. 53p cam .325 "389 cofloflfim Ea paofiuag 35: £28.. ..z v. .8313 3mm 6. m 6x03 m .dwl .Mmmmww don. .. .pnmfipndnflc. H33 93 vaoaoonwnS oaom Havana: no m33o§oo . m .3309 qofiHwhm .3980 .0333 5 30.302an Hmflom «Sam: 32m: asuunomfim no 2.3: 8am:8 mm 3238» as» an? .3; 5:33 Ba .36 {Rom .uo oo 32». 33° «3:3 .033 £83 .523: no 30.330 u #322 and: 833 383m «30 eggs .8392 «3 g ...»:3 3% 2: saw hafigfiaq .523 mm 339.8 5 393m “Beam 4 585234 3g. 933 gags? .30: 393» 33° 3:3 3 338 m R2 .oé .8333 c5 4.x .833 £35” . 3.1 33m macaw . muo . v.38 8 .333 n . .. 2 a. .3 «topic 3.5.0. :25 25.9 .82380 .m.~ £3. CHAP TIER III an"? TQTr‘f A." -Y"".‘ C‘"-"""' .mi. sol-JD. \.".‘ -3; chi-a}: C The De51gn The present study is designed to accomodate the variables known to be relevant in scale development and marital adjustment research. The assumptions underlying the independent variables are stated followed by a discussion of the derivation of the dependent variables. Identification of and methods of ceping with extraneous variance are presented followed by an overview of the sample and a diagrammatic design of the study. The Independent Variables Adjustment and maladjustment in marriage are the two independent 'variables of the current study. It was assumed that adjustment in Inarriage varies along a continuim from those few couples who approach “100 per cent adjustment to those few couples who are almost completely nualadjusted. It was also assumed that couples starting procedures for Clivorce as well as couples receiving marital counseling would represent time lower end of the continuim.while couples earning high scores on a reliable and valid marital adjustment test would represent the higher end <31? the continuim. If the two criteria represent the extremes of marital aCljustment. then a comparison of the two groups on a number of items would reveal those items which are neutral with repect to marital adjustment and those items which discriminate beWeen adjusted and maladjusted couples. 121S1_Lependent Variables The dependent variables of the experiment are the consensus of mat—63' perceptions on the truth or falsity of issues, the consensus of 73. 71+. mates on the perceived importance of issues, the configuration of the above variables and the content of the experimental scale--the Issues Scale. Consensus of Mates on Issues The consensus of mates on the truth or falsity of issues derives its status as a variable related to marriage from Terman's and from Burgess and Cottrell's nonnative studies and every major attempt to scale develOpment since the early studies. In each of these studies a high weighting was given to between six to fifteen areas in which consensus of mates was found significantly related to marital adjustment. The typical consensus or agreement item found on previous scales was constructed by, having a member of a couple rate his agreement. with his mate along a continuim from "always agree” to "alwws disagree” in an area, such as handling finances. me weights of the various ratings for marital adjustment were then determined by their reapective correlations With marital happiness. Irrevocably, the greater the agreement of rates on eight or nine areas, the greater was the marital happiness. In 1957 Locke and Williamson established the factorial valicfi.ty of consensus or agreement for marital adjustment. It was inferred from Kelly's corollaries numbered 10 and 11 (supra, pp. 5-6) that if gross areas such as "Intimate Relations" and "Philomphy of Life" are significantly related to marital adjustment, than items describing the Specifics of the gross areas would also be Significantly related to marital adjustment as well as pin pointing agreements and disagreements between mates. For example, in the area °f "Intimate Relations" the item. “Having sexual intercourse is some- tinIes annoying". because of its derivation from an area known to be 75- related to marital adjustment would be eicpected to be related to marital adjustment more so than would be an item~chosen at random. It is therefore ezqaected that if twenty items are chosen to represent an area of known merit. a portion of these titenty will discriminate between maritally adjusted and maladjusted couples. One of the unique aspects of the stuchr is the systematic exploration of areas known to be grossly related to marital adjustment. Consensus of Mates on the Perceived Iggortafigce of Issues 'lhe consensus of mates on the perceived importance of issues derives its status as a variable related to marriage from the personal construct theory of behavior--human behavior is a function of the manner in which individuals perceive events and anticipate consequences. The personal construct theory is applied to marriage by classifying the degree of marital adjustment as the anticipated consequences of the perceptions mates have towards various issues related to marriage. Thus, the behavioral consequence, marital adjustment or maladjustment is a function of the mates' perceptions on issues correlated with marital happiness. Perception in the study is Operationally defined as the strength , of feeling or discerned importance of opinions on issues. Therefore, considering the Operational definition of perception and the consensus theory described earlier, the personal construct theory applied to marriage may be reformulated: Marital adjustment is a function of the agreement of mates in their opinions and strength of feelings towards valr'ious issues correlated with marriage. 76. The Configural Scoring System Another unique aSpect of the study is that it is the first scale attempting to configurally score reSponses for marital adjustment in terms of the mates' consensus of the perceived importance as well as their consensus on the truth or falsity of the items. For example, Spousal agreement on the truth or falsity and importance of an issue W be more related to marital adjustment than Spousal agreement on the importance of an issue but disagreement on the truth of an issue. Ihe relationship suggested by theory is that spousal agreement on the truth or falsity of issues which both mates perceive as personally important are associated with adjustment in marriage. However, previous scale development in-marital adjustment. research has measured only spousal agreement without giving credence to the importance of the items to the couples. An empirical question which is studied in the current thesis is whether or not the perceptual dimension enhances item discrimination, thus supporting the applicability of the personal construct theory in marital scale research. The configural scoring system of the experimental scale is construc- ted to afford six types of Spousal reaponses which are illustrated in Figure 1. The first letter of the configural score always refers to the mates' reSponse on the truth or falsity of the item, the second letter refers to the mates' reSponse on their feelings about the items, and the third letter (when present) refers to the direction of their agreed upon feelings. Agreement on the truth or falsity of an item, re{egardless of whether the mates agree on their strength of feelings concerning the item is labeled an agreement (A). When mates agree on both the truth or falsity of an‘ item and feel strongly about the item, they are given a configural score of (It-AS). 77. Examples of the (A-AS) configuration are found in the first two rows of Column I in Figure 1. Figure 1. The Configural Scoring System for the Experimental Scale Strength of Feeling Concerning Item (Strong vs. Weak) Truth vs. Falsity of Items Both Mates Feel StronglL Both Mates Do Disagreement Net FseQ.§££22elz_i_ga_§2rensih t c . Agree-Msagree Both mates agree Agree-Agree Strong Agree-Agree Week on truth ALA-AS) (a.Aw1 ((A-D) Both mates agree Agree-Agree Strong Agree-Agree Weak Agree-msagree on falsity (It-AS) 41km ) (A—D) Disagreement on Disagree—Agree Strong Meagree—Agree Weak Disagree-Disagree; truth or falsm LD—ASJ (D—AW) (0-D) According to theory, marital adjustmentis a function of the agree.- ment of mates in their Opinions when they feel strongly towards various issues correlated with marriage. Spousal reSponse type A-AS meets the theoretical conditions and should therefore be indicative of adjustment in marriage. Spousal reSponse types LAW, A-D, D—Aw, D—AS and D-D fail to meet one or more of the theoretical conditions for adjustment. It is flxerefore expected that response type A-AS sould be chosen more often . bymaritally adjusted than maladjusted couples while the other reSponse types should be chosen more often by maritally maladjusted than adjusted couples. 4% ing with Extraneous Variables No sources of extraneous variance were identified in the study. The first source of extraneous variance is the contribution of socio- logical characteristics in marriage. The relative importance of these characteristics and their status as variables in the current study are 78. discussed in Chapter II under the subheading ”Background Factors in Marriage". They were treated as correlated variables in marital adjustment. No effort was made to control them because of their concomitant status in the study. The second source of extraneous variance is inextricably bound up in the reSponding to the scale. It is the error variance brought about by the susceptibility of all structured personality or attitude tests to "faking", "lying", or unconscious self deception. In addition, a further source of error variance is contributed by subjects who make errors in recording, having reading comprehension difficulties, lack understanding of what is eXpected, are very confused in their thought processes, and those who try to make themselves look worse than they are for various reasons. Meehl has written extensively about the last source of error which he labeled "faking bad” as Opposed to "faking good". A clear exanple of "faking bad" is that of draftees failing psychiatric exam- inations in order to be exempted from military service. 'lhweatt1 has extensively described the history and rationale of validity keys, their various merits and implications. For the current study two validity scales are being constructed. The decision of which type of validity scales should be constructed was based upon what are assumed to be the most serious types of errors peculiar to testing for marital adjustment. The validity of the Set ‘1‘ and F scales in the context of marital adjustment testing is assumed a priorally. One of these two scales was constructed in an effort to identify ‘ R. c. Thweatt, ”The Development and Validation of an F Scale for an Objective Test Battery on Motivation" (Unpublished Ph.D. dissertation, Michigan State University, 1961),, pp. 20-57. 79. rigid response.bias, i.e. the type of reSponding which falls into rigid monotony because of resistance, boredom, inflexibility, unconcern and other mlknown variables. This type of scale has been develOped by Frickez and.has been labeled the "Set T Scale". It is composed of items eliciting almost random reSponses because of their controversial nature. It is assumed that if a subject answers a significant number of contro- versial items in a singular direction that his peculiarity of reSponding is a source of error variance and this subject should be deleted from the sanple. The second validitar scale chosen for the current study is the "F lype Scale" which was originally developed for the MP1. It is a scale composed of items which are responded to with low frequency by predetermined samples. If a subject responds to a significant number of "F Scale" items in the statistically rare manner, it is assumed that he is either purposefully or unconsciously misunderstanding the task and contributing a source of error variance to the experimentaldesign. Therefore, this subject should be deleted from the sample. Within the lindts ofthe above assumptions, both validity scales were deve10ped for the ''Issues Scale". A further assumption was made regarding the cut-off point for the validity scales. The standard cut- Off point for validity scales is a '1‘ score of 70. It was assumed in the current study that the rejection of subjects from the sanple when they are not a contributory source of extraneous variance was a more serious error than retaining subjects whose reaponses were actually highly 2 B. Fricke, "ReSponse Bias Scale for the Minnesota Multiphasic Personality Inventory," Journal oi‘rgounselingPsmhology, IV (1957), pp- 149-53- 80. contributory to error variance. Three criteria were established for setting the cut-off points for the validity scales: (1) the cut-off point should be more stringent than the usual 1‘ score of 70, (2) the actual T score should be established at the point where there is a marked schism in the curve of the ‘1‘ scores, and (3) in the case where there is no schism in the curve of T scores the cut-off point should be established at the T score commensurate with a p<.01. Considering the lack of empirical validity for validity scales in marital adjustment scales, the above criteria would insure conservative interpretation of validity scale scores. The Sflle The sample consisted of a maritally adjusted group and maritally maladjusted group. The maladjusted group consisted of couples in which one or both mates were filing for divorce or couples who were obtaining professional marital counseling. The maritally adjusted couples were obtained primarily through the c00peration of various clergyman. In order to qualify as maritally adjusted, the couples had to earn a predetermined score or better on the criterion instrument (Slpra, p. 67). One-half of the maritally maladjusted and adjusted couples were assigned to the validation group and the remaining halves were assigned to the cross validation group according to a table of random numbers. In Chapter IV the selection and characteristics of the subjects are described in greater detail. I ' Diagrammzatic Plan of the Design The design .of the study is divided into eight steps which are outlined in Table 3.1. These steps refer to the major operations of the stucbr in their order of occurrence. In the table the hypotheses 81. .33» adorn? 05 5 EH95.“ mason» .8 92m 23 335.82. x... hugfiuspoum 53.“ wflfimpfldoo 5.805. vfiaoohmw * N . pawgm 05 55 noflmdscom .993: 933 m 3” speed”. 35.328 announce 05 .852? mo ago» HHH> Wm 53 co U33. «Has N M e5 Eugen... h3g3 apps conga» Q5523 398 on... 8053 39503454..» upswvaououfiu _ pnwgm Handmflzoo 983 @3338. 05 350:3 .«o 53:55.8. t . HH> pm ,. x x x N .m 8 nofiwuab M £8522 .2858»... , 380 N ha v5 533.3? N .3 33am mfifloon N pfiafinm aflemflaoo w pfifihm €33,203 "Ban 2»? 3:38 "Egg mo zofisamm N m5 manning 398 3 .830 5. 0.35% so: a £9 3 "$3 3 cheeses 05» .zoESHd; 880 fillllala .nofiwfimc M . no E85 pzoaoonmm wannabe 5033558 dons» . tango 3 mfiunooom o>3o€oha 0.32 m.“ 5ng ,, unnamed Emu.“ heaven: mo dofimgopou "20543; M _ .3 n 05 £95 38. 9.9m 058 we 532.23.? _ 05 B. mama.” gum. Ewan—”12> .mo Ragga .38 23 no 9.8 dad 2e. fi 83 3 3 £58» a 3 848 no zoflgmHg .95 e83. 33% even? mail 96.8 cosmefims $80 season 63895 98 33a .zobamszaa 982 H $6.5 dogged; _ ass .5 .s 52 engages in use ° 82. which are being put to test during the reSpective steps in the design as well as which groups are involved at each step are charted. Step I refers to those Operations conducted to prepare the items on an a priori basis so that they have the greatest chance of discrimina- ting between maritally adjusted. and maladjusted cowles. Step II is the administration of the ewerimental scale to all of the groups which will provide the data for the analysis. Step III is the development of the Validity Scales Set T and F described under "Control of Extraneous Variables“ in the present chapter (supra, p. 77). Steps IV and v are the comparisons of the maritally adjusted and maladjusted groups in order to determine the validity of the items and of the scale. In Step IV the scores of the first halves of the maritally adjusted group and maladjusted grow are compared on the ewerimental scale according to two different theories. In Step V the scores of the second halves of the maritally adjusted group and maladjusted group are compared according to the same two theories in order to cross validate the items. ' ihe sixth step is the determnation of reliability. It will be determined separately for each of the four grows in the sample twice: 1. Validation grow: Maritally adjustedp-configural scoring 2. Validation grow: Maritally adjustedp-straight agreement scoring 3. Validation grow: Maritally maladjusted-monfigural scoring ’4. Validation grow: Maritally maladjusted—straight agreement scoring 5. Cross validation grow: Maritale adjusted-..configural scoring 6. Cross validation group: Maritally adjustedm-straight agreement scoring - 83. 7. Cross validation grow: Maritally maladjus ted-..configural scoring 8. Cross validation grow: Mentally maladjusted-.8traight agreement scoring. In Step VII the cross validation grow maritally adjusted and maladjusted couples' scores on the validated items are compared in order to determine if the validated scale differentiates between a new sample of maritally adjusted and maladjusted cowles. Step VIII involves judgments regarding the configural scoring system. the straight agreement scoring system and the content of the scale. Judgments regarding the scoring systems are based on item productivity, validity and reliability. Judgments regarding the content of the scale are based upon the sufficiency of item productivity insuring systematic rather than chance discrimination in the Spousal agreement correlates of marital adjustment. The Hypotheses Three null hypotheses are formulated with their respective alter- nates. Null Hypothesis I is stated in reference to the scoring system inferred from the personal construct formulation of behavior. Null Hypothesis II is stated in reference to items scored for Spousal agreement on issues related to marriage regardless of the mates' agussement or disagreement on fire importance of the items. Null Hypothesis III is stated in reference to the content of the scale. {Jill Hypp thesis IuConfigural Sco_ri_gg ‘-' ear tally adj usted couples agree no more than mar}. tally maladw‘estei couples on the truth or falsity Ca. iss use perceived important by both mates. [:1 terrate I. mmtaMy adjusted couples agree more than riritaiily maladjusted cowles on the truth or falsity of res res perceived important by both mates. Null '{ypofhes SIL-Straight Agreemmt Scoring tunnels haigxteted couples agrte no more than maritly r‘-_a"*i;;-d couples on the truth or. fair. deity of ion" sues regzria less of the gerc ceived importance of the issues. .ilt arrows“. II: Mentally adjusted. couples ever» n: no than rauwrij Milk: oed couples on the truth or faint} of issues raga :dless of the pe err .eived impo- .‘cince of issues. Null Hypothesis III-Cor.tent of the Seal-L e Items derived from the spousal agreement correlates do not discriminate between maritally adjusted and maladjus ted cowles. Alternste "’7' Items derived from the Spousal agreement fixaxb O correlates discriminate between maritally adjusted and maladjusted cowles. lhe Statistical Analysis 'lhe basic statistic used in the eooperiment is the this-aquare.3 It is used to test the significance of the differences between maritally adjusted and maladjusted samples in the validation and cross validation grows for items scored in accordance with the personal construct formulation and the straight agreement approach. The six reSponse types of the Issues Scale are regrowed according to the two reSpective theories into contingency tables. The chi-square tests are conducted for each of the 1&0 items using both the configural and straight agreement scoring. All analyses were cross validated. It was decided that a p<.10 was necessary to consider items sigiificant at the validation stage and cross validatim a ge. lhe leniency assigned to the significance level required to reject the null hypotheses is Justified by two considerations: 1953). .3 1. lhe purpose of the study is to determine the relative merits of tin two scoring systems derived from two different theories. Every oppor‘bmity should be afforded H. M. Walker and J. Lev, Statistical Inference (New York: Holt, pp. 84-108. 85. for determining the differences in these scoring systems. 2. The replication in the design has the effect of insuring thattpredicted differences are reliable. Items which are significant for both the validation and cross validation groups but contain a cell or cells with less than a theoretical expectancy of five are reanalyzed using the Fisher's Exact Probability Testf‘ me t tests is used to determine whether or not the validated items differentiate significantly between the mentally adjusted and maladjusted subjects of the cross validation group. In addition to the chi-square statistic the Analysis of Variance Method of Determining Reliability6 is used. '1‘ scores and their reapective probabilities? are used to determine to what point scores on the validity keys are random. ‘ me statistical treatment of the validity keys warrant further comment. Items which are responded to in accordance with the character- is tics of the "Set '1‘" and "F" scales are separated into the respective validity scales. 1‘ scores are attributed to subjects in accordance with the number of validity key items they respond to in the statistically rare manner. Cut-off points are established for the scales according to the criteria (supra. p. 80) and subjects who score above the cut-off “I d. ( 5 A. L. Edwards, Statistical Methods for the Behavioral Sciences New York: Rinehart, 19 . pp. 111-3. 6 V C. J. Hoyt, . ”Test Reliability Estimated by the Analysis of Elmsince Method," Psychometrika VI (1941), pp. 267-87. 7 A. L. seems. 23. 213.. pp. 216.77. 86. point on either the "Set '1‘“ or "F" scales are discarded from the sample. The '1‘ score necessary for elimination from the sample was purposefully made high in order to avoid a lype I statistical error due to the relatively small N. Sunnnary The independent variables of the study are marital adjustment and marital maladjustment. The dependent variables are the consensus of mates' perceptions on the truth or falsity of issues, the consensus of mates in their strength of feelings about issues and the interaction of the above variables. 'mo types of extraneous variables were discussed. ’lhe sociological factors were considered concomitant variables and should therefore vary with the degree of adjustment and maladjustment in marriage. Secondly. the psychological factors such as response bias were discussed and two validity scales are being constructed to rule out this source of extraneous variables in cases vhere it is flagrantly manifested. The design of the study was rqaorted diagrammatically and in terms 0f temporal occurence. From first to last. the following eight steps were charted: Step I Pro-adnmxistration of scale Step II Administration of experimental scale to main body of subjects Step III Development of two validity scales Step IV Validation of items Step V Cross validation of items Step VI Estimation of reliability Step VII Validation of the experimental scale Step VIII Judgments concerning two scoring systems. 87. “flares null hypotheses and their alternates were stated. lhe first referred to the scoring system derived from the personal construct theory. Ihe second null hypothesis referred to the scoring system derived from the straight agreement theory. The third referred to the content of the scale. The statistical treatment of the data was discussed in terms of the various tests to be made during the operational steps of the design. Levels of significance required to reject the null hypotheses were stated in advance of the analysis of the results. A p 4 .10 was required at both the validation and cross validation stage of the study. CHAPTER IV MEIEODS AND PROCEDURES In'Chapter III the design for the stuchr was detailed. The purpose of the present chapter is to describe the nature of the variables. First, the criterion and experimental instruments are described followed by an explanation of subject selection. be age, number of years married. vocations, education and sundry characteristics of the subjects are given followed by a description of instrmnent administration. Instrumentation mo instruments are discussed in the following section. ’lhe first is the criterion instrument which was employed in the selection of maritally adjusted subjects. The second instrument is the experimental scale--the Issues Scale. The development and structure of the Issues Scale are discussed. Criterion Instrument The Short Marital Adjustment. Test (supra. p. 67) was the criterion instrument used to select maritally adjusted couples to participate in the current study. Marriages in which both mates earned a score of 100 or over on the criterion instrument were retained to comprise the mentally adjusted samples. Marriages in which one or both mates earned 3001588 of less than 100 on the criterion instrument were not considered maritally adjusted and were deleted from the stucv. The normative data on the Short Marital Adjustment Test suggested that 96 per cent of the maritally adjusted couples and 17 per cent of the 88. 89. maritally maladjusted couples score on or above the cut-off score of 100. line risk of including 17 per cent maladjusted couples in the adjusted samples was justified according to two consideration. 1. All previous attempts to scale marital adjustment indicated that the upper end of the marital adjustment curve falls off sharply, suggesting a modal range of adjustment followed by relatively few idealized adjustments. 2. A number of previous studies have found that maritally maladjusted couples who reSpond to scales similar to maritally adjusted couples are maladjusted only in particular aSpects of their marital life. he mean and standard deviations of the criterion instrument scores for maritally adjusted couples in the validation and cross validation groups were computed. Respectively, the mean scores for men were 124 and 127. and the mean scores for women were 122 and 127. The standard deviation for men was 13A in the validation group and 11.8 in the cross validation group. and 12.7 and 11.3 for the women in these respective groups. in F ratio was computed for the largest and smallest variances of the four groups. The ratio was not large enough to suggest systematic variation in the mean scores of marital adjustment between either men and women or between validation and cross validation groups. ...Ihe Qomm mstme t ‘Ihe "Issues Scale" is. composed of 1140 items subdivided into seven groups of 20 items. 'me seven subgroups of items were derived from the following seven spousal agreement correlates of marital adjustment: (1) handling finances. (2) demonstrations of affection, (3) intimate muons, (1+) friends, (5) dealing With in-laws, (6) philosophy of life, and (7) moreation. 90. In past studies agreement of mates on the above areas was correlated with marital adjustment. In Appendix A a list of the original pool of Issue Scale items is classified according to the above areas of marriage. Subjects were required to make bro responses to each item. First they were asked to designate whether they believed the item to be always true, usually true. rarely true. or never true. Secondly they were asked to designate whether or not they felt strongly about the item. In Appendix A the actual instruction sheet for responding to the scale is duplicated. 'lhe "Issues Scale” was administered to a sample of 15 couples from various socioeconomic groups in order to detemine any lack of clarity and anomalies in the scale. 'Ihe investigator read the items aloud to two of the pilot couples in order to observe their reactions to the statemelts. Several pecple listened to the items read in groups of taro, i.e. items numbered 1 and 2. 2 and 3, 3 and 14'. etc. in order to determine whether the ordering of the pairs provided a source of humor or confusion which might decrease the value of the items involved. Beoccurring ~ comments from the pilot sample concerning item clarity. the effects of oz-dering. and the format of the scale were carefully evaluated and considered in the revision of the scale. ‘ The next step in the preparation of the Issues Scale for the main both? of subjects who participated in the study was to establish the aEe-grede level of every word used in the items and the instructions. 10120 and ’Ihorndiltze1 suggest that words occurring in popular print at 1 E. L. morndike and I. Large, he Teacher‘s Word Book of 20.000 Words (New York: Bureau of Publications. Teachers College. Columbia K “malty. 19M). 91. the rate of from 10 to 100 per million are comprehendable to youngsters in the fifth or sixth grade or to people with a fifth or sixth grade education. Each of the words in the Issues Scale was checked against the list of words found in popular print at a rate of from 10 to 100 per million. An effort was made to replace any words in the scale occurring less frequently in popular print than 10 per million with words meeting this standard. With the exception of the words "sexual inte'rcourse'l which could not be replaced with suitable alternates. all of the words in the Issues Scale met the above criterion. Selection of Subjects In selecting subjects for a studv of marital adjustment the follow. ing biasing factors must be considered: 1. Maritally maladjusted couples are usually unwilling to participate in research which delves into aSpects of their life possibly illustrating their inadequacies and shortcomings. Attewts to attain mates from maladjusted marriages on a voluntary basis are often futile. I 2. Marital research necessarily involves an investigation of private or personal characteristics of the mates. 3. The average socioeconomic characteristics of maritally aijusted couples are known to be different from maritally maladjusted cowles. ‘Jherefore. attempts‘to match maritally adjusted and maladjusted cO‘Ielplles on sociological variables becomes a search for statistically rare conPlea in either the maritally adjusted or maladjusted populations. 1+. Both mates must volunteer or be persuaded to participate in a study of marriage. 92. In the current study the subjects were persuaded to participate by the investigator as well as a number of clergyman, social workers, and educators who helped obtain couples for the study. No restrictions were placed upon socioeconomic characteristics of the subjects with the exception that both the mates were required to be no older than 55 years. No claim was made about the subjects being a random sample of a specified population. Instead, the groups were regarded as ”chunks" from which hypothetical populations my be defined. The Maritigz Malag‘usted Goggles ‘me criteria of maladjustment in marriage were either that one or both mates were starting procedures for a divorce or one or both mates were diagnosed as having marital problems and were being counseled for the same by a professional social service agency. 'Diirty-nine couples, in which one or both mates were taking the preliminary legal steps leading to divorce were obtained from the Probation and Adjustment Divisions of the Detroit, Michigan Recorders Court. 'lhe Probation Division works with clientele who have broken one or more laws governing marriage such as non-support and incest. 'Die clients are placed on a period ofprobation during which they are helped to readjust and make restitution for their behavior. ‘Ihe Adjustment ItiVision is the initial office which hears complaints of one mate against the other regarding behavior which may be grounds for divorce. As part of the routine office procedure, each new couple interviewed within a week at the Adjustment Division and within three months at the ProElation Division was administered the Issues Scale. The administration °f the scale was conducted by the current investigator for two-thirds of the maritally maladjusted couples. Test administration for the remaining 93. one-third of the cowles was conducted by three caseworkers from the Probations Envision who were trained in administering the test. Thirty- nine cases were obtained from the Detroit Recorders Court. One of the thirty-nine couples was not included in the analysis of the data because their T score on the "1“ Scale” extended beyond the established cut-off point. Eight additional maritally maladjusted cases were obtained from the Catholic Social Serive of Lansing, Michigan. has caseworkers at this agency were instructed to administer the Issues Scale to clients where the diagnosis was clearly that of maladjustment in marriage. The Marital}; Adjusted Cogles The maritally adjusted couples were obtained primarily through the assistance of various clergymen. Two couples were persuaded to partici- pate directly through the efforts of the investigator and one couple was obtained through the efforts of an educator. An effort was made to enlist the cooperation of ministers from varying Protestant sects. mo of the couples were Catholic, and three were Negro Methodists. The various clergymen were instructed to select couples whom they~ believed to be maritally adjusted. Fifty-eight couples were obtained in all. Forty-seven of these couples qualified as maritally adjuswd by Scoring above the cut-off point on the Short Marital Adjustment Test. niece forty-seven couples comprise the maritally adjusted samples and have all been retained for the analysis of the data. 92+. 'lhe Validation and Cross vgdation Grogps Half of the maritally adjusted and half of the maritally maladjusted couples were assigned to the validation group. and the remaining halves were assigned to the cross validation group. The assignment of the couples to validation or cross validation was conducted according to a table of random numbers. The concomitant variables of marital adjustment: education, level of occupation, age of mates and years married were tabulated in order to describe the characteristics of the sanples. I Number of years of education is reported in Tables 11.1 and 11.2. Maritally adjusted men and women in both the validation and cross valida- tion groups had a higher number of mean years education than maritally maladjusted men and women. However. a computed F ratio for the maximal and minimal variances illustrated that differences among the eight groups were not significant. In Table 4.1 is reported the number of years of education for maritally adjusted and maladjusted men and women in the validation and cross validation groups. Table 11.1. Number of Years Education for Maritally Adjusted and Malad- justed Men and Womm in the Validation and Cross Validation Groups Number Validation Grog Cross Validation Gro of Adjusted Maladjusted Adjusted Maladjust-ed Years )1 F M F. M F M F 2-5 0 o 1 1 o o 2 o 6-9 . 2 3 5 5 0 1 7 3 10-13 111 15 12 12 17 18 13 19 No.17 5 5 3 2 6 1+ 1 1 48-21 2 o o o 1 1 o o M 23 3; 2i 20 211 211 23 23 Jie‘an 13 11.6 10.8 10.11 13.1125 10.: 11.2 3D” 2.- .2 2. 2.0 20 2.8 1.8 .- "a .... h ._ .. - .. _ ___ . - ».— __ "' Part of the sample having 13 years education spent an additional year in secondary school while part obtained one year of a college or trade school education. See Table 1+.2 for the frequency of subjects who attel’lded trade school or college. 95. It was noteworthy that maritally adjusted subjects attended college more frequently while maritally maladjusted subjects attended trade school more frequently. A chi-square was computed for these two groups, and the Mfermce in education is significant at a p<.01. there were no sigzificant differmces between men and women or between the validation and cross validation groups in the type of post high school education . obtained. In Table 11.2 the frequency of maritally adjusted and maladjus- ted men and women in the validation and cross validation groups who attended trade school or college is reported. Table 4.2. Frequency of Maritally Adjusted and Maladjusted Men and Woman in the Validation and Cross Validation Grows Who Have Attended Trade School or College Validation Group Cross Validation Grow Adjusted Maladjusted Adjusted Maladjusted 11 F M F M F 11 r _ Attended College 7 8 3 2 1O 9 3 3 Attended ‘I'I‘ade School 2 1 5 6 2 3 6 5 _fI_‘o_tal ' 9 9 8 8 12 12 g 9 s The occupational level of the participating cowles is recorded, and Classified according to the Dictionary of Occwational Titles Part IV2 in Table 11.3. The mantaJJy adjusted men's occwations tended to cluster around the professional. technical, managerial, sales. clerical and mechantical fields. me maritally maladjusted men‘s occupations tended to cluster around the mechanical and manual fields. mirerences between mi’lilritally adjusted and maladjusted women were not marked. In both grows aP'melmately two-thirds of the womai were unemployed. Considering the ‘ Go 2 Dictionary of Occwafional Titles (Part IV) United States Fremont Printing Office; Washington. D.C., 19144. 96. entire occupational classification. there was a rough inverse relationship between occwaticnal level and marital adjustment for the male subjects. this was not so for the female subjects. A chi-square was computed comparing men in both the validation and cross validation grows by dividing 0-3:, 1.1: and 2.x motionary of Occwafional Titles (D.O.‘I'.) classifications into a 'high" and 3.2:. 4-2: and 6.1: into a "low" classi- fication. ihe occupations of maritally adjusted men were significantly higher than those of maritally maladjusted men at a p< .005 for the validation group and a p< .001 for the cross validation grow. hble 11. 3. D. 0. '1‘. (Part IV) Occupational Classification of Maritally Adjusted and Maladjusted Men and Women in the Validation and Cross Validation Grows D.O.T. Validation Gro_w_ Cross Validation Gregg Part IV Adjusted Maladjusted Adjusted Maladjusted Classification M F M F M F M L 0.2: 6 0 1 1 12 3 1 1 1-x 1+ 4 3 1 5 3 l! 2 2.x 0 1 2 3 2 0 0 1 3.x 0 0 0 0 1 0 0 0 4-1 11 1 6 1 4 0 3 1 6-x A; 1 8 2 O 0 1b 1 Ungwloyed O 16 2 12 O 18 42‘ 18 Unknown 0 O 0 ‘2: O 0 O 0 Total 2L 21 22 22 21+ 24 21+ fl} The ages of the subjects are reported in Table L54. Ihe mean age for both maritally adjusted men and women was from three to seven years higher than for maritally maladjusted men and women. Although the maPitalfly adjusted men and women's standard deviations were lower than the maritally maladjusted men and women's in the validation grow. the 97o reverse was true for the cross validation grow. rDnemaritally adjusted subjects in the current [study are older than the maritally maladjusted subjects although the F ratio was not significant at p (.05. However, the F ratio for the age of maritally adjusted vs. maladjusted fanales in the cross validation gmw is significant at p<.01. Table 11.4. Ages of Maritally Adjusted and Maladjusted Men and Women in the Validation and Cross Validation Grows W Age Validation: Grow Cross Validation Grow in Adjusted Maladjusted Adjusted Maladjusted Years M F M F M F M F ‘ 16-23 0 0 3 4 1 2 2 2 2A—31 2 5 6 7 5 6 9 13 32-39 8 7 9 8 9 8 8 7 140-117 11 11 2 1 5 7 2 2 118-55 2 0 1 O 11- 1 2 O M’ 23 23 21 20 _gl+ 2A 23 24 ‘ Mean 110 37 42 30 37 33 31+ 30 SD .8 .8 .1 6. 8. 10. 6.8 6. Maritally maladjusted cowles differed from maritally adjusted ’ cowles in mean number of years married more so than on any of the other concomitant variables reported. 1116 number of years married for maritally adjusted and maladjusted cowles in the validation and cross validation groups are reported in Table 4. 5. In both validation and cross validation gmups the maritally adjusted subjects were married longer than the maritally maladjusted subjects. lhe computed t_ tests for both groups "91‘ e significant at a p4 .001 . However, the maritally adjusted couples “he no more variable than the maritally maladjusted couples in their number of years of marriage (all F ratios p< .05). 98. Table 1+. 5. Number of Years Married for Maritally Adjusted and Maladjusted Couples of the Validation and Cross Validation Grows Validation Group Cross Validation Grow Years Married Adjusted Maladjusted Adjusted Maladjusted 1-5 ' 1 1o 11 12 6.10 it 6 5 1+ 11-15 6 5 5 5 16.20 9 1 5 2 21-25 2 0 j 1 M 23 22 21} 21+ 6.8 13. 8.0 Mean 15.0 0 SD 5.0 5.0 2.0 5.2 Thirteen per cent of the subjects who participated in the study were married more than once. Eleven of these subjects were'men and thirteen were women. Although there was a tendency for more of these subjects to be maladjusted, the computed chi-square was p<.10 and not significant. 'me number of subjects married more than once and the tabulation of their sex and classification of adjustment is reported in Table L136. Table 14.6. Number of Marltally Adjusted and Maladjusted Male and Female Subjects Married More than Once . Validation Group Cross Validation Grog Adjusted Malagjusted fluted Mahmud N _8em Male. 3 5 1 2 11 Fbin-ales 3 3 2 5 13 Toad: 6 8 3 3 21+ 99. The Smle Characteristics In the actual composition of the sample there were a number of socioeconomic characteristics which appeared to be correlates of marital adjustment. Maritally adjusted men and women seemed to have a greater number of years education than maritally maladjusted men and women. However. it was the type of education rather than the length which was significantly differentiated. Greater numbers of the maritally adjusted attmded college mile greater numbers of the maritally maladjusted attended trade school. The adjusted men's occwations clustered around the professional. technical. managerial. sales. clerical and mechanical fields whereas the maladjusted men's occupations were almost entirely limited to the manual and mechanical fields. The age of the participants was not significantly associated with adjustment or maladjustment in marriage with the excwtion of the females in the cross validation group. In this grow there was greater variability among the maritally adjusted females than in the other grows. me most striking characteristic of adjusted marriages was that . they were older marriages. Since there was no restrictions placed won the selection of couples in accordance with their length of marriage. the correlation between marital adjustment and length of marriage appeared to be a valid one. There were too few participants in the Study who were married more than once to determine if marriage is systematically related to adjustment in marriage. In summary. the socioeconomic characteristics which were systemati- cally related to adjustment in marriage were the type of post high school edVacation obtained. the fields of work (for the men) and the number of years married. 100. Administration of the Instruments In each case the husband and wife of a marriage were instructed to reapond to the instruments without communication with one another. This condition was insumd for the maritally maladjusted couples because they were supervised by either the investigator or a social worker . trained in administering the instrummt. Imfortamately. the same safe- guard could not be instituted for the maritally adjusted couples who responded to the instruments at home in their leisure time. It is unlikely that the maritally adjusted couples failed to meet the require- ment of non-commication because there would be little reason for them to believe that they might benefit either directly or indirectly through comunicating with one another. All instruments were filled out anony- mously. and the maritally adjusted couples returned their protocols in a plain. sealed. white envelope. There was no time limit set for coupletion of the instruments nor were the couples promised any feedback on the results of the tests. be maritally maladjusted couples were administered only the Issues Scale. while the maritally adjusted couples were administered both the Issue Scale and the Short Marital Adjustment Test. They were instructed to fill out the Short Marital Adjustment Test first. Summary The Short Marital Adjustment Test was described as well as how it was used as a criteria instrument. The steps for preparing the experi- IBental instrument were indicated followed by a section on the selection 01’ subjects. First the nature of the maritally adjusted and maladjusted 101. subjects were listed followed by a discussion of the apportioning of subjects to the validation and cross validation grows. The number of years education. occupations. ages. number of years married. and second marriages were tabled for maritally adjusted and maladjusted subjects in the validation and cross validation grows. Obtaining a college as Opposed to a trade school education. working in the profes— sional. technical. managerial. sales and clerical fields as Opposed to the mechanical and manual fields. and the number of years married were all positively associated with adjustment in marriage. There was no systematic relation between adjustment in marriage and number of years education or marriage. Lastly. the conditions which were necessary for the administration of the instruments were explained. CHAPTER V ANALYSIS OF THE RESULTS The first section of Chapter V is a description of the develOpment of “Set T" and “F" scales for use in the present study. The validation and cross validation of the items and the validation of the scale with a new sample is reported. The various estimations of reliability for the alternate scoring systems of the scale for the sanples are reported followed by a discussion on the attribution of scores to the experimental scale. Develwment of the Set T and F Scales As a preliminary step in the develOpment 'of the validity scales. all of the validation grow's responses to the items were tabulated into two sets of contingency tables. The first set was constructed to receive the tabulation of responses scored according to the personal construct theory. and the second set was constructed to receive the tabulation of responses scored according to the straight agreement theory. The Set T Spite Through observation. items which seemed highly controversial. i.e. were responded to by both maritally adjusted and maladjusted cowles with almost equal probability in the direction of marital adjustment or maladjustment. were collected for the development of Set 1‘ Scale. The 1deal items for Set To Scale would contain an equal number of subjects in each of the four cells of the contingency table. A priorly. three criteria were established in order to select the 102. Vt 103. items for the Set T Scale: 1) the deviation between the maritally adjusted and maladjusted cowles' responses had to be no greater than 20 per cent of the total responses made per item; 2) the distribution of scores in the cells had to deviate no more than 30 per cent from chance ewectancies. and 3) the first two criteria had to be met by both the validation and cross validation grows. The above criteria were met by three items scored according to the personal construct theory and six items scored according to the straight agreanent formulation. Due to the small number of items meeung the criteria. the Set T Scale was discarded. It is highly possible that the lack.of items meeting the criteria was in part due to the nature of the item. Each item was carefully constructed to differentiate between maritally adjusted and maladjusted couples. The small number of items in the original pool as well as the stringency of the criteria also may have limited the number of items meeting the criteria. m P Sale Items were collected for the F Scale by selecting those itans in which one type of response was chosen by both maritally acUusted and maladjusted couples with alternate responses rarely chosen. The ideal 1" Scale item elicites near unifonn responding regardless of the degree of'adjustment in marriage. he procedures for developing the F Scale have been discussed (supra. p. 102). Eight items scored according to the personal construct theory met the established criteria. while ten items scored according to the straight agreanent formulation met the established criteria. In tame of 'the criteria the straight agreement itans were closer to the 101+. hypothetical ideal item when compared with the items scored according to the personal construct theory for the F Scale. me ten items scored according to the straight agreement formulation were chosen to cowrise the F Scale. These items and the direction of their respective statistically rare responses are listed in Table 5.1. Table 5.1. i The F Scale Itans and Rare Responses W m1 Item Rare Regonse 13 A person must be responsible for what he does. disagreement 26 ‘A marriage is happy when the husband and wife . love each other. disagreement 31 Children should be included in a husband and ' w . wife's plans for a good time. . disagreement 1&5 Husbands ani wives alould spend their free time together. disagreement 5o In-laws should be made to feel welcome in a cowle's home. ' ', disagreanent 57 Your husband's or wife's pamts are entitled to the same respect as your own pamts. disagreemalt 59 If a cowle would have good times together. thq would have fewer problems. disagreeth 81 The husband should be the breadwinner in the : family. disagreement 98 If people would stop worrying about how much better off their friends are. they would be disagreanent happier. 1 18 The spiritual part of life is iwortant. disagreement 105. The proportions of subjects who answered from one through seven itmns of the F Scale in the statistically rare manner were tabulated. . and T scores were attributed to the accumulated preportions. ‘me number of subjects responding to the F Scale items in the statistically rare manner. their respective accumulated proportions. and their T scores are reported in Table 5.2. . Table 5.2. Number. Accumulated Proportions. and T' Scores of Rare Responses to the F Scale Number of Number of Accumulated m Regpgnges Subjects Pygporgon T Score 0 61 .64893 54 1 21 .87233 61.5 2 9 .9680? 68 3 2 .9893“ 73 h o .9893“ 73 . 5 0 .9893“ 73 6 0 .98934 73 Z 1 .9999? 80 The cut-off point for the F Scale was based won the wave of the response distribution. Referring again to Table 5.2. it was noted that 98 per cent of the cowles scored below a T score of 71}. A natural break occurred in the curve of the distribution after a T score of 73 with one couple remaining at a T score of 80. The cut-off point was established at the T score of 74 which had the effect of dropping one couple who scored at 80 from the sample. The reliability of the 1“ Scale was computed for the entire group Qt 94 subjects according to Hoyt's Analysis of Variance Method.1 This v 1 c. J. Hoyt. “Test Reliability Estimated by the Analysis of Mame Method." thometrika. VI (1941). pp. 267-87. 106. method of estimation was chosen because it yields results based upon the error variance and affords a computational procedure derived from the theoretical definition of reliability. The estimated reliability of the F Scale for the respective samle was .503. 'me respective computa- tions and Analysis of Variance Table may be found in Appendix C. The a priori assunption concerning. the F Scale was that rarity of reSponse when actreme was associated with mismderstanding of directions. reading difficulties and "faking bad". A recent study by mweatt:a studied the F Scale in an objective test battery of motivation which was administered to 4200 Michigan elevajxth grade students. His conclu- sions were that underachievers select significantly more F itaus than overachievers .' further investigation with the F Scale should be . conducted before itis used to delete subjects from studies (particularly males). the F Scale represents. a measure of social'conffomity. and'it posseses the ability to tap e. academic masculinity-feminity‘ continuim. Ihe implication for the current study is that it is possible that the one cotple was deleted from the sample for spurious reasons. Analysis of the Data A restatulent of the Mpotheses and the findings tron the validation, and cross validation groups' data are found in this section of Chapter v. 1118 findings are discussed in term of the respective hypotheses. Rea guinea; of mmses MW Maritally a usted couples agree no more than maritally f 2 R. C. Thweatt. ”The Developmmt and Validation of An F Scale div he Objective Test Battery on filtration" (Unpublished Ph.D. ’ Seer-tenet. Michigan State University. 1961). pp. 87-90. ! 107. maladjusted couples on the truth or falsity of issues perceived important by both mates. Alternate I: Maritally adjusted couples agree more than maritally maladjusted couples on the truth or falsity of issues perceived important by both mates. ull es I --St t «sent 800 Maritally adjusted couples agree no more than maritally maladjusmd couples on the truth or falsity of issues regardless of the perceived inportance of the issues. W: Maritally adjusted cowles agree more than maritally maladjusted couplesen the truth or falsity of issues regardless of the perceived inportance of issues. Null mmesis III-e-Qnmt of 2e Sale Items derived fmm the spousal agreunalt correlates do not discriminate between maritally adjusted and maladjusted couples. ' _ W: Items derived from the spousal agreanent correlates discriminate between maritally adjusted and maladjusted couples. W Each of the 130 item of the Issues Scale (140 items minus the ten F Scale itans) were validated twice: first when scored according to the personal construct theory and secondly what scored according to the straight agreemalt formulation. The chi-square statistic was used to determine which itans discriminated between maritally adjusted and maladjusted combs of the validation group at a predetermined p (.10. The modified version of the MISTIC Digital Conputer Program K6»:3 was used to comute the chi-sqmres according to the formla .6 x2: %(01-E:-o5)2 131 E1 me sun of the blocks of mtrices kl t; = whereKisthenumherofrowsand1 i=1 is the number of calms . O. ' = The number of observed frequencies. 3 MISTIC Digital Computer Program K6M available upon request: gISTIC Office; Fifth Floor. Electrical Engineering Building: Michigan tats University; East Lansing. Michigan. 108. E1 = The number of expected frequencies. The modified version of K6M is recomended when using small numbers. Thirty-eight of the items scored according to the personal construct theory and forty-eight items scored according to the straight agreement formulation were ‘significant at a p < .10. Since the hypotheses were dimctional and the tests of the hypotheses one-tailed. the directions given by Mellow“ for. detennining one-tailed levels of significance from chi-square tables were followed. i.e.. tabled chi-squares for p<.10 were found in the p<.20 column. Cross Validation Group fig The computational procedures and statistics for the cross validation analysis were the same as those for the validation. Thirty-nine of the items scored according to the personal construct theory were significant at the p<.10 while forty-one of the items scored according to the straiglt agreement formulation were significant at the p (.10. again. item productivity was somewhat greater for the straight agreement scoring system . Twenty of the straight agreusent items and thirteen of the personal construct itans were significant at p < .10 for both the validation and cross validation groups. Ten of the twenty straight agreement items contained theoretical chi-square cell eXpectancies of less than five in either the validation or cross validation tests of significance. Walker and 131; state that chi-squares cemented on contingency tables “ a. mm. W (New York= “my- 1955) page 231 . 5 H. M. walker and J. J. Lev. Statistical Inference (New York: H0113: 1953: 1953): P389 104-, 109. with arm cells containing a theoretical expectancy of less than five may be spurious. They suggest that in order to rectify the above circum- stances. the Exact Probability Test should be used for items with cells containing less than five eXpected cases. 1 , In accoi'dance with Walker and Lev's recommdauons. the Exact Probability Test was canputed for the ten items" which contained cells with less than a five expectancy. All of these items discriminated betwoen maritally adij and maladjusted couples at a p <.1’o or better. . The steps through which items were eliminated is sunnarized in Table 5.3 while the computed chi-squares for all the items are reported in Appendix A. Only item which were significant in the validation and cross validation analyses were retained. The straight agrsumt scoring system was more productive in yielding items than the configural scoring systan. Ten of the forty straight agreement tests of significance for . retained items contained cells with less than five expected cases. None of the configural scomd retained items had less than five enacted cases per any of the cells. Table 5.3. Number of Items Generated by the Configural Scoring and Straight Agreemalt Scoring for Validation and Cross Validation Groups number of ;m fielded at p< .10 Configural 1 Straight Agreement ' m Scomg Seem . I Validation 38 148 II Crpss validation 39 41 III Significant items in ‘ validation and cross 13 20 M 110. The significant items were drawn from between four to six of the seven area correlates of marital adjustment from which they, originated. The retained items for both scoring systems. the area correlates from which they originated. and the percentage of items from each area correlate are reported in Appendix B. It may be noted here that all of the configural scored itmns and eighteen out of twanty of the straight agreement items originated from the areas ”Handling Finances". ”Dealin'g with In-Laws". "Recreation". ”Intimate Relations" and “Demonstrations of Affection”. Two of the significant straight agreement items originated from the area of "Friends”. No items. regardless of scoring system. originated from the area "Philosophy of Life". Estimates of Reliability The reliability of the Issues Scale was estimated for the maritally adjusted and maladjusted couples in the validation and cross validation groups for the two scoring systems by Hoyt's Analysis of Variance Technique.6 In all. eight reliability coefficients were calculated. These are reported in Table 5.1;. has computations and analysis of variance tables are found in Appendix c. Table 5.4. ‘me Hoyt's Analysis of Variance Estimate of Reliability for the Issues Scale _-__ Scoring Sfitem Couples Straight Agreement Configured Validation: . Maritally Adjusted .559 .803 Maritally Maladjusted .551 .866 Cross Validation Maritally Adjusted A77 .905 muggy Lagdjusted .279 .602 6 C. J. Hoyt. 22° 933.. pp. 267-87. 111. It may be noted that with the exception of the cross validation maritally maladjusted sanple each of the reliability coefficients for the configural scored items were markedly higher than the straight agreement scored items. The erratic reversal of the cross validation maritally maladjusted group coefficient mar be attributed to a greater degree of error variance present in this group's responses. Further sanpling would be necessary in order to determine if the reversal of the coefficient for the one maritally maladjusted sample was a sampling artifact. Considering the marked increase in reliatxility of the configural scored items for three out of the four samles. the difference between the two scoring systems and their respective formulations appears to be a qualitative rather than quantitative one. Although the straight agreement scoring system was more productive in yielding items than the configural scoring system. the latter afforded a more reliable measure of adjustment in marriage for that population hypothetically derived from the samples used in the stuck. Items which mates agree upon and perceive as important were more reliable measures of adjustment in marriage than items which yielded only spousal agreement regardless of the mates' perceptions of the items. Attribution of Scores to the has of the Experimental Scale The focus of the present thesis was to determine the value of developing a configural scoring systan for content derived from theory in the area of marital addustmmt. The investigator chose to limit scores inferred from both the personal construct and straight agreement theories to a simple “1"-JO“ dichotomy; "1" if mates responded in the direction of marital adjustment and '0" if'mates responded in the 112. direction of marital maladjustment. The frequency of positive or "1" scores for. maritally adjusted and maladjusted subjects in the validation ’ and cross validation groups were tallied. he means and standard deviations of the tallied scores are reported in Table 5. 5. In order to test the difference between maritally adjusted and maladjusted mean scores 1 tests were comuted. In all. eight~ tests were computed. All differences were significant at the p<.001. no standard deviations were relatively high for both maritally adjusted and maladjusted , samples. The high standard deviations reflected the high variability in small samples which was the case in the current stuw. Table 5. 5. Mean Score and Standard Deviations of Haritally Adjusted and Maladjusted Validation and Cross Validation Groups on the Issues Scale » §trgight firemen; Mm! . Miss SD N . 8 Validation . Haritally Adjusted 22 18.00 2.24 22 7.79 3.97 2.68 3.49 Maritally Maladjusted 22 9.25 2.98 22 Cross Validation ‘ Maritally Adjusted 21+ 7 17.45 2.00 21+ 7.33 1.58 us 21+ 11 . 2. .6 2.118 C s V o : sues S . n1. cross validation group couples' scores on the validated forty-' eight straight agreement items and thirty-eight configural items were tabulated. Forty-eight was the highest score that could be eamed on the straight wt items while thirty-eight was the highest score possible on the configural item. The ...... scores and standard deviations for maritally adjusud and maladjusted couples on the valida- 113. ted straigxt agreement and configural items are reported in Table 5.6. I Table 5.6. 1 Cross Validation Groups' Mean Scores and Standard Devia- ' tions on the Validated Items of the Issues Scale W ' ‘ W $1:th agreement Configugfl Cross Validation Couples. g Mean SD N Mean 32 Marit_a_ll_y Malagjusted 21» 28.2; £1.56 g4 3.5L» 5.70 The _t, and 1; tests were conputed in order to determine whether the validated Issues Scale itans differentiawd betweui maritally adjusted and maladjusted couples in a new sasple. Cross validation group maritally adjusted couples earn significantly higher scores on the validated straight agreement scored items (_t_ = 6.83; df =- 1+7: p<.0005) and the validated configural items (t= 3.11: df = 1+7: p <.005) when compared with cross validation group maritally maladjusted couples. Cross validation maritally adjusted couples were significantly more variable than cross validation maritally maladjusted couples in their responding to the validated configural scored items (:3 = 2.70; n, =- 23; n2 =- 23; p< .05). The computed g for cross validation couples' responding to validated straight agreement items was not sufficient to reject the null hypothesis of equivalent variance (g =- 1.28; N1 = 23: N2 = 23; 8.3.). In Tables 5.7 and 5.8 the overlap of the maritally adjusted and maladjusted couples' scores and the range of their scores are reported for the two reSpective scoring syetans. The percentage of overlap for the straight agreement scored items was 77. and the percentage of over- lap for the configural scomd items was 85. It may be noted that there is an overlap between maritally adjuswd 111+. Table 5.7. The Cross Validation Maritally Adij and Maladjusted Couples' Scores on the Validated Straight Agreement Scored Items Straighthgreement Cass Valli-Q1203 @5219; W Mutated team Musk; 21. - 23 1 , u 2% - 26 L. 27 - 29 5 30 - 32 2 u 33 - 35 3 5 36 - 38 7 2 39 - M 8 1&2 - M 3 n 2)), 211 Table 5.8. ms Cross Validation Maritally Adjusted and Haladjusted Couples' Scores on the Validated Configural Scomd Itans Configure]. Scogglntervals 0- 2 3- 5 3 6- 8 9-11 12-1l+ 15-17. ‘ ; 18-20 21-23 $1 81 reunites-Nat.) “631": II #333 115. and maladjusted couples' responding to both the straiyit amement and configural scored items. The extensive range of maritally adjusted . couples' scores in Table 5.8 seem to be a function of the configural scoring system. The range of scores was inure limited for the same subjects on the straight agreenent scored items. It seems that what couples' scores were penalised or enhanced according to their agreement or disagreement on the perceived importance 'of issues. a number of maritally adjusted couples responded to the scale like malitally maladjusted cowles. The reliability and validity of the 11:0 scoring systans which were the operational counterparts of the personal construct and straight agreement hypotheses were studied in order to confirm or reject the null hypotheses. The scoring systeus were considered valid if they produced a sufficient number of items to assure that differences were not occur- ' ring by chance. The mwritten standards in scale development is that itm retention should be about .333 at validation and .111 at cross validation. At the validation stage .292 personal construct item were retained. and .10 personal construct items were retained at the cross 1 validation stage. At the validation stage .369 straign agreement items were retained and .153 at the cross validation stage. Since both scoring systems approximated the unwritten standards and were sufficient in number to reject the possibility of chance significance (1 to 5 significant items per a pool of/100). the content twpothesis was accepted. The reliability for the configural itms was markedly greater than for the straight agreement items. However. the straight agreement scoring systus was more productive of items which differentiated significantly betwm maritally adjusud and maladjusted couples. The 116. straight agreement items afforded less of an overlap between maritally adjusted and maladjusted couples of a new sanple (the cross validation group). , Rejection or acceptance of either the personal construct or straight agreement hypotheses must therefore be considered in terms of requirements necessary for acceptance or rejection within the context of a pilot study. Both the personal construct and straight agreement hypotheses were accepted with the following reservations. be personal construct hypothesis provided a reliable framework for the prediction of marital adjustment. However. it failed to generate a scoring system which clearly distinguished between maritally adjusted and maladjusted couples of the cross validation group. ‘mis inadequate differentiation may be a function of either a faulty criterion of adjustmmt in marriage or an imperfection in the refinement of the personal construct hypothesis. the straight agreement hypothesis provided a framework for sealing adjustment in marriage with mediocre reliability. 'lhe relieve nemts of the two hypotheses may be oompamd by considering that the maximal validity of a scale can be no higher than the square of the mandmal reliability of a scale. In this ”Spect. the personal construct hypothesis when compared with. the straigxt agreement hypothesis provided a qualitatively superior framework for the scaling of adjust— ment in marriage. Summary 1119 results of the stuchr were reported in Chapter V. ‘Dle data relevant to the Set '1‘ and F scales were analyzed. ‘nle Set ‘1‘ scale was discarded due to the lack of items meeting the criteria for inclusion in 117. . the scale. Eight of the straight agreement scored items and ten of the configural scored items not the criteria for inclusion in the 1' scale. me straight agreement itans were chosen to comprise the F scale. and one couple was deleted from the sanple because of their significantly. high score on these items. . 1 Items were analyzed for the validation and then the cross valida- tion groups. At the validation stage forty-eight straight agreement scored items and thirty-eight configural scored itas differentiated significantly between maritally adjusted and maladjusted couples. Nenty of the straight agreuent itus and thirteen of the personal construct items remained significant when cross validated. ‘lhe reliability of the Issues Scale was estimated for the maritally adjusted and maladjusted couples in the validation and cross validation groups for the two scoring syetus by Hoyt's We of Variance Tedlnique. In all. eight reliability coefficiults were calculated. With the exception of the cross validation maladjusted eagle (r =- .602). each of the reliability coefficients for the configural scored items (r's :3 .905. .866. .803) were markedly higher than the straiglt agree- ment scored items (r's a .l+77. .511. .599. .799). Scores were attributed to the Issues Scale and t tests were coeputed in order to test for differences between the mean scores of cross validated maritally adjusted and maladjusted couples on the .validamd items of the Issues Scale. the differences between the means were significant at a p< .0005 for the straight agreement items and a p (.005 for the configural scored items. Cross validation group 118. maritally adjusted couples were siglificantly more variable than cross validation group maritally maladjusted couples in their responding to the configural scored validation items (p 4.05). CRAP TER VI S UMMARY AN D CONCLUSIONS ‘Ihe first part of Chapter VI consists of a summary of the problem, design. methods and pmcedures, resllts, and limitations of the study. The conclusions of this pilot study are then stated followed by suggestions for future research. Summary '13:? From as. The increasing recognition of rising divorce rates as a social problem and the abundant evidence for the success or failure of marriage depending upon the personal relationship of the mates have converged to provide a problem area for stuchr by the behavioral sciences. The study of marital success or failure is not new. However. studies anteceding 1950 have suffered from inapplicable instrumentation. faulty research design and meager theoretical foundations. Within the last ten years suitable instrumentation. appropriate research methodology and use of theory have increased the adequacy of research on success or failure in marriage. One barrier has remained sound-the problun of individual ferences in marriage. What is the panacea for one marriage may be the mination of another. ‘Ihe purpose of the present thesis was to construct a scale of marital adjustment and to determine whether the inclusion of the variable of perceived importance of the items by the mates would increase the reliability and validity of marital scale construction. A personal 119. 120. construct theory of behavior was employed in order to predict how the mates' perception of items would influence the measurement of their adjustment in marriage. It was hypothesized that the consequence, marital adjustment. was a function of the way the mates perceived each other and their marriage. A second purpose of the study was to deter- mine whether or not spousal agreement areas correlated with marital adjustment would provide a suitable foundation for the derivation of scale content. It was predicted that the spousal agreement areas correlated with marital adjustment which the mates agreed upon and both perceived as important were reliable indicators of marital adjust- ment. In order to determine whether the mates' perceived importances of issues warranted the consideration in the scaling of marital adjustment that was inferred from theory. two scoring systems were developed for the eacperilnental scale. ’me first took cognizance of the agreement of mates on issues correlated with marriage while the second gave scoring weight only to those issues agreed upon and perceived important by the mates. ‘Ihe Desim 'me independent variables of the study were adjustment and malad. justment in marriage. The former was determined by the mates' scores on the Locke-Wallace Short Marital Adjustment Test. The latter was determined by choosing mates who were starting procedures for divorce or receiving professional marital counseling for diagnosed marital difficulties. 121. 'lhe dependent variables of the study were agreement of mates on issues taken from areas known to be correlated with marital adjustment. the perceived importance of these issues. and the content of the scale. me Issues Scale, the Operational counterpart of the dependent variables“ was composed of 1160 items selected from seven spousal agreement areas known to be correlated with marital adjustment. These areas were (1) dealing with in-laws. (2) intimate relations. (3) friends, (it) recreation. (5) Philosophy of life, (6) handling finances. and (7) demonstrations of affection. No scoring systaus were constructed for the scale: the first measured mates' agreements on the items and was labeled straight agree. ment scoring; the second scoring system was configural and measured the mates' consensus on the perceived inportance of the items as well as their agreement on these items. For the latter, six differing combina— tions of agreement in perceptions of items and [agreement on items were tallied although those tallies were condensed into "1'--'0' dichotomies which were inferred from the personal construct and straight agreement ' formulations. In order to control extraneous variance. two validity scales were proposed. me first. an F Scale. was constructed by cospiling those items which elicited responses almost unanimously in one direction regardless of the degree of adjustment in marriage. The second. a ’1‘ Scale. was constructed by compiling those items which were contro- versial for the subjects regardless of the degree of adjustment in marriage. Accordingly. couples who would respond to the compiled items in a statistically significant rare manner were to be rejected from the sample in order to reduce error variance. Previous research on validity A 122. scales assumed that F type items measured misunderstanding of the scale, reading comprehension difficulties and confused thought processes. It was suggested that the Set T type items measured rigid reSponse bias attributable to resistance, boredom. inflexibility. unconcern and other unknown variables. There were eight temporal steps in the design of the study. The first involved the preliminary administration of the Issues Scale in order to refine the instrument. Following. the Issues Scale was administered to all the groups to be used in the analysis of the data. and the criterion instrument was administered to the maritally adjusted sample. The validity scales were analyzed followed by the validation analysis of the data. Cross validation followed in order to replicate the validation results. Reliability of the Issues Scale was estimated for each of the subgroups. The scale was then cross validated by analyzing cross validation group's scores on the validated items. The final step was a judgmental rather than statistical one. It involved the determination of whether the configural or straight agreement scoring should be used to determine marital adjustment and whether or not item productivity was sufficient to insure that the derivation of the reliable and valid items was not due to chance. The first two hypotheses referred to the alternate scoring systems as follows: (1) maritally adjusted couples agree more than maritally maladjusted couples on the truth or falsity of issues perceived important by both mates; (2) maritally adjusted couples agree more than maritally maladjusted couples on the truth or falsity of issues regardless of the perceived importance of issues. The third hypothesis referred to the content of the scale. It was predicted that item productivity was 123. sufficient to insure that reliable and valid items were not derived spuriously from the spousal agreement areas of marital adjustment. The chi-square statistic was used to test the significance of items in fiscriminating between maritally adjusted and maladjusted couples. When chi-square emectancies were below 5, the Fisher Exact Probability Test was used in addition to the chi-square statistic. T Scores were used in the comilation of validity scale scores, and Hoyt's Analysis of Variance Method of estimating reliability was used. Methods ad Procedures The Locke-Wallace Marital Adjustment Test was used as the criterion instrument for selecting maritally adjusted couples because of its inclusion of maritally maladjusted couples in the standardization. its reliability and validity. and its brevity. In the standardization of the criterion instrument 96 per cent of the maritally adjusted group soared above moons 17 per cent of the maritally maladjusted gimp‘ scored above 100. In the current stuw mates chosen for the maritally adjusted group had to earn a score of 100 or more on the criterion I instrument. , The development of the eacperimental scale involved the inventing of items to represent seven area correlates of marital adjustment. pilot administration of the original pool of items and revisions. determination of the effects of item order and age-grade level of words used in the scale. . Subjects were persuaded to participate in the study by various social workers. clergyman and one educator who cooperated in enlisting subjects. Cognizance was given to the hostility maritally maladjusted couples might feel in participating in such a stuw. to the private 121+. nature of the research. and to the known difference in socioeconomic characteristics for maritally adjusted and maladjusted couples. No 'attemt was made to restrict subject selection on any variable other than adjustment and maladjustment in marriage. The sample was regarded as the basis from which hypothetical pepulations were. inferred for generalisation purposes. Thirty-nine of the maritally maladjusted subjects conprise the entire group of mates filing for divorce or on probation pending divorce at the Detroit Recorder Court over a designated period of time. The remaining eight maritally maladjusted couples were in the beginning stages of marital counseling at the Catholic Social Service Agency: Lansing. Michigan. One of the couples from the Detroit Recorders Court was rejected from the sample because of their significantly high score on the 1' scale. Fifty-eight couples were persuaded by their clergyman and one educator to participate in the study. The Locke-wallace Marital Adjustment Scale was administemd to these couples, and forty-seven of them exceeded the pre-set score of 100. Half of the maritally adjusted couples and half of the maritally maladjusted couples were assigned to the validation group with the aid of a table of random numbers. The remaining halves were assigned to the cross validation group. Maritally adjusted men and women“ attended school longer than maritally maladjusted men and women; however. the differences between the means was not significant. It was worth noting that a significantly greater number of maritally adjusted men obtained college rather than A trade school training when compared with maritally maladjusted men. 125. more was a significant inverse relationship between the level of Ilictionary of Occupational Titles (Part IV) Classification for the men and their respective level of adjustment in marriage. Bio-thirds of .the women were unemployed. For those who were enployed. there was a traid in the relationship between marital adjutment and occupational classi- fication similar to the relationship for men. he mean number of years married for maritally adjusted couples was significantly higher when conpared with maritally maladjusted couples. mm per cent of the subjects were married more than‘once. There were no clear-cut indications that being married more than once wa‘s related to either marital adjustmentor maladjustment~for the subjects in~ a systematic manner. . The instruments of the stuck were administered to the maritally ‘ maladjusted subjects by the Investigator or social workers trained in the administration of the scale. Maritally adjusted couples filled out the scales at thdr own homes in their leisure time. They were instruc- tednotto talkwiththeirmates untilthe scales were returnedina provided plan. white. sealed envelope. 2 W The analysis of the Set ‘1‘ and l" Scales were given first priority in order to determine if any of the couples were to be eliminated from ' the major analyses of the study. Three criteria were developed for the selection of both validity scale itus: (1) the deviation betwem the maritally adjusted and maladjusted couples reponses had to be no greater than 20 per cent of the responses made per item. (2) the distribution of scores in the cells had to deviate no more than 30 per cent from chance expectancies in the direction of marital adjustmmt or maladjustment. FF" 126. and (3) the first two criteria had to be cross validated as well as validated. in insufficient number of itas (3) met the criteria for the Set T Scale regardless of scoring systaa. Eight configural items and ten straight agreanent items met the criteria for the 1! Scale. The straight agreement scored F Scale was retained. and T scores were computed for the sanples' responses to the scale. One maritally maladjusted couple was rejected at this point who earned a I score of 80. There were 98.9 per cent of the subjects who earned T scores no higher than 73. The '1‘ score 71} was established as the cut-off point for rejecting couples from further analyses. The reliability of the F Scale fe‘~ the total sample of 93 subjects was .503. ~ 3 In the next stage of analysis the validation group data was inspected for item productivity. Thirty-eight of the itms scored according to the personal construct theory and forty-eight of the items scored according to the straight agrecnent theory were significant. When the same items were cross validated. thirteen of the personal construct items and twanty of the straight agremmt items remained significant. . Ten of the twenty straight agreement items required further statistical analysis (Fisher Exact Probability Test). and these ten remained significant. The significant items from either scoring system originated from all of the area correlates of marital adjustment used in the study with the exception of “Philosophv of Life'. 'nie areas “In-Laws”. "Intimate Relations“. and ''Handling Finances' were the most productive of items. The Analysis of Variance Reliability Coefficients ranged from .599 to .779 for the straight agreement items and fret .602 to .803 for the 12?. personal construct items. The personal construct item were more reliable for three of the four groups. The Issues Scale was cross validated by using the cross validation group as a new sample to which the validated items were administered. The mean scores for the cross validation maritally adjusted and malade justed couples were computed. 1.; tests were used to test the significance of the means. and :1 tests were used to test the significance of the variance. The maritally adjusted couples earned significantly higher scores thah the maritally maladjusted couples on both the straight agreement and configural scored scales. me maritally adjusted couples' scores were significantly more variable than the maritally maladjusted couples' in .their responding to the configural scored scale. In sumary. the straight agreement scoring system produced more items while the personal construct scoring system produced a fewer number of more reliable items. Although both straight agreaamt and personal construct formulations afforded a valid measure of adjustment in marriage. the latter provided a scale which was somewhat qualitatively superior to the toner. The configural hypothesis. which referred to the perceived imertance of the itans to the mates and the straight agreement hypothesis which referred to the agreement of mates on the items regardless of their perceptions were both accepted with reserva- tions. no content lupethesis was accepted since there was a sufficient number of reliable and valid items generated from the spousal agreement correlates of marital adjustment to insure that the itas were not derived spuriously. 128. ta of the S The major limitations of the study were threefold: the small number of subjects engaged in the stuchr. the relatively gross scoring system. and the insufficient nmber of items developed for the original pool. 2 The restricted umber of subjects engaged inthe stuck tended to decrease the number of items which were found to be significant. The chi-square curve is dependent upon the number of subjects in the sample. and it is possible that if the analyses were carried out for larger groups. there would have been a greater number of items which remained significant through out validation and cross validation analyses. 'nle probability of lesser items remaining significant with larger groups is rare because of conservatism in the chi-square model involving a relatively few number of cases. A deeend limitation was attributable to the base of from 23 to 2A couples per group. A larger sample would have offered a larger and more easily definable base for gmeralizatien of the results to hype- thetical populations. ‘ The relatively gross scoring systems for the scale tended to hide interaction within the '1'-"0" dichotemies. The responses were originally classified into one of six classifications. 'nlese classifi- cations resembled a continuim in which mates' agreement on truth and mates' strong agreement on inpertance was the high point. Cemen sense might suggest that the opposite of the above configuration. mates' disagreement on tnlth and strong disagreeaent on strength of feelings would represent the other end of the continuim. Believer. theories of marital adjustnent have not been specific enough to provide a basis for 129. weighting varying reaponses along a continuim. The 1140 item starting pool, in interaction with the relatively small sample, did not yield a sufficient number of itans to systematically sample the composition of marital adjustment. The interaction beta-ream the original number of items and the size of the sample and their effect Upon item productivity was difficult to disentangle. A measure of safety would be afforded if both the original pool of items as well as the size of the sample were doubled. Minor lintations of the study included the failure of the validity scales to function in the assumed manner and the lack of homogenity of the sample. It must be recognized that the greater frequency of marital maladjustment among lower socioeconomic groups tends to influence definitions of marital adjustment so that they are useable for studying couples from these strata. “me question of whether or not marital adjustment entails different characteristics among the various socio- economic groups has yet to be determined empirically. Conclusions The purpose of the present study was to construct a marital adjust- ment scale from a conceptual framework of mate's agreement with content derived from spousal agreement correlates of marital adjustment. Tue scoring systems were compared. one inferred from the agreement of mates on issues correlated with marital adjustment and the second inferred from a theory of behavior which keynotes the mates' consensus on the Perceptual importance of issues correlated with marital adjustment. 130. It was concluded: 1. For a hypothetical population inferred from sanpling a wide range of Socioeconomic groups. consensus of mates on issues and their perceived importmce provided a Mr]: for reliable and valid scaling of marital addushmt. 2. Mates' consensus on issues regardless of their perceived imartance afforded a framework for scaling marital addustmmt of ‘ mediocre reliabiliw. 3. ’me trachtional interpretation of oomatability in war- rlageas afunctlonof spousalagremtincertainwelldefinedams is a misleading framwork fostering only partial scaling of marital adjustment. ’lhe traditional framework warrants alteration. Carpets.- bility in marriage is a function of spousal agreuent in the areas of "Recreation”. "Intimate Relations“. I'Dealing with In-Laws'. I'Handllihg Finances". and ”Friends" and the mates' perceived inportance of these areas. . 1+. Whm marital adjustment was construed as spousal agree- ment on issues and their perceived importance. 69.2 per cent of the items were drawn from the areas “Recreation“ and "Intimate Relations”, 30.6 per cat from the areas "Dealing with Ilia-Laws". ”Handling Finances" and 'inends". No itans were drawn from the area “Philosophy of Life'. Since the items were not derived- factorially. caution should be used in interpreting the item productivity of areas based on the percentages in the’current smchr. ' 131. Suggestions for Future Research Scaling marital adjustmult through the use of statistical methods and instmmentatlon typical of objective personality inventories appears to be a fruitful approach. However, the complexity of the area to'be researched necessitates the inclusion of the following safeguards. ‘ 'Ihe amount of items developed for an original pool should be inversely proportionate to the size of the sample. In order to develop, scaling methods of marital adjustment beyond the emloratory stage, the size of the sawle should be sufficiently large to provide representation from various socioeconomic groupings. Another reason for increasing the sanple over and above the amount acceptable for instruments scaling individuals is that couple interaction tends to increase the snout of variability found in a given sample, other things being equal. If validity keys are to be included in scales of marital adjustnmt. their mmirical value needs to be detemined. more is little empirical evidence that validity scales measure what they are purported to measure for individuals and no evidence of this kind for cmples. If eupirical validity is established for Such a scale. it is recommended that items be chosen especially for the validity scale rather than using the dis- carded items from the original pool. Again. the greater variability ‘ stalling from couple interaction tends to blur the distinctiveness of items ascriminaung bebwem marital‘adJushnmt and maladjustment and item discriminating bebceen faking and sincere responding. Factor analysis of items measuring mates" perceptions is another research project of value. It not only would cut down on the number of overlapping items but would also aid in the diamosis of the areas 132. perceived by the mates as contributory to marital adjustnent relative to current cultural values. . ' _ ¥ Another area which has clouded many attempts to scale marital adjustment but.has never been scrutinized.is the accomodation factor in marriage. In aimlest terms the accomodation factor is the label givm to the response of a mate when he or she responds in a manner which is believed to be favored by the spouse. The accomodation takes place not. out of conviction but out of adjustment to one another. It is usually more prevalent along women than mm and alum has been correlated with adjustment in.marriage. I . me most severe lack in scaling marital adjustment is the absence of systematized theory from which complex scoring systems could be .developed.which would.be commensurate with the complexity of the human relationship to be researched. In the present stuw there were eight response combinations possible. although only'tao of these combinations could be inferred clearly from.theory. If systematized.theory'on.marital adjustmnt is not forthcoming. the researcher in the field may be forced to work baclotards. i.e.. develop scoring systems. stuck the marriages they apply to, and make post hoc generalizations for the use of future investigators. Considering the preceding suggestions. further research on.marital scale development should involve a large starting pool of items. stratified sapling from a wide range of socioeconomic groups (500 to 1,000 calples). cognizance of the accomodation factor in marital adjustment and factorially derived scoring categories. ’me validity scales would require upirical validation. and a substantial amount of characterising data should be collected on the couples. BIBLIOGRAPHY Ahmed. E.S .. et. a1. "Factors in Marital Iliscordfl' Journal 0 W ILIV (July 1957). 193.222. Bernard. Jessie. ”me Distribution of Success in Marriage.“ m J;_eumal__ of Sociolon, XXIII (1933). 1915203. Bernard, Jessie. "Factors in Marital Adjustment.“ W W: XL (1931‘)- “9'50- I Brodbeck. new. "Models. some. and Theories, s e um W. (Ed.) Llewellyn Gross, Evanston. Effie“: 30'. Perterson, 1959s 373-“03- Burgess. E. W. and Cottrell, L. 3.. Jr. {mgm Success 2r £31.13}: W- New York: Prentice Hall, 1939. Burgess, E. W. and Walldn. P. - ”Predicting Adjustnent in Marriage from Adjustment 1n Engagement. " MW- mx (19%). 3211-30. Corsini ,Raynond J. "Understanding and Similarity in Marriage ng 9f Missal and Social Pmlggz. LII (1956), 327-32. mvis. Katherine B. F c rs the S Life Tu t - o Hundmd W ~ I on one. .8 campus a .3 £393. o8 5 ea. omega 4 .K 8.0 3.0 mafia 8.0 . owfioo Bahama dammed: 3 538 e 9:... 33-5 as omo.m .m.. eNe.o ma.o .oeaa o.oammoe a can delays oaoauaH .me was was 3.0 8.0 .33 smog handled has no mas o» nomoao away one conned 4 . mfl. 91 no.0 1.9.0 .oeeeo 8. face 5 does oeoea: ea. 98 e515 . No.0 2; 5.0 2.0 nae one on do» page no»? 3 on 25.5. fie. meadow we sateen 2e. .mm m. .fi 13.: «sum .xeaufidofioa a 833s 3 Hayseed ea .3 . a 0?.0 . om. mow.w .mcwaom.eammoo any ad new one» . m o>me easoom are» .hvcoe oaqeoo m osam mamaica :en3 .0 also Ed has «5; H333". ea 93-5 SS .2 .a $3-5 .<.m Menominee do. Meaemfihe 2de I . tease 8383; 395 defined; . .83 one 8.2 Mateo... .<.m .. 5333...» 93.6 .e wnmnoom Henewwmcoo n codumUAHm> nmouo .m A.<.m noaaeo nonommouoov madness pcosoonwe womanhon u ceavmpsnsoo coaumcaad> .N madnoom awaswawcoo : coavmvdmsoo soapmvdaw> .— eoeH Sea o8 883-20 condos". .< 5.33 no «.92 2.2: 3 583 “caresses oaeem 833 one «o 283 .HH 0000000000 00000 2000 0000000000 mpdqfioo mummmm-Hnu 00000 00.0 ‘0mm.0 000.0 .00000000 0000 00 001000 000 000000 000 .000 0000 0000 000.0 00.0 .0000: 0000 000-0000000 00 000 0000 00000 00 .000 00.0 000.0 000.0 000.0 .00000-00-000 .000 00.0 00.0 ‘00.0 00.0 .0.0-mm 000 x00 00 0000 0000-0000000 000 000mwm-0003 .000 i000.0 00.0 000.0 00.0 .00 00 0000 000 0000 00 000000 00000000000 000000-0000 -00.0 00.0 00.0 000.0 An-n- -000000 000 0000 0000 0000 000 0.00000 .000 000.: mmm.0 m0.0 mmm.: .00803 can» Eoummpm HmSXom whoa cm>00.mn cadocm cm: .00 000.0 000 .0 00.0 00.0 00.000235 00.00000 Q» mnflvmoa 000000 @005 vasonm vcanmsn m haco wmm 00.0 0%0 $0.0 _ 00.0.: .memmfluuwe .00 .0000000 Umoauompm up 002 casonm mmnsooumpca Hmsxmm .00 000.0 00.0 00.0 00.0 000000 000000000-00-000 000 000 0000-0000000 000 00003 .00 00.0 000.0 - 000000 000000 .0000000 0000000 000 mammm-mw-Nmmm-mmm-fluu 00.0 00.0 000.0 000.0 .00000 000000 00 000 000000 0000-000 0.000000w0 000 . 000.0 00.0 --00.0 00.0.00000000000 000xwm,0>00 00 .00 mm 00.0 -00.0 000.0 00.0 . .00000000 00000 0 1. you pcwpuomsfi 00 mHaddKom 02000 new 0» «H00 mcamm .3: 000.00 00.0 00.0 00.0 .0000000 000 0000 00000 . o>0pomuppm 0Hamsxmm mcamn 030nm mama 0000 amsoz .mm 0000 00.0 000.0 00.0 .000 00000 0000 0000 000 00000000 00000 0000 000000 00003 .00 0000 000.0 00.0 0000 .00000 0000 000000000 1,-1. 00 :02: tho mchmmHQ ma mwudooumpcw adsxmm .mm 00.0 30 0001.0 000.0 .0058 00 00000000000 Hmmmmm psozvfiz m000000 0000 now 0:000 .00 .rmmewrt mowsm- 000.0 00.0 .002009 0 mp :00 omudoupopC0 000000 mmfi>wm .0 0:000mamm 00.0 00000 000.0 0000 .000 000-000 0000 00 000000000 000 000: 00-000 .0 00000000 000.0 00.0 000.0 00.0 ---, .000000-0000 00000 000 0000 0000 000_0000 00 000000 0000-00-000000 .000 000000 00.0 00.0 00.0 .0000: 00 00000000 0000 0.000000 0 0>00 0000-0H.0000 0000-00 .0.0 . .00 0 .<.0 0000 00000 00000 000002 EmpH 0:0 0004 E “+2. 9.0 «mm; 910 2.0 .84: uza campus: can anon no 900 on» ma zvnos madaucmm .Nm N3 «5.0 an; awed . 892 no syncs mHHEdm wcmmm o» wmum Hmmm vdsonm mvcmomsm .mN 00.0 «0.0 00.0 .0.0 Nn.0 mm:.n‘ mmw.w «mm.m .wams ma macs on manawmo mm mQOH mm maafip mcazmg mmo 0H0: on «and 000m 0 m.pH ._~ mwocmcfim "0.0 mm.o 00.0 No.0 .:0803 zany who D Hmowpomu whoa mp0 nos mafiavcdm 00.0 mm.v wim;m mow.m .uonvo numm Bonn .3m mean mpmcm nflmgp mammmw ufidoo hwgp ma mmo nmavwn an “H503 Qua: ho Ucwpmsz < . m— mw.0 mm.4 00.0 Fm.— ‘ mmuo 00.0 mNm.m mN.— .gHAmL rimsos Scum . www.w mm.r m—m.dr wwN.ww‘ .054; mmum pamnp caudv wands» “sons xawp 00 find: pa “cam mo>a3 wed mvcmnmzm .m—_ HQ.“ mama max: mwmfl .55 you 2a 333 3 39$ mamas pcaom may on mamsoo m wpoahpmmp mmmflgud: .wOF 900.0 «find 3?: «EN .2833 593 ago: 053 63 «wuw no.0 00.0 m0.0 n0." 9.0 mmfifl m6. F mem;m‘ mom.N «00.0" «:0. N mp.0 mwm.m mm.” me.N . mmmmewwammmwlmlllllu oxdp vases bony ma voxwamp ouoe up 0H503 mamoom .mm —0.0 “WWW“ . quN.m 00 0 no.0 9.0 «9.0 a: m mmm.N F0.0 88 .323 8m 32° 3 £3 93% x53 «@5an .& nmm.P 00.0 no: 0 www— 3.0 9.. 8.0 flu.— a; o» aha: anon as» no one ma :oamw>oaov wawnopdz .n soapa¢uoom msopH uwnssz :oapmvaam> .mmouo coaumvumm>3 EopH wad moh< . . :omflfl. s. Mallou Imlgdmumoi ...I ll no.0 00.0 3.0 mN.0 ......" 0x5 H00.“ 0.:00 Sch 0023 00.000.300.090 on 00 3.5.0 03H .F 3; 3.0 9.0 3.0 .0 .23 5 00300.30 g 00:0 0.00 0 0.3a. . 00.0 0MN.q 00.0 N_.— .0090: 00.00005 0 :02» 0.85 5003 3.3m 03.0.3. 03% 0H .3 “mm 0: FFoo F000 3.0 0880: 5.5. :05 3 0.03009: 0.33 0“ 090.2003 5 00.30093 .3 @04“ . 90.0 00.“ no.0 60.03....00001. H0880 mafia: 0505.": 305300.30 ....00H :00 000900 4 .mm 00.0 08.0 00.0 No.0 .0 0.3003 5 00.300009”: H0380 3 003 0630034 . _. 000.0 00.0 000... #04 32.030 00 5.30030 no 30H 35: .3 0.33 no 00300.: 4 .N_. c0300.“: mo 00.0 5.0 “MWJN 3.0 8.0 0.0.2830 0.,” 500 000.330.0850 momN 00.0 03.: 3.0 .3085 393......“ 05 ago :93 02 80. 530.05. 23 0? 05 053009 0.3 .09 bod ...0 RN .— . 5.0 8.0 modd 0 .fi M2 0500000 0.3000. 0.0000 0H3”? 00>; 0:0 gmgm .nNF 1 9.0 3.0 00; 3.0 .0005 005 0x05 p. 500 0a 0023 000.0 030 030.0 300 0.9000 00020:: < . I 00.0 8.0 0.0.0 3.0 . 0005 05. 0000.0: 090:... 0mg 050 05000: 0.3. ion . or 8.0 6; 5.0 00.0 8. 533 08 0 58 8 0.8: 0.. 0H .00. 000.0. a; 000.: 0mm.N .0003 so: 0055 05. a 00. .0008 g 00:0 020: 3 0H wcfiom $03.85 0 30x 00 083 0009 05. .mm M0; 9.0 00.0 00.0 O«.O 30.0 fi0.0 Pm..o .wMMfihth CH 0mg Hdecmmmo Cd md. moggwfi HMOHVGZ .3 mnmé 5.0 2.0 0:3 .000; .6: 00% :05 320E 000d :0 03A 3 5.3 00. 000.30: 0H pH .N. $010 030 $.10 00.0 505 005 00530 00 0.35 000 0 350:0 00:03 .0 «0.0 00.0 m... o .06 300002 300.80 3.20:0 .005 .0035 0H 30:00 m 005.. . 03.: 9.0 m0. 0 ..Hé .0H 00 0 000.500 0>0H 0.3000 000 0009... M0 $93.. 4.. g No.0 I; 0%. 0 5.0 .9899: 05 3 0.000 0:0 0000.0ch 0.3.3 05 3 0.98.30. 2.0039900 Engam 000. no 0.000 .mm mcfluucmm .<.m H0050...» 0.010 REES 0800H 008.3: c030030> 0020 5300mm; 50.3 0:0 00.2 cosgsflroo 0L§Umufi£o 1hQ. 8 0 0.0.“ 00.0 9.0 .002 0803.. you hu0000000 00 909090 0 m0 009808 0>0900 c0 wca0m .mm wm.o Nm.o 00.0 03.0 .0003 0300 ca 00 m0 H1.JH009000 >MOH00m .om —o.o 090.: 00.0 m0.o .0090009 9009 099 09 00:0Hu0mmw amw me o m0.0 .Imm.o No.0 .0009 0H00HE 099 00x09 0MHH H0000.099.I0N OO O m0.0 NF.O O—.O ...lefioow: EOD 08m. mdflomm .{w m0 0 wwwo .mmno 0mm.N .H0000 0090090 000 208 add .N mm o mm.o Fo.— No.0 .0000000 50109H0 09 090m 950». v 00.0 3n." 00.0 m0.o .wmopz Song I. 99000 300x 900 000 5099 m0 0H0EH¢0 09 0H003 0:030m .mm 0¢M.N 0mw.o 09.0 MN.o .990009 09 900900 0Mfla 0H 0m0099 900990050 099 ..3 00.0 0MN.~ 00.0 xmd.o .auumw0x 0 .9099009 ms 50 H: .3 00 0 no.0 wN.o 00.0 .xuozn0909.w9 000900 000 000a :0 0mca99 0000 099 .NM 00 0 No.0 000.: 0FMdN .0009000 no 000200009 930 00099 0x05 0H0M0m 0N 0ma9 mo 00.0 00.0 0NM¢N aw.o II; .90990019 00090 00990009 0H0mm,9099 mafls0u 099 .m 090000Hfl9m 00.0 000; no.0 00.0 .0000: 00 00009005 099 .m:a>0H 0A 0M0: 90 0209009 0 0093 . 00.0 m0.0 mmwo 0nmwmr.r .00H90002H x00c.mwwmsoo 0093 . F 0qun .mw.o N000 00.0 .0030: :09» .1: rir 0905 :08 0909909 0m0auu0a ind 000900mm0 mo xoma <.le— 0mm.p Nm.0 —o.o 00.0 .005 0099 0030: 09 9:09000Ea 0908 0a 0w00090€ cw 009900mm< .09? mw.0 mm.o no.0 emao .90990009 00009006 0 0H0: 09 0>0H G099 0908 00x09 9H .09? 100.0 id 810 00.0 0000.005 you 0x0H9 009 090 00090H0000u0m 0H00 9903 0Hao0m .mo— Nw.— 0pm.m 00.0 3040 .0>0H 0000,3090 09 003 000m 0 0H 0ma: 90 0909009 A00» 09 mpmww mca>ao nmm mm.r 00.0 00.0 00.0 .9003 0099 9093 00m 00 00s 2099 0905 200900MM0 000 c0503 .00 Iiwm40 wwwqw “0.0 00.0.0H0930wM09w0 Hmm09 0 0050009 0003 90 0009009 0mmM0mmx .Nm 0Nm.w 00.01, 1M0.o 00.0 .09003 09 9093 000900.00990 000 r, 0>0w 09 0Hmo0a 039 0003909 900800900 :0 0a 0m009002 .mm £009009m< mo No.0 .rmm. 0 map 09.0 .0900Hw0a>090 00050009 0900i00H>0H 4 .mm 00009099000209 .<.m 910.900.4140 440 9.03.04.09.00 . 08094 909052 00090mMW0m Wmamo coawmumwmwiu E09H 000 0004 coflu0ummaoo 09010mtfl9u4 0 5 .4 1 III?!" :owpmwzafioo mQMSdmlflno 3.0! 3.0 EA 10.x .maaommwwflfl ofi I! pme pom ado: cmo Avnp mmdmoon sewage mum mUcmHum .mp— 3. _ wmflfi 0% P 9 .0 £338 imaaon mum sch.cm:3 sch Ham» Haas Ucodum Hmop < .NFF 00.0 No.0 mmm.# NN.O .=0CHS Mo nnmw.vflo: - ummm mxfln mUcmEMrUfiu B :Uumg ma .2“ mmmflums wimp? .mo— .wd.o 00.0 .wo.o Fo.o .qunmdn . mg» swdonnv mUme mum mUchum soc mmmaupme umpud .rm mo.o oo.o do. No.0 .mUcumwm ndQngpflz mamanopm. ommaupwa ndoh.uw>o xamp o» manm mp ansonw :0» .dm ammw_ .rmo.o No.0 .m_.o .xmoz a mono nHon may Spa: gso ow.UHsonmi Ucmnms: mge umm ow.o #0.? wmd.w mmm.ow .mUcmauw acme .mwnu pnwphoasa macs mum mUcoHum Uoom 03» yo 0:0 .om mmm.m oo.o mm.o mmw.m .cmeoz I. camp ems op pcwenoasa macs ma «Ucmfipm mmoao mcfl>mm .mm No.0 nw.o .o.o mo.oi|. .Ucowpu mep m ow.mcHnMHcm WMm :mo so» nmm om.o md.e Fwd" mmn.: .mmaz no Ucmnmss 950h1pmcammw II: co>o mUam use» mxmpl daaz or: mco ma Unmapm Hmmu 4 .m3 .2m.o mmo.w Fo.o mm.. .mm>a3 many mcomamm pamhmmmaU how nUmmem xoag.U:wpmsm .m: pp.o m_o.m. mm.o .1ww.o .mmoCflmsp 2:0 ufionp Unae mUCQHum pmon one ~wm waw;w No.0 oo.o ow.o .co UchmU :mo 90H.og3 mnomcom ma Ucoflnmy< .mm Ilwo.o mQUJWI, www.ml. m..o .mUcmwpm :20 pawn» mbmn Some UHsonw mafia Una Ucmpmsn 4 .FN 8.0 9.0 «Mm; No.0 .352.“ :8 $8.... 8. I; Una: pa mUcam mmwxmmso: m .mponnmamc scum anm< .ar ”w.o Po.o mv.o Fo.o .OU mUcmHgm HH03 30: psonm mxamp mafia no Ucwpmsn a can: mcahocam w.vH .m mUcoHLm no.0 8.0 8.0 8.0 .38... Sofia? B 8 RE 2. RH, .mfl L; mo.ml. mm.o oo.w Fo.o .sQN.:H:p ado pam3.somwaam Mdfl>wm .NM« mo.o wFN.w NP.o 3m.o .Ucflw op Uumg ma m>oa maps .mNF «8.: $6 3.0 «ma 1 53 fi 238.... 8:951»on «.38.. J: 83 go. 5.0 «2. .N 5.0 .2383 «853 8 ea: :9. 2w 8 v84.W 2 ..H .3... 23832.. o mmopo coaumUHHm smuH Una moh< II“ ”I 3.0 NN.0 no.0 w0.0 .nUcoflC Bahamas: mafia: mmmpm 3.90% mm: .3 Ucmnumw < .03? mm. «:3 $6 984 mom; 55880 .80.... 1. I pmmuou so» 3m: 58 ~85 3533 £393 0.8 mUcmflm .mm— «Mmé N30 mmwi mm .0 .mmwmflpms pom 2....me 0.3593 on 53 mUcmflm .UNF mUsmflm jfiufifisu .<.m Mahamflwuou mgpuww 900,552 2033-23 30.6 c3333, 5: USm 8.: cofimpsyfiwoo mamfidmaflno 147. APPENDIX B I. Validated and Cross Validated Items Classified According to Their Area of Origin A. Percentage of Retained Items Per Area .0343 009m :20. 05.90 0055. 0:80 033. 8. 0.8: .3 050 8...? 0:... 005.020 0 : .55 now 03.3 0.30.3 0.“ 0.35 00003 0500 009 0» 0amboo 0 mgoahumou 0m0ahh02 m0— 13% 0030 madam—0w 00.003 0:2.» Uoom < «or I .3' 05.3 Uoom 0.0000 3 A020: 03H 0m £05 0 .3.“ SC 0.00 m.“ as.“ 0.830000 8.0803 0 00:3 mm Javaod. .00 8.6 3.000 .8.“ .300 3 050 0.8% .22... @0330 .2 ..HNON NMO>0 .fldflll 08.3 0.30.3“ 0.“ 000:... 0.333 05.00 B... 3 03:8 a 3005.08 000.2002 02 .303..le h030 390m 0:008 05.3 Uoom 4 For 0% 0a....» Uoom 0 00.00 3 0020: 0....H 0w do: no 0 0 0.903 5 mm «a mango." 0.85 00 0.903 03000 2. &.mm 0.0.3 00H 1148. 0003000., 0x00 3.0000 003000 0' :05» 08.3 000m 0 020: 3 U90: 0.3” 0N 00.308000 0 .0me09: .00 so . 05.80 0.3 005 can: to 33 00 0.9000 micelfldmfioag H9 .3075 3 0009.9 0m0hzooc0 3. 0.0.0 0.003 ma pH mm dwfl 0.3900 0 5.5 3 .05 030.75 Q! .smaufiupofioa 0 9393b» 0230.. 3 0H 3 mm fiaflawwuofioe .3. .0355 «H 9.01.0. 02% a £303 8. 3300: 3 0H 3 031mm“.& 0000083841 05 u0>0 00x00 00g 033 :00 £0003” no .8003 0.330.“ U500 -205 05 .3: 20 0.3.303: 0.3 R. 8 00.5 H000 0.920 000330 mm 3 0.3 03 B o. .3 0.3 .8005 000000 0>0n 8. 0..” 0580 0.3.3000 00 mcom 00 a3 05 0002056 00002.2. 0 002. 8. 0.... 080 0E a :80 to 32 3 $3. 080 a 0.3 t 0:30:00 300.“ * 0:00 80.: m 0.000 00.2 33H 000 50H 000 0880 0500804 200800 U0uoom 0.00. 0002. m x52??? 1&9. .050H900Q¢ 1w. 000% pompom do» 0H0: :00 m0np 0000000 :000:0 0:0 00:0flpm mm— .00M0flr33000m 0— 000x05 0Hndoup 0Q :00 00:0aum mm? 00:0wum .nmanaano 0a :oapo0wm0 me n .0m0anums :« 00:000 m.m :0H000MH< m0 000a 00:03 0:: 0ww3 :0 0:0900: 4 N_. 3:0»:H Hmsx00 Op 0000a :0H000mm< or 0:00»0:00:080Q .00000 00 x00 _00 .cmuuaflno .0080: :0np E00000m 0>0: 00 0a x00 how :0000: 0:9 mm— dexmm 0:0: :0>Hm 0: nanosm :02 mm .00 pa :0:0 :29 0:08 .0mwfippme mo 00H0¢50.mwofiMNMMMIIIIII 0Q vasonm 00000000ucfi H0:X0m mo— 09 no: 0H00£0 0mpsoou0p:a H02K0m mm . .00:009 0 my .00u0500 .l m.om 0:0pra0m 0p :00 00::00:00:a Hmsxmm w:fl>0m m 0000005 Low hmawm 00 Edam m0m no 000EH0:H 000m -0 0000 000m. 0. 0000 0004 80.: :00 50.; :00 00:00m 0:0:00pm4 unmamupm 00:00m 0:00002 E 4.0. I. 150. APPERDIX C Analysis of Variance Tables and Computations A. B. ‘1; n Scale - All Subjects 5’ "I e Configural Scored “Issues Scale" 1. Validation - blaritamy adjusted group 2. Validation o Marimlly m ladjusted group 3- Cross Vail tion - I-iaritally adjusted group 4. Cross Validation - Maritally maladjusted group Straight Agreement Scored "Issues Scale." 1. Validation - Maritally adjus bed group 2. Validation - Martitally maladjusted group 3. Cross Validation - I~1aritally adjusted group 4. Cross Validation - Mantalh' raladjusted group A. B. 151. Analysis of Variance Tables and Computations Analysis of Variance Estimate of Reliability of ”F“ Scale for All Subjects W Source of Variance df Sum of Sguares Variance Ind}.vidu‘aflsflw E53 _— ~~ = p 2 z 1 h__,:;§ows) 931 DDR ,.u7021 s R .13183 Items C-1 55 = 4. 1 1 52 = .u (3.17(c-17* _ 2 _ , 30-1 _ , Total A939 ssT - 50.17021 N items = 10 r = s - s = . 02 6 N individuals = 9b tt _—§§Z;_—§' 5 7 Configural Scored "Issues Scale" 1. Analysis of Variance Estimate of Reliability of Configural Scored a”Issues Scale’3 for Validation Maritally'Adjusted Group v“— ‘ 30u§92,°f Variance ' df Sum of Sgnares Variance IndiViduals R-1 55 = 20.608 32 = . '6 (Rows) 22 ' R R 93 73 Items C-1 53 = 2.622 32 = .218 o _§coiumns) 12 C C 5 (R-1)(C-1) . 2 _ RC-1 - _ Total 228 °sr ‘ 71°839 N items = 13 = 2 2 = . . m individuals = 23 $2 a Mean = 7.79 SD = 3.97 152. 2. Analysis of Variance Estimate of Reliability of Configural Scored ”Issues Scale" for Validation Maritally Adjusted Group Source_of Variance df Sum of Sguares . Variance Individuals R-1 55 = 16. 8 52 = .808 1 (Rows) ' 21 R 9 3 R 7 Items c-1 as = 2, 11 52 = .20 2 .1colpmns) 13%c1; C 5 C 9 5 (R- - - 2 .. # Em)? 25,2. SSE — 27.33117 5 E — .1084? 30.1 Total 2§5_ SST = 1+633287 N items = 13 rtt = 32R — 32E = .8658? nmunmas=a —:?-— ‘ R Mean = 2.68 so = 3.49 3. Analysis of Variance Estimate of Reliability of Configural Scored "Issues Scale" for Cross Validation Maritally Adjusted Group Source of Variance df Sum of Squares Variance IndiViduals R-1 S" = 1 . 8 $2 = 1 . 6 16 Items C-1 2 = (R-1 0-1) 33 = .268 s = .12 46 Error 276 E 9 3 E 9 RC" 53 = Total 311 T 75'3987 N items = 13 rtt = $2 - 32E = .90516 N individuals = 24 -117;-- S R Mean = 7.33 SD = 1.58 —-b \11 K» 0 4. Analysis of Variance Estimate of Reliability of Configural Scored “Issue es Sc ale " for Cross Vali tion Maladjusted Group Source o§_Variance df Sum of Square§__ Variance IndiVidualS “-1 55 = 10.2 88 52 = .44691 (Rows) 23 R 7 5 11 Items -C-1 as = . 010 52 = .29751 AfiColunns) 12 C 3 57 R ' (R-1)(C-1) ' _ 2 _ 1 , 3301‘ 276 _ssE - 49.15105 3 E — .17803 RC-1 Total 411 331 = 63.00000 32 $2 Nitems= 13 rm: R" E = .6015 N individuals= 24 S2.q Mean = 2.63 SD = 2.48 C. Straight Agreement Scored "Issues Scale" 1. Ana13-isc cf Variance Es+ inate of Reliability of StraightrAgreement Scored ”Issues Scale" for V lidation maritally'Adjusted Group _A Source of Variance ._, df fi_ Sum of Squares, Variance fizégiggals RE;_F____SSR = 5.1826 SZR = .2355?} ' Ifzgzlumns) 0-1 SSC = ‘9‘0522 526 = '21329? Error (R-;3:C 17”SSE = 39'“78 SEE = CQ“”“‘ Total Rig; SST =_&?'1526 N i+3ns = 20 rtt = 32R - 52E = .5991 N individuals = 23 —s—2—— R Mean = 16.83 SD = 5.72 154. 2. Analysis of Variance S..ti..1ate cf Feliabi lity cf Straight agreement S3r red ” issues Scale” for Valida non Maritally Vaiaijusted Group fiSource of Variance qgfi Sum of Squares Variance Individuals R—1 cs = . 1+ 52 = . (Rows) 21 a R 9 7 773 R .464178 Items C-1 _ 2 - (3-1)(0-1) = 2 - Error 399 SSE 83°10227 S E - .208276 RC-1 _ Total 439 SST ~109-99773 N items = 20 2 ‘ r 3 S - S = 11 individuals = 22 tt .17.}. 5513 5 Mean = 9 95 R SD = 2. 98 3. Analysis of Variance Estimate of Reliability of Straight Agreement Scored ”Issues Scale” for Cross Validation Maritally Adjusted Group source of Variance df Sum-of Squares Variance Individuals R-1 .2 SS = 4. 8 S = .1 6 (Rows) 23 R “9 R 955 5 Items C- SS = 4.0 52 = .212614 (Columns)__ 12_ C 397 C Error (R-;%§C-1) SSE = 44.7103 323 = .1023119 R - Total :7; SST = 53.248 N items = 20 r = 2 2 _ L A N individuals = 24 tt 3 R .E_§, ‘ ~%?63J9 52 Mean 2 17.45 R SD = 2.00 4. 155. knalysis of Variance Estimate of Reliability of Straighthgreement Scored “leanes Seal “#— e" for Cross Validation Maritally Maladjusted Group _Source of Variance df Sum of Sgpares A7Variance Individuals R-1 55 = 21. 0266 52 = . 48 8 Items 0" $3 = 20.0 66 32 =1.0 1 Lgplumns) 19 c 77 a 567 (3-1)(0-1) _ 2 _ Error 43? SSE - 90.37234 5 E - .206802 30.1 _ Total #29 SST -131.9521 N items = 20 _ 2 2 _ ‘ N individuals a 20 rtt ‘ 5 R - 5 a ‘ '77879 2 Mean = 11.29 5 R SD = 2.36 -‘,':a_ ~ , $7.le O 5‘? I ‘g: to! !?5‘\ Pr “($13, \'I .' X 'I .. 1 r‘ W “t“ ...... W-k5he~w " 5" 4‘1“?" ‘ ’ 1:1 '1- :3" 1‘“. .-‘-'-.-. , . I ‘. . 11‘ .. .- u - \— u'k) . - 5M“ ?//é¢, ‘1/ 7 r '1, ”’9 NR - 3 .fl J‘tfifi-M 1954'? 0W4 3 1L MICIIWIWININTITIHTHIEWIUIWIIVHEIMIW||1|W|Vs 3 1ll3 03082 6899