AN INVESTIGATION 'OF ATTITUDES 0F PARENTS TOWARDS THEMSELVES, THEIR MATES, AND THEIR CHILDREN Thesis for the Degree of M. A. MICHIGAN STATE UNIVERSITY JANICE 'OPPENHEIM FISHER 1972 LIBRAR Y ' Michigan State ABSTRAC T ATTITUDES OF PARENTS TOWARDS THEMSELVES, THEIR MATES, AND THEIR CHILDREN By Janice Oppenheim Fisher The main theoretical position of this thesis is that parental attitudes are the result of more basic needs and motives which encompass the larger family constellation. It was pre- dicted that because of the influence of these basic needs striving towards gratification, parents' attitudes towards themselves and their families are interrelated. Specifically, it was hypothesized that self satisfaction, marital satisfaction, and wishes for one' 3 children to possess characteristics similar to their parents are all interrelated; that parents with positive self images will tend to have satisfactory marital relationships; that these parents will approve of their children's identification with themselves or their mates. Con- versely, when self image is negative, marital satisfaction will Janice Oppenheim Fisher be low. In such situations, parents will want their children to develop characteristics similar to the ideal self or ideal mate, depending on the sex of the child. Thus, these parents will use their children to satisfy needs ordinarily filled by another person. It was also theorized that children fill more pervasive needs. for mothers than they do for fathers. Therefore, it was hypothesized that all indices of interrelationships of parental attitudes will be stronger for mothers than for fathers. A sample of 25 couples living in the Michigan State University married student housing facility was used to test these predictions. Semantic Differential descriptions which included three factors of connotative meaning were used to obtain ratings of actual and ideal family members. Discrepancy scores were tallied between certain of these ratings taken two at a time and these D scores compared by means of product-moment corre - lations. Correlation matrices among the three factors making up the ratings were developed for each hypothesis. The results partly supported the hypotheses in that evidence for partial interrelationship of attitudes was discovered. Specifically, positive correlations were found between self satisfaction and marital satisfaction. Positive relationships also were found between self satisfaction and the desire that the same Janice Oppenheim Fisher sex child resemble the actual self, and between marital satis- faction and the desire that the opposite sex child resemble the valued mate. Ideal self and ideal mate were not used as models. Contrary to expectation, mothers' data did not produce higher correlations than fathers' data. AN INVESTIGATION OF ATTITUDES OF PARENTS TOWARDS THEMSELVES, THEIR MATES, AND THEIR CHILDREN By Janice Oppenheim Fisher A THE SIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1972 ACKNOWLEDGEMENTS For helping me to plan and complete this thesis I would like to thank the members of my committee, Drs. Albert Rabin, Bertrum Karon, and Robert Zucker. I am especially grateful to Dr. Rabin for his con- sistent patience, encouragement, and support. In addition, I am happy to acknowledge the assistance of Dr. Arthur Blue, of the University of Saskatchewan, in analyzing some of the results, and the invaluable assistance of Robert L. Rodensky, of the University of Western Ontario, in obtaining and interpreting the computer -developed treat- ment of the data. ii TABLE OF CONTENTS LIST OF TABLES . INTRODUCTION Statement of Purpose Theoretical Concerns Review of the Literature . A Psychodynamic Approach to Parental Attitudes . Summary. PROBLEM AND HYPOTHESES . Problem . Hypotheses . METHOD. Description of Instrument Subjects Procedure Prediction in Terms of Instrument Employed. STATISTICAL TREATMENT OF DATA RESULTS. DISCUSSION OF RESULTS . The Findings . Theory and Design Sample Directions for Further Research SUMMARY . BIBLIOGRAPHY iii Page 19 19 22 24 24 27 30 44 44 49 54 56 61 64 Page APPENDICES 70 A. Directions for Administration of Instrument 70 7 1 B. Sample Semantic Differential Rating Form . iv Table LIST OF TABLES List of Adjective Pairs Included in the Semantic Differential . Sample Variables Coding of Variables Mean D Scores for Variables Coefficients of Correlation Between Variables Coefficients of Correlation Among Separate Factors of Variables . Comparison of Coefficients of Correlation for Mothers and Fathers . Comparison of Coefficients of Correlation among Separate Factors for Mothers and Fathers Page 21 23 29 31 32 33 34 35 IN TRODUC TION Statement of Purpose The problem of how parental adjustment determines attitudes towards children is one that has received much attention in the clinical literature. It has long been agreed that parents' attitudes greatly influence child adjustment. That these attitudes reflect the parents' own personal adjustment and that this adjust- ment in turn influences these attitudes towards other family members has not been as thoroughly explored. The present paper will examine this problem in depth. A review of the relevant literature will explore work in the area of parental attitudes and more basic motives encompassing the larger family constellation. Based on conclusions drawn from this review, a theo- retical position will be proposed and assessed by means of a correlational study . Theoretical Concerns Attitudes as a topic of scientific research came into prominence in the 1930' s with Thurstone' s pioneering work (Thurstone, 1936) and has since become firmly entrenched as an integral area of study in both social and clinical psychology al- though the topic continues to be fraught with problems. Research has taken several directions such as investigation into the function and structure of attitudes, investigation of those attitudes which affect the quality of life, and investigation of the problems of attitude measurement. For example, opinions, which can be measured, do not lead to the reliable prediction of behavior. Attitudes are not entirely conscious, but are probably dependent on more covert motives which are often not measured (Allport, 1953; Katz, 1960; Rabin, 1965). Even when attitudes are largely conscious, respondents can choose not to answer items honestly. All these problems are present in attitude study in general and in the con- sideration of the particular topic under investigation. Attitudes of parents toward children and the effects of these attitudes on child development have been especially difficult toresearch because in addition to the above problems, these attitudes are subject to change, difficult to isolate and define, and have varying effects on the child. Further theoretical compli- cation stems from the fact that agreement with certain state- ments may serve a psychodynamic function. For instance, respondents will often agree with contradictory attitudes if both statements seem to express expert opinion (Stogdill, 1936). Most parental attitude studies have not viewed the attitudes as an expression of needs, but as a variable independent from other facets of the personality. According to Katz (1960) however, an attitude cannot be understood without consideration of the function it performs for the individual. Despite these limitations, much speculation and research has focussed around the nature of the influence of parental attitudes upon child adjustment. Most theories of child development place great emphasis on the psychological environment generated by attitudes which the parents provide for the child. A prevalent thesis is that healthy parental attitudes pro— duce healthy children, while malevolent attitudes inhibit and dis- tort psychological growth. Presumably, the attitude is somehow expressed through behavior, but how this occurs is not clear. It has not been demonstrated either in theory or in empirical re- search. In fact, many studies confuse the two concepts of attitude and behavior, treating them as a single construct. This approach has not produced a clear understanding of the factors operating in the situations under investigation and led Shaefer and Bell (1958) to conclude that results of parental attitude studies have been inconclusive. Instead of focussing on attitudes per se, a more fruitful approach has concentrated on parental adjustment as the prime factor in first determining these parental attitudes which then in turn somehow affect child adjustment. Results using this approach are more consistent, but still cannot explain why different attitudes and behavior are associated with similar levels of adjustment. The most promising approach recognizes parental attitudes to be composed of interrelated feelings and predisposi- tions which are expressions of psychodynamic needs. It is assumed that the deep underlying motivation, rather than its surface expression as an attitude determines the healthy or path- ogenic quality of a parent-child interaction. When the interde- pendence of attitudes and their origins in more basic motives are considered, a larger picture of familial interrelationships can emerge. An understanding of these interrelationships is necessary to appreciate the dynamics operating in a family, for it is this set of dynamics which produces the attitudes parents have toward their children. It is the thesis of this paper that instead of concentrating on individual behavior or attitudes per se, research must focus on a larger network of parents' needs for personal and marital satisfaction in order to fully comprehend the function of attitudes and expectations which operate in the typical family constellation. Only when such motivational structures underlying parental attitudes are made clear can the long range goals of need satis- faction towards which action is directed be completely understood. The following review of the literature will examine these issues more closely and will provide support for the latter theoretical position. Review of the Literature Concentrating on the attitude itself rather than on its function as a need -serving vehicle, several instruments have been developed for the purpose of measuring parental attitudes (Leton, 1958; Lovinger and Sweet, 1961; Shaefer and Bell, 1958; Shoben, 1949). One of these, a questionnaire called the Parent Attitude Research Instrument (PARI), was constructed by Shaefer and Bell (1958). Since its publication, the PARI has been used extensively in attitude research. Factor analysis of data from the instrument has yielded two dimensions: 1) love -hostility, and 2) autonomy- control. These two dimensions are similar to those found in several other investigations (Brody, 1956; Krung, 1965; Levy, 1948; Sears et a1. , 1957). The PARI has become the preferred instrument in that it is used most frequently for research relating to parental attitudes and child adjustment. Studies using the PARI often analyze results of mothers of poorly adjusted children and compare them to a control group of mothers whose children are judged to be emotionally stable. Many studies using this paradigm find parents of maladjusted children to score higher on the hostility and control dimensions that controls (Madoff, 1950; Shaefer, 1959, Shaefer and Bell, 1957). However, other studies have shown in- conclusive or nonsignificant results (Brody, 1963; Zunich, 1966). This inconclusiveness can partly be explained by the finding that scores on the PARI are correlated with education of respondents (Shaefer and Bell, 1958). The better educated mothers often report healthier child -rearing attitudes while the less well- educated mothers show more authoritarian, suppressive, and hostile attitudes. Possibly the better educated mothers recognize and check the socially approved answers. Other studies, using a variety of instruments exhibit a diversity of findings. Horowitz and Lovell (1960) found mothers of schizophrenics less approving of hostility in children and more approving of maternal control than mothers of normal subjects. However, contradictory results were obtained by Guertin (1961) who reported significant results but in the reverse direction, while Zunich (1966) reported that his experimental group could not be differentiated from normals on those scales of the PARI. Using a variety of techniques, Mark (1953), Ross (1955), and Freeman et a1. , (1959) all found mothers of schizophrenics and other diagnostic groups to have restrictive, possessive attitudes towards children and to demonstrate both devotion and detachment. Other studies investigating parental attitudes and behavior as perceived by the children themselves have shown the most consistent results in the area. Poorly adjpsted children view their parents as having malevolent, arbitrary, and rejecting attitudes, while well —adjusted children see their parents as generally benevolent (Ausubel et a1. , 1954; Roe and Siegelman, 1963; Siegelman, 1965; and Williams, 1957). .Heilbrun (1960) found that the PARI, although it was developed originally for parents, could differentiate between normals and schizophrenics when the subjects answered the questionnaire the way they thought their mothers would. Schizophrenics in this study indicated their mOthers possessed authoritarian attitudes. Other experiments found mothers of schizophrenics to be described by their offspring as controlling, demanding, engulfing, and subtly rejecting (Kohn and Clausen, 1956; Lane and Singer, 1959; Lu, 1961). The rationale for this method of using children as infor- mants of their parents' attitudes is made explicit by Ausubel et a1. , (1954). They argued that children' s perceptions of parental be- havior and attitudes can be measured more validly than these latter phenomena themselves. Children have less reason or ability to falsify reports of their parents' attitudes and behavior than their parents and so are more honest respondents. Considering all these studies as a group, the one main conclusion to be made is that there are definite indications that parental attitudes constitute an important variable which affects the child' s psychological adjustment. Unhealthy parental attitudes are mirrored in maladjustment of the child. Very little in this re« search, however, contributes to an understanding of the attitudes themselves. How, for instance, do attitudes exert their influence? Presumably, they are somehow expressed through behavior, but none of the studies have concentrated on this point. What are the factors which determine whether a parents' attitudes towards his children will be healthy or malevolent? Are particular attitudes towards one member of the family influenced by attitudes towards other family members? In order to answer these questions, a broader concept- ualization of the parental attitude as part of a complex persona- lity process must be employed. An unconscious defensive function must be recognized (Allport, 1953; Katz, 1960; Rabin, 1959) and instruments employed which are capable of tapping these dynamic aspects of parental attitudes and opinions. A Psychodynamic Approach to Parental Attitudes When the interdependence of attitudes and their origins in more basic motives are considered, a larger picture of familial interrelationships emerges. Many separate research studies support the conclusion that parental attitudes are directly related to such diversified areas as self image, view of others, sexual happiness, and marital compatibility (Ellis, 1954; Helper, 1955; Parsons andBales, 1955; Sears et al. , 1957; Zuckerman and Oltean, 1959). It may be that all parents use their children to fulfill certain needs (Freedman, 1960). Whether or not the outcome is pathological depends on the nature of these needs. Freedman' s study concluded that all parents use their children to relive past 10 experiences and vicariously enjoy new experiences. Carter (1967) felt it is only when parents focus primarily on their own needs at the expense of the child' s that pathology results. Meyer and Karon (1967) found that when mothers consistently gratified their own wishes in preference to those of their children, serious child pathology resulted. More specifically, it has been shown that when parents use their children to fulfill needs ordinarily gratified by the marital partner, the children react adversely (Ackerman,. 1958; Ellis, 1954). Parental adjustment as a variable in attitudes towards children has received separate attention. Personality theorists havelong suspected that in order to accept other people, an individual must possess a reasonably good self concept and be reasonably well —adjusted (Adler, 1921; Arieti, 1955; Fromm, 1939; Horney, 1937 Rogers, 1949). Research into the process of psychotherapy has shown that during successful treatment, the discrepancy between a person' 3 ratings of his actual and ideal self decreases, while his acceptance and tolerance of others increases (Berger, 1952; Sheerer, 1949; Stock, 1949). These results have been applied to the field of family relationships. Well —adjusted people have reasonably good self concepts and are more accepting of their families. Conversely, poorly adjusted individuals show a wide discrepancy between 11 actual self and ideal self and are also intolerant of the other members of their families. Zuckerman et al. (1956) found that hospitalized patients were less approving of their parents than a group of normals. Using mothers as subjects, Zuckerman and Oltean (1959) found that self-acceptance correlated negatively with the hostility —rejection factor of the PARI. Similarly, Block (1955) found that fathers who were constricted, submissive, and had little self -assurance expressed restrictive punitive attitudes, while well —adjusted fathers were more permissive and accepting in their attitudes toward their children. Instead of concentrating on individual adjustment, many workers have focussed on the marital relationship as an important factor influencing the way children are perceived and treated. According to Symonds (1949), a child cannot be fully accepted by his parents if their marriage is not stable. Parents often displace negative feelings towards the mate onto the child, or the child is blamed for the marriage failure. Ackerman (1959) says much the same thing, but within the framework of role analysis. He states that the relations of the parents to the child must consist- ently be weighed within the framework of the relationships of the mother and father in roles other than the parental one, and (p. 181) that the parents will love the child in the same measure 12 as they love each other. Several other theorists (Ellis, 1954; Freedman, 1960; Hamilton, 1950) agree that parents who have poor sexual relations will use their children to fulfill these needs in that their behavior towards their opposite sex children will often be seductive and conflict producing. One of the earliest studies to show the poor psycho- logical adjustment of children whose parents are unhappily married is that of Baruch and Wilcox (1944) who found this correlation in preschoolers. Later,-Krung (1965), using a sample of parents of children with behavior problems, reported similar results. These parents, besides having rejecting, hostile, and authoritarian attitudes on the Parental Attitude and Behavior Checklist, were also characterized as having a negative attitude toward their spouses. Another interesting approach to the study of detrimental effects of interparental tension concentrates on the disturbance created .in the sexual identification of the children involved. Whenthe parent is dissatisfied with the spouse, he or she will prevent the child from identifying with that spouse. This pro— cess has been suspected in theory, especially as applied to the dissatisfied mother preventing her son from identifying with his father. This view holds that the little boy cannot identify with 13 his father unless Mother permits him to do so by her encourage— ment (Adler, 1921; Ansbacher and Ansbacher, 1958). The same viewpoint has also been interpreted according to strict reinforce - ment theory (Helper, 1955; Sanford, 1955). If the little boy is negatively reinforced by Mother for acting like Father, he will soon learn not to do it. Helper (1955), using the Semantic Differential to measure likeness to the father, found that, ”Self concept modeling after Father is positively related to its reward by Mother. " Baxter et al. (1964) found similar results using a forced choice test on which adolescent boys compared themselves with their fathers. Of interest is that this process works differently with girls than with boys. All findings agree that mothers have more influence in affecting the masculine identification of their sons than fathers have on the feminine identification of their daughters (Hamilton, 1950; Helper, 1955; Parsons and Bales, 1955). Thus when Mother is not pleased with her husband, the effects of her son' 3 development are more pronounced, while there will be less disturbance to the daughter when the father is dissatisfied with his wife. 14 Summary As indicated by the precedingreview, the literature leaves little doubt about the importance of attitudes in affecting the child' s psychological development and adjustment. There is evidence that unhealthy parental attitudes, through their mani» festation in behavior lead to, or even cause, deviant development of the child. Thus parents whose attitudes about themselves and their spouses are negative, act in such a way as to induce mal- adjustment in their children. There is some evidence that in these homes where parents are dissatisfied with themselves and their mates, the children fulfill certain needs and are viewed with attitudes which are different from those children in families where parents' opinions of self and spouse are more favorable. PROB LEM AND HYPOTHE SE S Problem The first problem in investigating the interrelationship of parental attitudes is obtaining a valid index of these attitudes which can be used to elucidate these interrelationships. The next step in this problem is to analyze this information in a way which will lead to understanding how behavior follows from these attitudes, needs, and expections, From the preceding review, it appears that the attitudes of parents are influential factors in the child' s psychological environment. These attitudes affect the child' s emotional development in either a healthy or unhealthy direction. Because of this important relationship, the different attitudes, needs, and expectations of parents regarding themselves, their mates, and their children should be investigated. The literature already contains many empirical studies demonstrating that parental attitudes and children' s behavior are directly related (Gildea et al. , 1961; Shaefer, 1959; Shoben, 1944). However, there is less work in the area of the interaction of 15 16 attitudes, needs, and expectations themselves. It is this type of strictly attitudinal research which is necessary to help dis- cover the reasons behind the known relationship of attitudes and behavior. Only when the motivational structure of parental behavior is made clear can the long range goals of need satis- faction towards which action is directed be completely under- stood. In this way, also, will more understanding be gained about family living, parent -child interaction, and human motivation in general. Hypotheses It has been suggested that the attitudes of parents toward themselves, their marriages, and their children are interrelated and that the explication of these interrelationships will allow one to understand the motivational forces behind the behavior of parents toward their children. This thesis will attempt to make more explicit the inter- relationships among attitudes of parents. From the previous review and theoretical orientation the following hypotheses were formula ted: 17 1. Self satisfaction of parents correlates positively with marital satisfaction. It is argued that when a person is relatively satisfied with himself as an individual, he will be able to accept and like another person. Conversely, when an individual is dissatisfied with himself, he will not be as able to accept another person, especially in a close interpersonal relationship such as marriage. 2. Self satisfaction of parents is negatively correlated with the Wish for the ideal same sex child to resemble the ideal self. When a parent's self esteem is high, he will be suffi- ciently free from conflict to allow his children to develop along directions which do not have to satisfy his unfulfilled narcissistic wishes. When, on the other hand, a parent feels himself to be far below his ego ideal, he will have a greater need for his children of the same sex to become as close as possible to his own ego ideal. In this way he can achieve self satisfaction ‘2' vicariously, through his children. 3. Marital satisfaction correlates negatively with the degree of congruence between ideal opposite sex child and ideal spouse. When marital happiness is low, the parent will use the child of the opposite sex to satisfy needs ordinarily filled by the spouse. On the other hand, when satisfaction with the marriage partner is high, the parent will not need the child to be like the ideal spouse since the actual partner fills that role. - 4. Marital satisfaction correlates positively with the dfiree of congruence between ideal opposite sex child and actual spouse. When partners in a marriage are satisfied with each other, they will each want their opposite sex child to identify with the spouse. Husbands and wives who are not satisfied with 18 their spouses will not want their opposite sex child to resemble the disliked partner. 5. There is a positive relationship between marital satisfaction and agreement of parents' ratings of their ideal children. When marital satisfaction of parents is high, it is hypo- thesized that their goals for their children will be in agreement. Conversely, when marital satisfaction is low, parents will more likely have differing expectations of their children. 6. Marital satisfaction and self satisfaction of mothers as opposed to fathers are more highly correlated to other measures involving their ratings of their children. In this culture, mothers are more involved with their children' s personality development than fathers. Mothers have clearer and more specific ideas of how they want their children to develop. Also, children fill more pervasive needs for mothers than they do for fathers. Therefore, it is reasoned that mothers' ratings of their children will correlate more closely to other measures than will ratings by fathers. ME THOD Description of Instrument Semantic Differential Although the Semantic Differential was initially developed by Osgood (1952) to measure the meaning of concepts, it has been successfully used in personality‘research (Osgood et al. , 1957). The procedure consists of rating a set of concepts on aseries of bi-polar adjectival scales separated by seven spaces. The subject checks the location on the scale which he thinks is closest to the meaning of the particular concept. If he thinks the scale is ir- relevant, he checks the middle Space. Osgoodet al. (1957) and Rabin (1959) suggest selecting appropriate scales on the basis of face validity to measure particular concepts of interest. Semantic Differential data have been factor analyzed consistently yielding three major factors: 1) Activity, 2) Eval- uation, and 3) Potency. It is assumed that these three factors are an integral part of the connotative meaning of all attitudes, with Evaluation being foremost. The inclusion of several scales 19 20 representing each factor is necessary to insure the valid des- scription of an attitude. Ratings may be obtained from the same subject concerning his perception of various aspects of his own and others' per- sonalities and then correlated to determine a measure of similarity among these perceptions (Cronback and Gleser, 1953). Ratings of the same concepts by different subjects can be compared for overall similarity on each of the three aforementioned factors (Rabin, 1959). In this project, a list of twenty pairs of opposite adjectives were selected (See Table 1) on the basis of face validity from the suggested lists printed in Osgood et al. (1957) for inclusion in this study. In order to insure that the attitudes to be measured would be conceptually valid, adjective pairs were included to represent independently, the factors isolated by Osgood et al. (1957). Ten pairs had high loadings on the evaluative factor, five pairs had high loadings on potency, and four with high load- ings on the activity dimension. They are listed in Table 1. One pair, young -old is high on both activity and evaluation and is listed twice. The pairs of adjectives were arranged in random order on rating sheets with a space at the top of each page for the concept to be rated (See Appendix B). 21 EEBaSGASBBO Hosanna? “mosom noon- scam firmware? eardrum 30: 950% com: human 30.. munch £008? nwsom “Gammoflgu 3mm «3% 33m- “man.— 333:? doom wouflwwm- SRO cocgouu am~=wn< «mow- 9.84 Eh? :NSO ofimmma- 953w Mama? wsohm Ems- Emfismom :5- 82m :8. Ram can- ease grosses. 5333‘. 13:8 0% sofldgflwm souowm REEZSEQ ongfiom 65 9.1 €335 33m. 33334 .8 33 H 3nt 22 Each subject filled out the checklists to describe the characteristics of his actual self ideal self actual spouse ideal spouse ideal son ideal daughter. Biographical Information In addition to the Semantic Differential descriptions. the following information about each subject was obtained. years of marriage number of children sex of children age of children religion. Subjects The subjects used in this study were a group of twenty- five married couples living in the married student housing facilities on the campus of Michigan State University. Each couple had at least one child or were expecting one shortly. All the husbands were students. Nineteen of the couples were drawn from a list of people who had participated the previous year in a personality research project. Six couples were included who were not on the list. Their names were drawn at random from the population of the same housing facility. Subject variables are shown in Table 2. 23 2 ofiwaom schemes m 332 we Now m 232 H £333. mm scamBoE coamnom m oflofimu Z 033.3 > n .v 2. .3 - s .c c835 .3 was. a . m u o .o :oREEO mo Roach—Z o .m 3 .. m .o omfiuawz m0 930% 7 smog omsmm Basic > 8m. ... zv SEARS, «Assam N 3nt 24 Procedure The subjects (88) were contacted by telephone, and an appointment time was arranged when both husband and wife would be present. At the prearranged time, the examiner (E) called at the 85' apartment. Ss were informed that the experiment con- cerned attitudes people have about their families. Ss were given instructions explaining the rating procedure. The Ss sat in different areas of the room, worked separately, and did. not confer with each other. As a precaution against any order effects in the ratings, Ss rated the concepts in random order and were given the checklists one at a time. After one checklist was completed, it was removed and another handed to the S. Predictions in Terms of Instrument Employed To test the hypotheses outlined, the following specific predictions were formulated in terms of the scores obtained with the Semantic Differential. Each corresponds with the pre- viously stated hypothesis with the same number. The variable of marital satisfactionis operationally defined as the discrepancy between ratings of ideal spouse and 25 actual spouse. The variable of self satisfaction-is operationally defined as the discrepancy between ratings of actual self and ideal self. 1. Self satisfaction of parents correlates positively with marital satisfaction. 2. Self satisfaction is negatively correlated with the degree of congruence between ideal same sex child and ideal self. 3. Marital satisfaction correlates negatively with the degree of congruence between ratings of ideal opposite sex child and ideal spouse. 4. Marital satisfaction correlates positively with the degree of congruence between ratings of ideal opposite sex child and actual spouse. 5. There is a positive relationship between marital satisfaction and agreement of parents' ratings of their ideal children. a. Marital satisfaction correlates positively with the degree of congruence between ratings of ideal son as rated by both parents . 26 b. Marital satisfaction correlates positively with the degree of congruence between ratings of ideal daughter as rated by both parents. 6. Marital satisfaction and self satisfaction of mothers as opposed to fathers are more highly correlated to other measures involving ratings of their children. Hypotheses 2, 3, and 4 will yield signi- ficantly higher correlations when ratings of mothers are compared to ratings of fathers. STATISTICAL TREATMENT OF DATA Discrepancy scores were obtained betweenratings to compare them for similarity. This method is recommended by Osgood and Suci (1952) to compare the semantic distance between two concepts. The bi-polar adjectives are separated by seven intervals, and a difference score is obtained by counting the number of spaces which separate the ratings of the two concepts on each individual scale. These individual discrepancies or dif- ferences are the summed to obtain the total discrepancy score (D score) for each concept. Mean D scores were thencalculated across subjects for each variable. The D scores were correlated to examine the relationships between the attitudes under investigation. Product-moment (r) correlations were computed between each pair of variables to be compared. For Hypotheses one through five, data for mothers (n = 25) and fathers (n = 25) were combined (N = 50). Eachr was tested for significance by using the formula t = ' r tests were one tailed and df = 48 for N = 50 and df = 23 for n = 25. 27 28 To test Hypothesis six, ratings of mothers and fathers were divided, andseparate correlations were obtained for the variables under consideration, using the formula for t given above. To test whether the correlations differedwfor mothers and fathers. the Subjects. were treated as samples representing two independent normally distributed populations. The correlation coefficients were transformed to z scores and the standard deviation of this difference obtained using the formula: All tests were one -tailed. To determine the interrelationships of the three factors of connotative meaning of eachvariable, Activity, . Potency, and Evaluation, correlation matrices betweenthe factors of each variable were developedfor all hypotheses and the diagonal of the matrices. evaluated for significance using the formulas given above. In addition, mean D scores for each of the factors (A, .E, P) of the several variables under analysis were reported.(See Table 4). Variables were coded (See Table 3) for efficiency of reporting. 29 uncommon: <1 flow ~30flo< moaoom Q osflsm> 353800 non—ooh ha monoom Q La 8m u 75 mofinwwam> pom meadow D smog v gash. 32 Table 5 Coefficients of Correlation Between Variables (N = 50) Correlation Hypothesis Variables Coefficient (r) 1 A vs B . 31* 2 A vs C -. 02 3 B vs D . 15 4 B vs E . 87**** 5 (a) B vs F . 012 5 (a) B vs G . 17 7 a A vs H .81 a a posteriori hypothesis * p<.05 *Mak p<,001 33 Table 6 Coefficients of Correlation Among Separate Factors of Variables (N = 50) Correlations (r) of Factors Hypothesis Variables ctivity Potency Evaluation 1 A vs B . 34 . 17 . 40* 2 AvsC -.21 .45b .00 3 B vs D . 17 -. 02 . 08 4 B VS E , 70**** . 61*:9: . 82**** 5 (a) B .VSL-F -. 13 -. 02 . 09 5 (b) B vs G -. 12 -. 31 b -.06 7 a A vs H .74**** .. 75**** .80**** a a posteriori hypothesis b All tests were one tailed 2 tests for predicted directions only. * p<.05 ** p<.01 *** p<.005 **** p<.001 Comparison of Coefficients of Correlation 34 Table 7 for Mothers and Fathers (n = 25) Significance Hypothesis Variables Mothers Fathers of Difference 1 A vs B .29 . 46** NS 6 (2) A vs C . l3 -. 02 NS 6 (3) B vs D -. 10 .27 'NS 6 (4) B vs E . 91**** . 88**** “NS 7 a A vs H .84**** .86**** NS a ** p <.01 **** r><:.001 a posteriori hypothesis 35 moo.V a $3 fio.V 9%." 8.Vn* .sfino meow—coho con—Buoy.“ no.“ 3mm» ,N “5:3 28 one? 3mm“ 24» a mwmofiogs EoEoumom .m w b m2 3&3 . "wimp . n wz ***mm . *snvm . 94 33."? . “wimp . m m> as w b m2 "wing . “gig . mz *ib . “.33.": . mz 333m . “Swim . m m> m G; o mz mo. mo. m2 5.7 . 3.. m2 *3. mo; D m> m A8 m m2 8. 2 .- n mz cm. :3. m2 3 .- 2. u 3 < AS a 2 m2 *3 . mm. m2 3 . 3. m2 3 . mo . m m> < H moose muwfimhv 29:32 moose mumfidm mama—02 _ moose mumfimm 20502 - ammun— - .8me u .5th coflgfiwspm monoaom 3:504 mofinflunfiw mflmofioamm muouomm Ho munowoflmoou :ofimfionnoU ‘___JL__ A 3N n 5 mum—bdrm was mama—3.2 no.“ whoaowm ouwuwamm macaw coESmuhoU we madmwofluoov mo noflumafiou n Bass 36 r = . 46, significant beyond the . 05 level of confidence. ,However, for wives,“ however, the correlation bettveen self and marital satisfac-tion‘was r = . 29. This result is not significant, but in the predicted direction. Therewas no significant difference between these correlations. See Table 7. Looking at the individual “Ti factors, E, P, and A, it appears that when mothers and fathers I. T are considered separately, it is E which is most sensitive to variations in these variables, with r = . 46, beyond p< . 05. All L" correlations among other factors are insignificant, with r = . 25 (Evaluation) for mothers again approaching significance. ‘Hypothesis 1 then remains. largely confirmed. Fathers who are satisfied with themselves tend to be satisfied with their mates, and conversely, fathers who have poor self concepts tend to be dissatisfied with their wives. The difference between husbands and wives, however, was not great enough to reach significance. Hypothesis-2. Self satisfaction is negatively correlated with the degree of congruence between ideal same sex child and ideal self. This hypothesis was tested by a product-moment corre - lation test between difference scores of ratings of actual and ideal self (self satisfaction) and ideal self and ideal same sex child. The correlation obtained was only r = -. 02, and t was not significant. 37 Correlations among the factors (see Table 6) of Activity, Potency, and Evaluation for each variable were calculated. These correlations range from r = -. 14 to r = +. 45 and offer further support for the rejection of Hypothesis 2. For the Potency dimen- sion, r = +. 45 was significant at p < . 05, but in the opposite direction predicted relation between self satisfaction and the parents' ideal self was not found. Hypothesis 3. Marital satisfaction correlates positively with the size of the discrepancy scores between ratings of ideal opposite sex child and ideal spouse. Hypothesis 3 was tested by a product-moment corre- lation between discrepancy scores obtained with ratings of actual and ideal spouse and ratings of ideal opposite sex child. As seen in Table 5, the correlation was only r = . 15, and t was not signi— ficant. An intercorrelation matrix also failed to produce signifi- cant results (see Table 6). Thus, Hypothesis 3 was not confirmed, There was no relation between marital satisfaction and the desire for the opposite sex child to possess the characteristics of an ideal mate . 38 Hypothesis 4. Marital satisfaction correlates negatively with the size of the discrepancy scores between ratings of ideal opposite sex child and actual spouse. This hypothesis was tested by a product-moment corre- lation between discrepancy scores between ratings of actual and ideal spouse and ideal opposite sex child and actual spouse. The correlation derived was r = . 87. A significance test gave t = 12. 22 highly significant beyond the . 005 level of confidence. Corre- lations for all three factors (see Table 6) A, P, and E, of the variables are significant beyond the . 01- level for confidence. These findings lend support to the theory that as marital satis- faction increases, parents will increasingly desire their opposite sex child to possess characteristics similar to the valued mate. Hypothesis 5. There is a positive relationship between marital satisfaction and agreement of parents' ratings of their ideal children. Product-moment correlation coefficients for this hypo- thesis were obtained separately for agreement of parents' ratings of ideal sons and ideal daughters. One correlation was obtained between discrepancy scores of actual and ideal spouse and ideal daughter as rated by both parents. A similar correlation was computed using discrepancy scores of ideal son as rated by both 39 parents. The correlations obtained were r = . 012 for daughters (Hypothesis 5b) and r = . 17 for sons (Hypothesis 5a). In both cases the results are not significant. Intercorrelation matrices (see Table 6) produced one significant correlation, but in the I opposite direction of that predicted. For Potency, there was a negative correlation (r = -. 31) between marital satisfaction and agreement of ratings of ideal daughters (see Table 6). There- fore, Hypothesis 5, that parental agreement of ideal character- istics of sons and daughters is positively related to marital satisfaction cannot be accepted. Hypothesis 6. Marital satisfaction and self satisfaction of mothers as opposed to fathers are more highly correlated to other measures involving ratings of their children. .Hypotheses 2, 3, and 4 will yield significantly higher correlations when ratings of mothers are compared to ratings of fathers. To obtain comparisons of ratings of mothers and fathers, the data was separated into ratings made by mothers and fathers. Product-moment correlation coefficients were separately ob- tained from mother and father ratings. These correlations were then converted to z scores using Fisher' s r to z transformation tables, and a 2 test was performed to determine the significance of their difference. See Table 7. Then intercorrelation matrices 40 were developed for the three factors, A, P, and E, for mothers and fathers separately and 2 tests performed on the diagonals to determine whether these differences were significant. The re- sults are summarized in Table 8 and are as follows: Hypothesis 2. Self satisfaction of mothers as opposed to fathers, is more highly negatively correlated with the wish for the ideal same sex child to resemble the ideal self. The correlation obtained with mothers' data was r = . 13, and t was not significant. The correlation obtained with fathers' ratings was r = -. 02, and twas not significant. Analysis of the diagonal correlations yielded one significant result, in mothers! data, but in the opposite direction of that predicted (Potency: r = . 54, p< . 005). It appears that there is not a strong negative relation between self satisfaction and the wish for the same sex child to resemble the ideal self for either mothers or fathers. For mothers thewrelationship along the Potency dimension appears to be positive. These results lend further confirmation to those obtained for Hypothesis 2 above. Hypothesis 3. There will be a higher positive correlation be- tween marital satisfaction and the size of discrepancy scores between-ratings of ideal opposite sex child and ideal spouse of mothers as opposed to fathers. The correlation obtained with mothers' data was r = -. 10. The correlation obtained using fathers' data was r = . 27. t tests showed neither correlation was significant. Neither was the difference between these two correlations significant. Analysis of the diagonal correlations of E, P, and A indicated that only the 41 Activity dimension for fathers was sensitive (r = . 38, p <. 05) to these variables. Thus for fathers only, there was a positive correlation between marital satisfaction and the size of the dis- crepancy scores between ideal opposite sex child and ideal spouse. Again, however, the difference between mothers and fathers was not significant. Hypothesis 4. Marital satisfaction of mothers, as opposed to fathers, is more highly negatively correlated with the size of the descrepancy, scores between ratings of ideal opposite sex child and actual spouse. Product-moment correlations were performed separately using mothers' and fathers' ratings. For mothers, r = . 91, = 10. 52. For fathers, r = . 88, t = 8. 88. In both cases 1: is significant beyond the .005 level of confidence. The difference between these two correlations was not significant, however. Analysis of the factors E, P, and A for mothers and fathers yielded significant correlations on all three dimensions; again, none of the differences between mothers and fathers on E, P, and A were significant. Thus, there is no difference between mothers and fathers in the tendency for both parents to desire the opposite sex child to resemble the actual spouse as marital satisfaction increases. 42 Hypothesis 7. Self satisfaction correlates positively with con- giuence of ratings of ideal same sex child and actual self. Hypothesis 7 was developed on the basis of preliminary; analysis of the data which suggested that there might be a positive relationship between self satisfaction and the desire for the same sex child to resemble the actual self. Thus additional corre- lations were obtained between self satisfaction and ratings of ideal same sex child and actual self. The correlation for both husbands and wives combined was r = . 81. = 9. 57 and was signi- ficant beyond the . 005 level of confidence. Thus, as self satis- faction increases, so does the desire for the same sex child to possess characteristics similar to the actual self. Further analysis of the variables was made to determine the relationships of the factors E, P, and A (see Table 6). These correlation matrices further support the conclusions obtained in the results, all corre- lations being significant beyond the . 001 level. For mothers, r = . 84, t = 7. 42. For fathers, r = . 86, t = 8. 08. Both correlations are significant beyond the . 005 level of confidence. However, similar .to other findings involving com- parison of mothers' and fathers, there was no significant difference between mothers' and fathers' ratings. Similar results were also found when correlations were analyzed in terms of E, P, and A. 43 For both mothers and fathers, correlations were significantly positive on all three factors, but the difference between them was nonsignificant. 1 r" . .5 , ‘. o DISCUSSION OF RESULTS Overall support of the predictions did not occur. Of the seven hypotheses three were upheld. The following several paragraphs explore the implications of the results taken as a whole. The remainder of the chapter will examine theoretical variables of design and sample, and directions for further research. The Finding_s_ Hypothesis 1, that self satisfaction is positively related to marital satisfaction has been corroboratedby this research and therefore, adds support to a relationship reported elsewhere in the literature (see Review of the Literature above). Although the question of causality in this relationship cannot be answered with this type. of research, an alternative, heuristic enquiry is whether one variable, either self satisfaction or marital happiness, were improved, would the other likewise improve. It is a fre- quent observation by therapists that such indeed is the case, but the answer should be explored more formally. The question of whether the marital satisfaction-self satisfactionrelationship 44 45 directly. influences attitudes towards children should also be investigated for application to clinical practice. If self and marital satisfaction could be improved, for instance, would attitudes towards children becOme more healthful as a by -product. It may be that marital satisfaction, self satisfaction, and satisfaction with one' 3 children are all positively related. Unfortunately, ratings of actual children were not gathered in this study, so any such relationship cannot be shown here. Neither Hypothesis 2 nor Hypothesis 3 was confirmed. It appears that neither the ideal self nor the ideal spouse is used as a model for children. However, as shown by the confirmation of Hypotheses 4 and 7, as self and marital satisfaction increase, it is the actual self and spouse, rather than the ideals which are used as models. This is especially interesting because in cases where marital satisfaction is high, ideal and actual spouse are similar. Ideal self and actual self are also similar when self satisfaction is high. Only one correlation, the Potency or strength factor, indicated a possible positive connection between self satisfaction and the desire for the ideal same sex child to resemble the ideal self, and this correlation could possibly be picking up the similarity between ideal and actual self as self satisfactionincreases. There was no such positive relationship Vt W‘Z— I b 46 with the ideal -actua1 spouse correlation, however. So, parents, in general, indicated that they do not wish their children to re- semble either their ideal selves or ideal mates, but desire them to model after actual family members when those members are considered suitable. Perhaps the desire for the opposite sex child to resemble the actual spouse is a manifestation of healthy admiration of the mate. Likewise, when a person's self esteem is high, the desire that his child take after him (or her) is a healthy narcissistic expression. In these homes, because there is a concrete, approved model for the child, identification may proceed according to proper sequence. In homes where marital and self satisfaction are low, the child would have a more difficult time growing up. Here the parents do not approve of themselves or theirmates as models. They do not necessarily wish that the child become like their ideal selves or spouses either. The child is then left with the problem of having no consistently approved models in the home. Perhaps there is another model which is held.up to the child in such cases. But it is one which can never be as concrete and available as are his parents. Worse still, one parent may be used as a negative model and the child ethrted to avoid many of his characteristics, creating a 47 situation which leads with high probability to confusion, conflict, and emotional difficulties for the child. Hypothesis 5 must be rejected in totality. Not one correlation produced the slightest evidence of confirmation. It appears that the premises leading to the prediction might be invalid. In homes where there are poor marriages, parents disagree about many aspects of family life, but goals for the children is not necessarily one of them. If so, this is fortunate, because it suggests that children of poor marriages do not neces - sarily have to be confused or conflicted by their parents' disagree- ments in this area. However, it may also be that while parents may agree or not agree in the abstnact about ideal characteristics for their children, in real life, their behavior may be different from what their ratings imply. This discrepancy between self reports and actual behavior, unfortunately, is one of the unsolvable problems of attitude research. Hypothesis 6 also was not confirmed. It appears that mothers' wishes for their children' 3 development are not more closely related .to self satisfaction or marital happiness than'are fathers' wishes. In only one correlation, and that one in the opposite direction predicted (see Hypothesis 2 above) did the relationship hold. In Hypothesis 6 (2) on the Strength or Potency 48 factor only, mothers indicated a positive relation between self satisfaction and the desire for their daughters to resemble their ideal selves which was stronger than the corresponding father- son relationship. Contrary to expectation, several correlations indicated that fathers' , more than mothers' , desires for their children more closely parallel their satisfaction with themselves and their mates. On the Activity factor, fathers indicated a stronger negative correlation between marital satisfaction and the desire that their daughters possess characteristics in common withideal mates. Similarly, on both the Potency factor and the evaluative dimension, fathers desired their sons to resemble them more closely as their self esteem increased. Fathers also indicated a stronger positive relationship than did mothers be- tween self satisfaction and marital happiness. However, fourteen of the twenty comparisons investigated produces no difference between the ratings of mothers and fathers. Since other research cited in the introduction supports the fact that mothers influence child development more, the process by which this occurs appears not to be one of motivation. Other variables must enter into maxi- mizing mothers' influence while minimizing that of the father. Simple proximity is the most economical explanation. Mothers and fathers may have equally strongideas about the developmental 49 directions which they want their offspring to follow; indeed, under certain circumstances, fathers may have even more definite ideas of particular characteristics they desire in their children, but since the father is usually away at work all day, if the parents' ideas differ, the mother has more time with the child, and there- fore, more opportunity to influence him. More research is needed to show the separate effects of each parents' influence on the child. Theory and Design The one main theoretical position of this study‘was that family attitudes are determined by more basic motives which also influence the behavior of parents in their familial situation. It was theorized that because of the impetus of basic needs, parents' attitudes towards themselves and their families would all be interrelated. Eventually these needs and attitude constellations would be mirrored in the behavior of the family--in the behavior of the parents towards their children and thenin the children' s response. It was theorized further, that if the parents were relatively satisfied in their needs regarding their marital partners and themselves, their attitudes towards their children would differ from those cases. in which self esteem needs and marital satisfaction motives were less fulfilled. To examine these relationships more 50 closely, several hypotheses were formulated from these premises. Three findings seemed to lend confirmation to these premises. The remaining four hypotheses did not. The theoretical position has been heavily supported by earlier research, as seen from the Review of the Literature. Perhaps, then, the reasons for the 3 lack of positive results here can be found partly in the nature of 1 the research design and instrument used. The Semantic Differential is a flexible instrument which has been used to elucidate the meaning of many different concepts. The experimenter is able to choose from a large number of polar adjectives those which are most relevant to the topic under con- sideration. This very flexibility, however, can constitute a weakness of the instrument as there is no way of knowing before- hand, other than guessing on the basis of face validity, whether the items chosen are sensitive to the concept examined. Usually if the results are positive, one concludes afterwards that the items were valid. In this particular case, the items chosen may have been inappropriate. However, since the selection procedure has been validated before, the items were probably relevant and the negative results valid. The Semantic Differential has much support as an attitude measuring devise (Osgood et al. , 1957) because of its ability to tap levels of meaning below the surface of awareness. In other words, 51 Semantic Differential data are interpreted as reflective material . ordinarily garnered by a more traditional projective test. An attitude, when conceived as being the result of basic needs and motives, is the product of many thought processes from varying levels of consciousness. However, it may be that in the area of familial attitudes, a more traditional projective device would better uncover some of their more unconscious aspects. One value of this study was that it allowed for four ways of looking at the relationships among attitude variables. Overall correlations and separate correlations for the three factors con- tributing to the connotative meaning of the variables rated were computed. The generally consistent results across all corre- lations seem to indicate that the overall correlationis as sensitive to the interrelationships as are the factors E, P, and A taken separately. This is slightly at variance with Osgood et al. , (1957). who suggested that the Evaluative dimension is the only factor sensitive to the meaning of attitudes. These results suggest that part of the meaning is also carried along Potency and Activity dimensions . Thisproject was planned to be strictlyattitudinal, in- vestigating only the'relationships among attitudes. However, this design has certain inherent limitations, which could be overcome by a different research design. Most important of these is the 52 absence of a behavioral reference point with which to compare subjects.- With definite information as to the actual behavior of the parents in this study towards each other and their children, the attitudinal information would be more meaningful. If the parents were separated into groups on the basis of this behavior, then their attitudes could be compared, with their behavior used as a reference point. This paradigm is one which is valid and reliable in the area of parental attitude research because of the wealth of positive findings reported. Given that the study remained on the level of cognitive investigation, the variables chosen to illustrate the attitudinal interrelationships could have been unproductive from the be- ginning. For example, a measure of actual child, if included, might have given rise to more productive results. A discrepancy score between actual and ideal child could have been used as a measure of child rejection and might have shown interesting relationships with self and marital satisfaction. This way be- havior towards children could have been more easily inferred. If, in conjunction, the children had been of known adjustment levels, then more concrete inferences could have been made about treatment of the child by the parents. way-1L” 'Eq... 53 In trying to assess a measure of approved identification, the discrepancy between actual spouse or actual self and ideal child was used. However, this might not be valid. For instance, the descriptions may vary according to age, making abstract assumption of similarity between variables meaningless. In speaking of identification, a measure of similarity of ratings of ideal self or ideal spouse to ideal child was used. It turned out that similarity of ideal child to actual self and actual spouse was more productive. Even so, identification has more complicated aspects, and this design did not take them into account. Parents, for instance, may desire their children to resemble themselves or spouses on certain qualities and not others. Therefore, they probably selectively encourage and discourage modeling behavior when they observe it intheir children. Ambivalence is another characteristic of attitudes not considered in this design. It has been observed that parents sometimes consciously disapprove of ,certain characteristics in their child, but on another level be pleased to see it. One well known example of this is when a parent decries his child' s stubbornness, but is covertly pleased to see it as ”a chip off the old block. " ' 54 The design of this research, then, was deficient in these three factors: a behavioral reference point, clearly valid scales, and a consideration of ambivalence and selec- tiveness on the part of parents. Sample The sample of twenty -five Spartan Village couples used as subjects in this project were probably somewhat atypical when compared with subjects usually used in parental attitude studies. They were probably younger, although specific age data was not gathered. As seen in Table 2, they were married for only an average of five years and had only . 9 children. Eleven of the twenty -five were expecting their first child. Of the fourteen couples who had at least one child, the average number of children was 1. 4, far fewer than the national average. The meanage of these children was only 4 years 6 months. One couple differed from the rest in that they were much older (in their forties) and had been married nineteen years. All these variables could easily have affected the results of the study in some significant but unknown way. The fact that so many couples were Still childless could have great ramifications. It is possible that their thinking about their unborn child could 55 have been in terms more idealistic or unrealistic than couples who already had children. Since many of the marriages were relatively recent, under three years, they could conceivably be happier than most, making the ratings on marital satisfaction quite skewed. On the other hand, living under the near poverty cir- cumstances of the average married student, in crowded housing, .v.. supporting a mate and child, under constant pressure of studies, Is} and the stigma of being subjected to the whims of an enormous bureaucratic institution, self satisfaction could very well be lower than average, skewing this rating in the opposite direction. The table of means (see Table 4) cannot answer these questions as there is no data with which to compare it. The writer noticed gross differences in the behavior of the subjects during the interview. Some subjects displayed a polite interest in the project and were courteous and friendly. Their apartments were pleasingly arranged, and in general, the living arrangements seemed comfortable. Other subjects were annoyed at the imposition of the rating tasks and the examiner' 3 presence, although they had agreed beforehand to participate and knew why they had been chosen. These subjects derogated the project as much as possible, criticizing the test items and 56 acting in a disagreeable fashion. The writer was struck by the total disorganization of several apartments. Clean and dirty laundry lay in heaps on the floor and chairs, and in general, the atmosphere was depressing. It is reasonable to assume that the ratings of all these respondents, indeed their attitudes and behavior would differ. Unfortunately, the design of this research does not separate these subjects, so this must remain a guess. However, the results may be influenced in some un- known fashion. It is suggested that in the future, subject variables be more carefully controlled, especially when using such an atypical sample as married students. Directions for Further Research Attitude investigation has already made valuable con- tributions to the field of psychology. The area of family life, and child development in particular, has been broadened by the knowledge gained from studies of parental attitudes and expec- tations. But despite many years of attention, attitude research is still plagued with the same basic problem which Thurstone (1936) wrestled with years before. Attitudes are still difficult to measure. Because of the multi -level nature of attitudes, and the practice of many people to deny their attitudes under if 57 certain conditions, contemporary research is making more and more use of projective techniques in the investigation of attitudes, either along with, or instead of conventional questionnaires. These projective techniques follow a continuum from greater to less structure. Greene (1967), in studying motivation for parenthood, used a straightforward questionnaire with the intro- duction, "Most people believe. . . " The answers of the respondents 5‘ were assumed to be projections of their own motives. Further along the continuum, .Major (1967), investigating the same topic, used TAT -like pictures, but accompanied them with stories to which the subjects chose from among four alternative endings. There were no differences between her adjusted and maladjusted groups, although the results were not as uniformly negative as Greene' 8. Carter (1968), used a variation of Major' 8 procedure. Her stories were similar, but they were open ended, and the subjects completed them in any fashion they chose. Carter's results using this more unstructured technique were more positive than the other two studies. The Semantic Differential, used in this project, probably falls somewhere between the objective questionnaire of Greene and the semi -structured technique of Carter on a continuum of ”projectiveness. " The results produces fall somewhere in between, also, in terms of 58 confirmation of the hypotheses, Greene' 5 being wholly negative, these being partly negative, and Carter's being most productive. ,, If the above group of studies is at all representative, it appears that family attitude measuring devices with some degree of projectiveness are most productive in uncovering multi -level information about this subject. A completely un- structured "wide band" test, such as the Rorschach is not appropriate, however, because, by its nature, it cannot answer specific attitude questions. In the future, then, the use of semi -structured projective techniques should be investigated more thoroughly in family attitude research. In the future, also, attitude studies might profitly explore more fully the basic determinants of attitudes and opinions. Unlike the problem of measurement, this is a rela- tively new concern which mirrors the dynamic approach to any human characteristic. If attitudes are conceived as the defen- sive end -product of more basic motives, then they can be the outcome of many mechanism of defense which people employ. As with other defenses, people may exhibit the same attitude while their underlying need structure may be quite different. Conversely, similar attitudes canlead to different behaviors. In addition, it would be no surprise if the effects of the attitudes on other people would vary, also. This area is one which could be 59 quite useful, and it is hoped that it will continue to become a focus of interest. With projective methods applied to the inves- tigation of attitudinal determinants, there exists a real possi- bility of productive findings. Another area which deserves more intensive investi- gation is the process by which familial attitudes are translated into behavior. Ultimately, it is not important that people hold certain opinions if the process ends there. What causes great ramifications in the family is what the individual does with his opinion. It is the individual' 3 behavior, whether subtle or easy to observe, which causes reactions in his children and mate. Until now, research in this area has usually stopped after the attitude has been delineated and a few correlations have been discussed. However, the variable by which the attitude has been transmitted or expressed, the behavior of the person who possesses the attitude, is often neglected. Concentration on what response it meets in the environment would help answer many puzzling questions of family interaction. In the future, research into this area should consider the change of family attitudes. Studies seem to emphasize delineating more clearly and exposing more fully the nature of family attitudes and their interrelationships and manifestations in behavior. Hopefully, research in this area eventually will 60 isolate those attitudes which are healthful to family living from those which are pathogenic to the ongoing family. The ultimate goal, however, should be to learn how to change those pathological attitudes and opinions, or the more basic needs which give rise m to them. In this way psychology can truly contribute to the happiness of human existence. It is not enough, although it is a great deal, to know the process by which attitudes are formed and expressed ; real contribution requires an ability to put such it" knowledge to work improving human living. In conclusion, research directions in the field of familial attitudes should concentrate attention on the following problems: measurement using suitable projective techniques, the 'baSic determinants of attitudes, the process by which attitudes are translated into behavior, and ultimately, the methods by which pathogenic attitude constellations can be made more health producing. SUMMARY This study was designed to explore the interrelationships of attitudes which parents have towards themselves, each other, and their children. . . The main theoretical position of this paper conceives of 61 family attitudes as the end products of more basic motives which also influence the behavior of parents in their family situation. It was hypothesized that because of the impetus of these basic motives striving towards gratification, parents' attitudes towards themselves and their families would all be interrelated. It was further theorized that if parents were relatively satisfied in their relationship with their mate and possessed a relatively good self image, then they would be favorably inclined to allow their children to identify with themselves or their mates, according to the sex of the child. Conversely, in homes where the parent is unhappy with himself and his mate, he would hope that his child become like his idealized image of himself or like anideal mate, depending on the sex of the child. In these homes, it was theorized, the children would be used to satisfy needs ordinarily met by the self or the spouse. 61 62 Also considered in the project was the confirmed observation that mothers have more influence than fathers in the development of their child' s masculine or feminine identification. The reason for this greater influence was thought to be the mother' s greater emotional involvement with her children and the closer ties which all aspects of parenthood have to her self image and marital relationship. Because of this, it was hypo- thesized that all measures regarding parents' attitudes toward their children which are associated with marriage or self image would be stronger for mothers than for fathers. To test this viewpoint, the Semantic Differential was used to gather descriptions of various actual and ideal family members as rated by a sample of twenty-five Spartan Village couples. The ratings were compared by product-moment corre - lations and the results observed. These results partly supported the theoretical position in that evidence for at least partial interrelationship of attitudes was discovered. Specifically, a positive correlation was found betweenself satisfaction and marital satisfaction. As an in- dividual became more self satisfied, the more he desired his same sex child to resemble himself. Similarly, the more satisfied he was with his mate, the more he wanted his opposite sex child to possess characteristics similar to that mate. 63 Not supported was the prediction that the ideal self or ideal mate would be used as models for the child. There was also no evidence that mothers data would show stronger corre— lations than would fathers' ratings. These results were discussed in terms of identification ramifications for the child, and marital and personal satisfaction for his parents. Suggestions for improvement of similar attitude research was discussed in terms of sample control and measurement by semi —struotured projective techniques. Directions for future research emphasized the area of basic determinants of attitudes, the process by which attitudes are formed, and the benign transformation of pathogenic attitudes. BIBLIOGRAPHY -_t HI» A- , . E BIBLIOGRAPHY Ackerman, N. W. The Psychodynamics of Family Life. New York: Basic Books, 1958. Adler, A. The Neurotic Constitution. New York: Moffatt Yard, 1921. Allport, G. The trend in motivational theory. American J. Orthopsychiat., 1953, 23, 107-119. Ansbacher, H. and Ansbacher, R. (Eds.). The Individual Psychiatry of Alfred Adler, New York: Basic Books, 1956, p. 373. Arieti, S. Interpretation of Schizophrenia. New York: Robert Briner, 1955. Ausubel, D, et al. Perceived parental attitudes as determinants of children' s ego structure. Child Dev. , 1954, 35, 173- 184. Baruch, D. and Wilcox, J. A study of sex differences in pre- school children' s adjustment coexistent with interparental tensions. J. Genetic Psychiat., 1944, 64, 281-303. Baxter, J. et al. Father identification as a function of mother- father relationship. J. Indiv. Psychol. ,. 1964, 20, 167- 171. _ Berger, E. M. The relationship between expressed acceptance of self and expressed acceptance of others. J. Abnorm. Soc. Psychol., 1952, :11, 778-782, Block, J. Personality characteristics associated with fathers" attitudes toward child rearing. Child Dev. ,, 1955, 26, 41 -48. 64 65 Brody, Sylvia. .Patterns of mothering. New York: International Press, 1956. Buchmueller, A. D. and Gildea, M. C. Group therapy with parents of behavior problem children in public school: report of work. in schools in St. Louis county. Int. J. Soc. Psychiat. , 1955, 1, 51-56. Carter, Charlene. Motivation for parenthood and parental be- havior as perceived by disturbed and normal children. Unpublished Ph. D. dissertation, Michigan State University, East Lansing, Michigan, 1968. Cronbach, L. J. and Gleser, G. Assessing similarity between profiles. Psychol. Bull., 1953, _5_0, 456-473. Ellis, A. The American Sexual Tragedy. New York: Twayne Publishers, 1954. Farber, B. Marital integration as afactor in parent-child relations. Child Dev. , 1962, 33, 1-14. Freeman, Howard et a1. Possessiveness as acharacteristic of mothers of schizophrenics. J. Abnorm. Soc. Psychol. , 1959, _5_3, 271-273. Freedman, Deborah S. et al. Size of family and preference for children of each sex. Amer. J. Sociol., 1960, 66, 141- 146. '— Fromm, E. Selfishness and self love. Psychiatry, 1939, 2, 507-523. Gildea, Margaret et al. Maternal attitudes and general adjust- ment in school children. In J. C. Glidewell (Ed.) Parental Attitudes and Child Behavior. Springfield, Illinois: Charles C. Thomas, 1961, 42 -89. Guertin, W. Are differences in schizophrenic symptoms related to the mother' s avowed attitudes toward child rearing? J. Abnorm. Soc. Psychol. , 1961, 63, 440 -442. 66 Greene, R. J. Motivation for parenthood in mothers of disturbed and mothers of normal children: An exploratory study. Unpublished Master' 3 thesis: Michigan State University, East Lansing, Michigan, 1967. Hamilton, G. A research in marriage. In M. Kovonarovsky , Functional analysis of sex roles. American. Soc. Rev. , 1950, 15, 508 —516. Heilbrun, A. Perception of maternal child rearing attitudes in schizophrenics. J. Consult. Psychol., 1960, 24, 169- 173. Helper, M. Learning theory and the self concept. J. Abnorm. Soc. Psychol., 1955, 5_1_, 184-194. Horney, Karen. The Neurotic Personality of Our Time. New York: Norton, 1937. Horowitz, F. and Lovell, L. Attitudes of mothers of female schizophrenics. Child Dev., 1960, 31, 299-303. Katz, Daniel. The functional approach to the study of attitudes. The Public Opinion Quarterly, 1960 2_4, 164-204. Kohn, M. and Clausen, J. Parental authority behavior and schizophrenia. American J. Orthopsychiat. , 1956, 22, 297 -313. Krung, R. S. Attitudinal and behavioral differences in child rearing practices between parents of clinic and non- clinic children: A questionnaire assessment. Dissertation Abstracts, 1965, 2_6, 2869-2870. Lane, R. C. and Singer, J. L. Familial attitudes in paranoid schizophrenics and normals from two socioeconomic classes. J. Abnorm. Soc. Psychol., 1959, _5_£_), 328-339. Leton, D. A. A study of the validity. of parent attitude measure- ment. Child Dev. , 1958, 2_9_, 2869-2870. Levy, D. Maternal Overprotection. .New York: Columbia University Press, 1943. 67 Loevinger, Jane and Sweet, Blanche. Construction of a test of mothers' attitudes. In J. C.‘ Glidewell (Ed.) Parental Attitudes and Child Behavior. Springfield, DIW— Charles C. Thomas, 1961, 110-123. Lu, Yi-Chuang. Mother -child relations in schizophrenia. Psychiatry, 1961,21, 133-142. Madoff, J. M. The attitudes of mothers of juvenile delinquents _ toward child rearing. J. Consult. Psychol., 1959, 23,, 518-520. '"' Major, M. A. Assessment of motivation for parenthood in parents of disturbed and normal children. Unpublished Ph.D. dissertation, Michigan'State University, East Lansing, Michigan, 1967. Mark, J. The attitudes of the mothers of male schizophrenics toward child behavior. J. Abnorm. Soc. Psychol., 1953, :1_8_, 185-189. Meyer, R. and'Karon, B. The schizoprenogenic mother concept andthe TAT. Psychiatry, 1967, 3_0_, 173-179. Osgood, C. The nature and measurement of meaning. Psychol. Bull. , 1952, 49, 197-237. Osgood, C. and Suci, G. J. A measure of relation determined by mean difference and profile information. Psychol. Bull. , 1952, 49, 251-262. Osgood, C. et al. The Measurement of Meaning. Urbana, Illinois: University of Illinois Press, 1957. Parsons, T. and Bales, R. Family,.Socialization, and Inter- actionProcess. New York: Free Press, 1955. Rabin, A. I. A contribution to the "Meaning" of Rorschach' s inkblots via the semantic differential. J. Consult. Psychol., 1959, 23, 368-372. Rabin, A. 1. Motivation for' parenthood. J. Proj. Tech., 1965, is, 405-411.“ 68 Roe, Ann and Siegelman, M. A parent -child relationship question- naire. Child Dev., 1963, 34, 355-369. Rogers, C. A coordinated research in psychotherapy: A non- objective introduction. J. Consult. Psychol., 1949, 1_3_, 149-153. Ross, A. A schizophrenic child and his mother. J. Abnorm. Soc. Psychol., 1955, _51, 133-139. Sanford, N. The hynamics of identification. Psychol. Rev. , 1955, 62, 106-118. ‘ ' Schaefer, E. S. A circumplex model for maternal behavior. J. Abnorm. Soc. Psychol.,1959, _5_9, 226-235. Schaefer, E. S. and Bell, R. G. Patterns of attitude toward child rearing and the family. J. Abnorm. Soc. Psychol. , 1957, 544.: 391-395. Schaefer, E. S. and Bell, R. G. Development of a parental attitude research instrument. Child Dev. ,. 1958, 29, 339-361. Sears, R. R. et al. Patterns of Child Rearing. Evanston, Illinois: Row Peterson, 1957. Sheerer, E. An anlysis of the relationship between acceptance of and respect for the self and acceptance of and respect for others in ten counseling cases. J. Consult. Psychol. , 1949, 13, 169-175. Shoben, E. J. The assessment of parental attitudes inarelation to child adjustment. Genet. Psychol. Monog., 1949, 39, 101-148. Siegelman, M. Evaluation of Bronfenbrenner' s questionnaire. Child Dev. , 1965,, 3_6_,_ 163-174. Stock, D. An investigation into the interrelations between self - concept and feelings directed toward other persons and groups. J. Consult. Psychol. , 1949, 1_1_3_, 176-180. 69 Stogdill, R. M. The measurement of attitudes toward parental ‘ control and the social adjustment of children. J. Appl. Psychol. , 1936, 2_0_, 359-367. Symonds, P. The Dynamics of_1:arent-Child Relationships. New York: Bureau of Publications, Teachers College, Columbia University, 1949. Thurstone, L. L. The Measurement of Attitude. Chicago: University Press, 1936. Williams, ‘W. W. The PALS Test: A technique for children to evaluate both parents. J. Consult. Psychol. , 1957, 54, 229-249. Zuckerman, B. and Oltean, M. Some relationships between maternal attitude factors and authoritarianism, per- sonality needs, psychopathology and self acceptance. Child Dev., 1959, g), 27-36. Zuckerman, B. et al. Acceptance of self, parents and people in patients and normals. J. Clin. Psychol., 1956, 12, 327-332. Zunich, - M. Child behavior and parental attitudes, J. of Psychol. , 1966, 32, 41-46. APPENDICES APPENDDi A DIRECTIONS The purpose of this study is to investigate some of the expectations parents have regarding their families. In filling out this question- naire, please make your judgements on the basis of what applies to you. You are to rate the particular family member named at the top of each page of the scales in order. If you feel that the family member at the top of the page is very close to one end of the scale, you should place your check -mark as follows: Kind X : : : : : : cruel Kind : : : : : : X cruel In each case, put your check in the space which best describes the person being rated. If you consider a person to be neutral on the scale, both sides of the scale equally associated with the person, or if the scale is completely irrelevant, then place your check in the middle space. Kind : : : X : : : cruel Please remember to 1. be sure to check every scale for every person 2. never put more than one check on a single scale. Although some items may be similar, they are all different. For each person make a separate and independent judgement. Work at fairly high speed through this test. Do not worry or puzzle over individual items. It is your first impression which is important here. Thank you very much for your cooperation. Please note. Although the middle scale indicates that both sides are equally characteristic of a person or that the scale is ir- relevant, you are requested not to use it unless absolutely necessary. 70 APPENDIX B PLEASE RATE Active Angular Beautiful Brave : : : : Calm __ __ __ __ __ __ __ Deep Dirty Fast Happy , Hard Hot Honest Loud Obedient Pleasant Poor Rough Sharp Weak Young 71 passive rounded ugly cowardly agitated shallow clean slow sad soft cold dishonest soft disobedient unpleasant rich smooth dull strong old MICHIGAN STQTE UNIV. LIBRQRIES NH 1 I llll lllll lllll ll ‘llll llllll‘ 9 7 312 30098 547