u... .3. m.,‘. I ‘6... in!“ ‘uxaw v 'cmlioh f } find any.- anuqu o. .. .a E v03»! 3.53pm?" .‘2.0 2/601 ABSTRACT ~MULTIVARIATE PREDICTION OF DRINKING IN ADOLESCENT GIRLS: AN INTERGENERATIONAL ANALYSIS By Katharine Anna de Vaux The present study is primarily an inductive search, designed to investigate the correlates of problem drinking among adolescent girls. A multivariate statistical approach was utilized in order to examine the interrelationships as well as the combined predictive ability of the adolescent's personality, parental personality and drinking patterns. family atmosphere, child rearing practices, and demographic character- istics with regard to the quantity and frequency of alcoholic intake, alcoholic antisocial behavior, and social motivations for drinking. "The adolescent subjects for this study were seventy-five girls who were juniors or seniors in high school and selected so as to ade- quately represent the full range of the drinking continuum from non- drinking to heavy drinking. The Park Problem Drinking Scale (Park, 1962) was used to measure problem drinking while the drinking quantity and frequency measure was derived from Cahalan and Cisin (1969). Moti- vations for drinking were assessed using attitude scales taken from Grossman's cluster analytic study (Grossman, 1965). The Bronfenbrenner Parent Activity Inventory (Bronfenbrenner, l961) was administered as a ‘hl I .-“3 I (”‘“"\. .‘ f/ Katharine Anna de Vaux measure of child rearing practices experienced by the adolescent subjects while they were growing up. The subjects' antisocial behavior was assessed through a modified version of the Antisocial Behavior Check- list (Kulik, §t_al,, 1968); and a personality inventory was obtained through the use of the Rutgers Attitude Survey which includes scales taken from Kalin and Williams (1967) and Moore and Holtzman (1965), the Marlowe-Crowne Social Desirability Scale (Crowne and Marlowe, 1960), and the Sociability and Socialization Scales from the California Per- sonality Inventory (Gough, 1957). A background questionnaire assessing demographic characteristics completed the adolescent data collection. A parent sample was also obtained including fifty-six per cent of the natural parents of the adolescent subjects. Thirty mothers and thirty-six fathers were either not living at home or unwilling to complete the study material. Each parent filled out the Bronfenbrenner Parent Activity Inventory (Bronfenbrenner, 1961) for himself as applied to the adolescent in the study, the six Kalin and Williams (1967) scales included in the adolescent data collection, the Marlowe-Crowne Social Desirability Scale (Crowne and Marlowe, 1960), a background question- naire, and measures of parental drinking behavior. The drinking ques- tions were taken from Cahalan and Cisin (1969). A series of questions regarding "implicative drinking" taken from Haberman and Sheinberg (1967) was added to assess possible drinking problems. The statistical analysis by stepwise multiple regressions consistently demonstrated a relationship between asocial or antisocial behavior and the consumption of alcoholic beverages. The data further suggest that the development of undercontrolled and impulsive behavior may emanate from the identification by our teenage subjects with Katharine Anna de Vaux mothers who are themselves deviant and impulsive, together with a lack of adequate sex-role socialization through instrumental or affiliative companionship with either parent, and the lack of appropriate conformity or impulse control training on the part of the father. Adolescents who scored high on the three drinking measures of Quantity-frequency, Alco- holic Antisocial Behavior, and Social Reasons for Drinking appeared to have prematurely assumed adult responsibilities and prerogatives with relative financial and moral independence and the adoption of the sexual, drinking and smoking habits of maturity. While there was a general commonality in the relationships found between the independent variables and each of the three dependent variables, some differences were present in the degree to which certain behaviors were exhibited as well as in their relative salience for the three categories. Girls scoring high on Alcoholic Antisocial Behavior appeared to be most completely and visibly identified with an asocial or antisocial role and to have the most perceived difficulty with both parents. Girls scoring high on Social Reasons for Drinking appeared to exhibit less antisocial behavior and to report less friction with parents than did girls scoring high on the other two measures. They also appeared to have positive or neutral relationships with their fathers in distinction to the negative father- daughter interactions reported for the Quantity-frequency and Alcoholic Antisocial Behavior measures. In addition, there appeared to be a rela- tively broader spectrum of influence from the various classes of inde- pendent variables upon the Social Reasons for Drinking category while there was the implication that Quantity-frequency was less dependent Katharine Anna de Vaux upon environmental and social influence than it was upon the child's personality make-up and perception of child rearing practices. MULTIVARIATE PREDICTION OF DRINKING IN ADOLESCENT GIRLS: AN INTERGENERATIONAL ANALYSIS By Katharine Anna de Vaux A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1976 ACKNOWLEDGMENTS I would like to express my gratitude to Dr. Robert A. Zucker, chairman of my thesis committee, for his continuing encouragement and guidance in this work as well as in my general development as a clinical psychologist. I would also like to thank Dr. Ellen Strommen and Dr. Joel Aronoff for their contributions as members of my committee. Per- haps most importantly, I would like to thank my friends, Michael Teixeira and Sheila Bienenfeld, for their love and support and Mitchell Bloomfield for accepting full responsibility.‘ ii TABLE OF CONTENTS Page LIST OF TABLES ........................ INTRODUCTION ......................... 1 THEORETICAL BACKGROUND .................... 3 Adolescent Personality .................. 3 Parental Drinking ...... . . ............. 6 Parental Personality ................... 7 Family Atmosphere . . . . ................. 8 Child Rearing Practices .................. 9 Peer Influences ..................... lO Demographic Characteristics ................ 11 The Problem ........................ 13 METHOD . . . . ........................ 16 Subjects . . . . . . ................... 16 Procedure ......................... 17 Statistical Analysis ................... 19 RESULTS AND DISCUSSION . . . . . . . . . . . . . . . . . . . . 20 Multivariate Analyses ................... 34 GENERAL DISCUSSION .......... . ...... , ..... 53 BIBLIOGRAPHY ......................... 6O APPENDICES A. MISSING DATA ..................... 64 B. SIMPLE CORRELATIONS .................. 72 C. DESCRIPTION OF MEASURES USED IN DATA COLLECTION . . . . 81 iii LIST OF TABLES Table Page 1. Report of missing data analyses ............ 21 2. Simple correlations of adolescent drinking measures with demographic variables . ............ 24 3. Simple correlations of adolescent drinking measures with adolescent personality variables . . ..... 26 4. Simple correlations of adolescent drinking measures with child's perception of child rearing practices ......... . ........... 28 5. Simple correlations of adolescent drinking measures with parental personality and drinking indices . . . 31 6. Simple correlations of adolescent drinkin measures with parental child rearing practices Ias reported by parents) ........ . . . . . . . . 32 7. Stepwise multiple regressions of adolescent drinking measures with demographic variables . . . . . . . . 35 8. Stepwise multiple regressions of adolescent drinking measures with adolescent personality variables . . . 37 9. Stepwise multiple regressions of adolescent drinking measures with adolescent perceptions of child rearing practices ......... . ....... 39 10. Stepwise multiple regressions of adolescent drinking measures with parental perceptions of child rearing practices ....... . ......... 41 ll. Stepwise multiple regressions of adolescent drinking measures with parental personality and drinking variables ......... . . . . . . . ..... 42 12. Multiple correlation coefficients obtained through stepwise multiple regressions of adolescent drinking measures with each class of independent variables . . . . . . . . . . ........... 45 iv Table Page 13. Stepwise multiple regressions of adolescent drinking measures across classes of adolescent variables . . . 46 14. Stepwise multiple regressions of adolescent drinking measures across all classes of independent variables . . . . . . . . . . . . . ......... 50 A.l. Missing data analyses ................. 65 3.1. Simple correlations of adolescent drinking measures with all demographic variables . . . . . . . . . . . 72 8.2. Simple correlations of adolescent drinking measures with all adolescent personality variables ...... 74 3.3. Simple correlations of adolescent drinking measures with all child's perception of child rearing practices ........... . .......... 75 3.4. Simple correlations of adolescent drinking measures with all parental personality and drinking indices ....... . . . . . .......... 77 3.5. Simple correlations of adolescent drinking measures with parental child rearing practices as reported by parents) ................ 79 6.1. Description of measures used in data collection . . . . 81 Table Page 13. Stepwise multiple regressions of adolescent drinking measures across classes of adolescent variables . . . 46 14. Stepwise multiple regressions of adolescent drinking measures across all classes of independent variables . . . . .................. 50 A.1. Missing data analyses ................. 65 8.1. Simple correlations of adolescent drinking measures with all demographic variables ........... 72 8.2. Simple correlations of adolescent drinking measures with all adolescent personality variables ...... 74 8.3. Simple correlations of adolescent drinking measures with all child's perception of child rearing practices ........... . .......... 75 8.4. Simple correlations of adolescent drinking measures with all parental personality and drinking ‘ indices ............ . .......... 77 8.5. Simple correlations of adolescent drinkin measures with parental child rearing practices Tas reported by parents) ................ 79 0.1. Description of measures used in data collection . . . . 81 INTRODUCTION Two reviews of the psychological test literature (Sutherland, et_al,, 1950; Syme, 1957) have concluded that there is no "alcoholic personality" and that research should be directed at other variables than personality in order.to.understand and explain alcoholism in men and women. This viewpoint seems extreme when one considers the con- sistency with which certain personality characteristics and problem drinking are linked in.the.1iterature. In a recent study of marijuana use in high school and college youth, Jessor, §t_al, (1973) noted that " . . . the system of.personality variables alone contributed signifi- cantly to accounting for the variance in marijuana use . . . ," and concluded that " .1. . personality is, indeed, central to variation in drug use, whether there.is.high social support for it or not." Both Jessor, et_al. and Riester and Zucker (1968) found a positive relation- ship between drug use and the use of alcoholic beverages in their young populations, and one.wou1d expect to find that personality variables are as meaningful in relationship to problem drinking as they are in relationship to drug use. The present study is concerned with person- ality variables as they relate to problem drinking in adolescent girls; but keeping in mind the results reported by Sutherland and Syme, the study will use multivariate methods in considering the combined predic- tive ability of not only the adolescent's personality, but also parental.personality and drinking patterns, family atmosphere, child rearing practices, peer influences, and demographic characteristics. THEORETICAL BACKGROUND The review of the literature which follows is not intended to be discursive.but.to.present the research findings in the field of . alcoholism as they pertaio.to the above variables in explaining problem drinking in adolescent girls. .Much of the background data had to be gleaned.from the case.histories of adult female alcoholics. Of course, this is.largely.retrospective.information, subject to the vagaries of memory and imagination and difficult.or impossible to verify with secondary sources. Since this area of problem drinking among women is so little researched, this.study.is primarily an inductive search; rather than proposing.theory.to predict data results, it will develop theoretical.explanations in the discussion as they follow from our findings. Adolescent Personality Most of the.literature pertaining to personality variables. of female problem drinkers are concerned with adult women. The adoles- cent years may be described in.terms of family atmosphere, relationships within the family, and parental attitudes and characteristics, but little is.mentioned of the personality traits in adolescence which correlate with present.or.later drinking patterns. Two writers, Jones (1971) and Zucker (unpublished data) have conducted research on social and personality factors with adolescent girls. Based on current drinking practices, Jones classified the members of her longitudinal study as problem.drinkers, moderate drinkers, light drinkers and abstainers...Personality characteristics distinguishing the five cate- gories of.Subjects were evident in the adolescent ratings collected during the early or the middle adolescent years; and the characteriza- tions in the.junior high school period and particularly those in the eighth grade.were more clearly related to adult drinking behavior than the ratings in the high school.period. Problem drinkers and abstainers were found to be.similar. .Both were described as vulnerable, withdrawn, pessimistic, and selfsdefeating. The important differences were that the abstainers were able-to-accept dependency relationships and were emotionally.controlled,.consistent,.responsible, conventional and ethi- cal. Problem.drinkers were judged to be undercontrolled, impulsive and submissive in.youth.whi1e-rebellious in adulthood. Jones'.case summary sketch of the problem drinker describes a girl who changes drastically-over.the junior high school years. At. age twelve,.she is.confident of her appearance, intelligence and the opinions.others have.of.her.. By fifteen she is confused, fearful and distrusts herself and.others. Her religion is one stressing judgment and punishment. She attempts to solve her problems through super- femininity. kFor the girl who will later become a problem drinker, adolescence appears to be a more critical and painful period than it is for those who.will.follow.other drinking patterns. , Zucker's.cross.validation.sample of seventy-four adolescent girls is largely consistent with Jones' findings. In this study, all of the antisocial behavior variables were significantly correlated with problem drinking, supporting Jones' description of problem drinkers as undercontrolled and impulsive. However, Jones depicts problem drinkers as submissive in youth and rebellious in adult life, whereas Zucker's data indicates.rebellious activity during adolescence. This may be a generational difference between the 1940's and 1970's as well as possibly reflecting the absence in Zucker's problem drinking sample of girls who.will.begin.their drinking careers after adolescence. Zucker's sexual behavior variable concerning heterosexual activities such as petting.and intercourse was significantly correlated with problem drinking and.is.suggestive.of one possible aspect of the escape into femininity which Jones reports. Items concerned with resentment of dependency in.reference to parents and distrust of others were sig- nificantly correlated with problem.drinking in Zucker's adolescent girls. This finding is supportive of the difficulty in accepting dependency relationships reported-for problem drinkers by Jones. Zucker also found an.association between cynicism and problem drinking.corroborating.Jonesi description of problem drinkers as pessi- mistic. .Jonesf problem.drinkers were described as withdrawn, while Zucker's variable of "leaving the field" which assesses such behaviors as running away, quitting.jobs and breaking dates was strongly related to problem drinking in his subjects. Zucker further found negative correlations between problem drinking and both the Socialization and Sociability scales from the California Personality Inventory. Parental Drinking The high incidence of alcoholism in the family backgrounds of alcoholics is.widely.reported in the literature (Winokur and Clayton, 1967; Sherfey, 1955; Schuckit,.gt_al., 1968; Lisansky, 1957). Riester and Zucker.(l968) found.that.high use of alcohol by teenagers was more likely when both parents drank.. The number of female alcoholics whose fathers.drink excessively.is.extremely high with most authors reporting around fifty per cent (Rathod, 1971; Kinsey, 1966, Lisansky, 1957; Boss, 1929; Wood and.Duffy,-1964;.Rosenbaum, 1958); and Parker (1972) found that sixty-nine per cent of his Spree drinkers had alcoholic fathers. Jones (1968) found someindication in her case histories that drinking.by.the opposite.sexed.parent had greater effect upon the child's future.drinking pattern than did drinking by the same sexed parent. However, Boss (1929).reported that his female alcoholics had twice as many alcoholic mothers as.did.his male alcoholics. Rathod's (1971) female alcoholics.also.had.more alcoholic mothers as compared to his male alcoholics, but the difference was slight. Jones (1971).found.that.many girls in her study reacted to an excessively drinking parent by.abstaining with disapproval as youngsters and either continue to abstain as adults or become light drinkers in accordance with their social group or spouse. Wood and Duffy (1964) suggest.that the crucial factor determining whether or not the daughter imitates an alcoholic parent is identification. If she identifies with the nonalcoholic parent who usually disapproves of the abuse (and often use) of alcohol and who has "a need to demonstrate competence,‘l she is unlikely to become a problem drinker. The early.moral training which Kinsey's (1966) female alco- holics received was.either abstinent or ambivalent. Jackson and Connor (1953) found that.ambivalence toward drinking was more prevalent among parents of.a1coholics than.among parents of non-drinkers and moderates. However, the data from Zucker's sample of adolescent girls (unpublished data) suggests that parental drinking patterns and attitudes are not related to the girls' drinking during adolescence. Parental Personality The literature tends to describe the parents of alcoholic women in deviant terms...Jones-(197l).notes that for both female and male problem drinkers,.the-mothers were rated as less agreeable, less satisfied with.life and less mentally alert than were mothers whose children later followed other drinking.patterns. Other authors have described the mothers of female alcoholics as cold and domineering (Lolli, 1953),.religious,.hostile.and bitter (Kinsey, 1966), strict and controlling (Lisansky, 1957), rigid, perfectionistic, emotionally distant, strict and unloving (Wood and Duffy, 1964), aggressive and severe (Massot, 1956),.and strict, moralistic, nagging or nervous (Rosenbaum, 1958). Wood and Duffy (1964) reported that seventy-eight per cent.of.their subjects described their mothers as dominant in the household. The fathers of alcoholic women are most frequently described as alcoholic, psychiatrically ill or absent (Kinsey, 1966; Lolli, 1953; Rathod, 1971; Rosenbaum,.l958). Their passivity or weakness is often stressed.(Lolli,.1953;.Rosenbaum, 1958;.Wood and Duffy, 1964; Kinsey, 1966). However,.their.alcoholic daughters tend to portray them in a better light.than they do their mothers (Kinsey, 1966; Karpman, 1948), and Massot (1956) noted that fathers were seen as gentle and lovable. . Although the overall.impression from the above is one of dominant mothers.and.weak fathers,.Bleuler (1955) found that either parent.may be.depicted.as.too dominant or too weak by the alcoholic daughter. Mothers!.characteristics.receive.greater focus than do fathers' characteristics in the literature probably because of the frequency of alcoholic.and.absent fathers. Jones (1971) found that the mothers' characteristics were.more closely related to the later drinking patterns of bqys than of girls and noted a cross sex effect of parental drinking...However, in Zucker's sample of adolescent girls (unpublished data), the only.parental personality variable which was correlated with problem drinking was aggressive sociability on the part of the mothers. Family Atmosphere The family atmosphere portrayed in the literature on female alcoholics suggests tension and unhappiness in these homes. We have already mentioned the high rates of alcoholism and absenteeism for the fathers of alcoholics, as well as the tendency for the mothers to be the dominant figures in the households. Broken homes are typical in the background of female alcoholics (Rathod, 1971; Lisansky, 1957; Kinsey, 1966), and Bleuler (1955) found that only ten of his fifty subjects had been raised by both parents under favorable circumstances. Wall (1937) reported.that in.thirteen cases out of fifty, the parents were incompatible.and provided an unwholesome environment for their 'children. Kinsey (1966).found a high degree of tension between the family members in the backgrounds of his subjects and noted that dis- agreement.over.re1igious.activities was characteristic of the parents. Sixty-three per cent of his subjects felt that another sibling was preferred by one or both parents. Karpman (1948), Kinsey (1966) and Widseth (1971) reported friction between the mother and daughter during adolescence. Kinsey (1966).found that his subjects tended to reject one or both of their parents and that the mother was the.parent most frequently rejected. In Widseth's (1971) sample of delinquent girls, heavy drinking girls experienced.more intense feelings of hostility and rejection toward their mothers.as well as feelings of being rejected by their mothers than did other girls. .In-his.samp1e of adolescent girls, Zucker (unpublished data) found that family tension, parental defiance, and resentment of dependency were all related to problem drinking. Child Rearing Practices Jones (1971) found that mOderate drinkers were more likely to describe their parents.as egalitarian whereas we have mentioned that female alcoholics often felt that favoritism was shown by their parents and that their mothers were strict and controlling. Rathod (1971) found that while many fathers were weak and easy-going, some were genuinely cruel. Kinsey (1966) reported that his subjects were 10 generally restricted in.regard to social activities during adolescence. He also found some evidence that parental demands were either unrealis- tically.high or low. The implication is that the child rearing practices experienced by female alcoholics may be either neglectful or harsh. Peer Influences Many female alcoholics reportedly drink at home and alone (Wood and Duffy,.l964;.Johnson,.e£;al,, 1965). Kinsey's (1966) data suggest that age may be a factor in the type of drinking situation selected. About half.of.his.subjects.did their drinking in public places, such as bars and cocktail lounges, and most of these drank with heavily.drinking.friends...Straus.(1958) found that the heavy drinkers among his college population tended to drink in groups; and Zucker's unpublished data on.adolescent girls suggest that their consumption of alcoholic.beverages is strongly.predicted by peer consumption and some- what less strongly by peer motivations for.drinking. However, problem drinking in this group of girls does not appear to be primarily under peer control, and is better predicted by internal personality influences. Riester and Zucker (1968) found that the occurrence of heavy or frequent drinking.among their high school subjects was related to drinking in a small group.of friends with no adults present. Heavy drinkers among their subjects.tended.to.belong to either the "colle- giate" group in school.which.contained.individuals who were active and successful in school-related.activities and who were headed for college or the “leather" group in school which contained adolescents who were already withdrawing from the system. Riester and Zucker suggest that 11 the "leathers" are more likely to develop into problem drinkers than are the "collegiates" who will most likely find useful and satisfying positions in society as adults. .Cahalan (1970) noted that a feeling of alienation from society appears related to problem drinking. Demographic Characteristics Demographic characteristics such as age, race, social class, parental drinking, and ethnic and.re1igious affiliation appear related to problem drinking. .Age is certainly a factor in studying drinking among adolescents since many may not have begun drinking by this period.. However, it is notable that Riester and Zucker (1968) found that ninety- one per.cent of the community's.juniors and seniors in high school had consumed.alcoholic beverages,.and.Widseth (1971) found a high incidence of problem drinking among delinquent girls only thirteen to sixteen years old. Cahalan (1970) reported a high rate of drinking connected with problems of social consequence among blacks; but Riester and Zucker (1968) found that nonwhites in their high school sample tended to be low users of alcohol.. Cahalan's high incidence of drinking- related problems among blacks may be related to social class differences. He also reported that while.there is.a larger proportion of persons with drinking problems among.those of low socioeconomic status, indi- viduals of.upper socioeconomic status.are more likely to have been exposed to heavy drinking than are the former. Riester and Zucker (1968) found high use of alcohol unrelated to lower-class background and reported that students who drank more frequently and in larger 12 quantities tended to have parents.who were semi-professionals. For his college population, Straus (1958) found that his quantity-frequency index rose as family income rose.. Cramer and Blacker (1963) compared women's reformatory inmates whose.drinking problems started in adolescence with inmates who were later problem drinkers and found that early drinkers were more likely to have.had disturbing childhood experiences and seemed more disorganized. These differences.seemed to be related to social class differences. Parker (1972) found some indication of social class differences between his more severely disturbed."spree" drinkers and his "non-spree" drinkers in a population of adult alcoholic women. We have already mentioned the consistency with which broken homes and exposure to a1co- holic parents are linked to problem drinking. Ethnic background is frequently reported as being related.to alcoholism with the well known low incidence of problem drinking among the Jews and the high incidence of problem drinking among the Irish. Kinsey (1966) and Cahalan (1970) both stress cultural differences in attitudes toward the normative use.of alcohol in explaining these dif- ferences. Anthropologists and sociologists tend to stress family and societal structure in the formation of differential vulnerability to alcoholism. McClelland (1972) suggests that for men the crucial vari- ables seem to be clear cut roles in society and ways of achieving suc- cess. The relationship of religious affiliation to problem drinking is unclear. Riester and Zucker (1968) found it unrelated to either the 13 use or nonuse of alcohol in their high school subjects, while in Zucker's.adolescent sample (unpublished data), religiosity was nega- tively related to problem drinking. However, Jones (1971) reported in her sketch of the problem drinker that the girl was affiliated with a religion which stressed judgment and punishment, and Kinsey (1966) reported that ninety-three per cent of his female alcoholics were active in religious activities during childhood. Straus (1958) found that affiliation with a religion which prohibited drinking was negatively correlated with drinking in college, but that individuals with such an affiliation who did drink tended toward heavy or problem alcohol usage. The Problem Although we have seen that a number of variables are strongly related to problem drinking, no one variable is sufficient to explain or predict its occurrence. Furthermore, factors such as impulsivity and cold, controlling mothers correlate with other disorders as well as with alcoholism, and so, while descriptive of problem drinkers, do not explain their particular choice of symptoms. What we need to know is how the various factors of adolescent psychology, parental psychology, parental drinking, family atmosphere, child rearing practices, peer influences and demographic features combine to produce problem drinking in adolescent girls. Field theory presupposes that there are multiple determinants of behavior and that the best understanding of a particular phenomenon is achieved through the discovery of the unique combination of contribut- ing factors producing it. While one or another determinant may be more 14 heavily weighted, it is rather pointless to pursue the partitioning out of factors into environmental versus personality variables when the evidence clearly indicates that both types of variables are related to the development of problem drinking. Since certain personality or back- ground factors are also related to behavior other than problems with alcohol, it follows that it is the combination of particular features rather than the presence of any one of them which culminates in the abuse of alcoholic beverages. The one study in the area which operated from a rigorously field theory approach is that of Cahalan (1970). While looking at those variables closely related to drinking behavior, he attempted to ferret out the relationships between problem area measures as well as determining which patterns of association best accounted for the variance in alcohol-related behavior. Since_his sample was confined to adults and his analyses primarily concerned with males, specific findings need not concern us here. What is of importance is his demonstration that certain combinations of variables more reliably than others predicted the occurrence of particular patterns of drinking and that some variables frequently found in association with drinking problems actually contrib- uted very little to the reduction of variance. The statistical analysis by stepwise multiple regressions which Cahalan utilized is particularly suited to analyzing a battery of items for optimal prediction while isolating the resultant correlates of the dependent variable so that their independent effects may be examined. The present study has accordingly followed his statistical as well as his theoretical approach 15 in its own investigation into the correlates of problem drinking among adolescent girls. METHOD Subjects The subjects are seventy-five adolescent girls who were selected from a drinking survey made of the entire public high school population in the area of study. They were selected so as to adequately reflect the full range of the drinking continuum from nondrinking to heavy drinking. All subjects were juniors and seniors attending the one public high school in a Middle Atlantic community which is characterized by a wide distribution of social class, ethnic, and religious groups within its population of just under 15,000 people. They are members of Zucker's (1970) cross validation sample, part of his continuing research on the development of normal and problem drinking in adolescents. A parent sample was also obtained including the natural parents of the children in the cross validation sample. Separated parents and parental surrogates were included provided that the child was living with them and that they had played an active part in raising her for three or more years. Because of the possible difficulty in obtaining parental cooperation, the initial contact was made at the family's resi- dence by trained interviewers who then set up an appointment time when both parents could complete the study materials. A follow-up and modi- fied approach using more experienced interviewers was utilized in an effort to convert outright refusals to participate. 16 17 Procedure The adolescent subjects each received five dollars for taking part in the study. They were contacted at home and requested to take part in a two-hour questionnaire session in a local church educational building. They were told that the study was interested in "teenagers' leisure time activities." The following measures were administered during the questionnaire session: 1. A background questionnaire to obtain demographic informa- tion, drinking quantity and frequency levels, frequencies of situational contexts for drinking, age of first drink, information on drinking moti- vations and indications of problem drinking. The subjects were also asked to rate themselves as hard, moderate, light or nondrinkers and then to list all those teenagers attending the session whom they knew reasonably well, following each name on their lists with an assessment of the individual's drinking habits using the same categories with which they rated their own drinking patterns. The drinking quantity and fre- quency measure was derived from Cahalan and Cisin (l969),motivations- for drinking were assessed using attitude scales taken from Grossman's cluster analytic study (Grossman, 1965), and-the Park (1962) Problem Drinking Scale was used to measure problem drinking. 2. The Bronfenbrenner (1961) Parent Activity Inventory, a measure of child rearing practices, was filled out for each parent, indicating how applicable the adolescent felt each of 100 statements was about his parents while he was growing up. This instrument provides 18 information on such instrumental and expressive activity as affection giving and parental absence or rejection. 3. The Antisocial Behavior Checklist (Kulik, 1968) was adapted to distinguish antisocial activity when drinking from antisocial behavior when not drinking. This checklist also measures experience with other kinds of drugs. 4. The Rutgers Attitude Survey, consisting of MMPI type true- false items. It contains the following scales, the first six of which were taken from Kalin and Williams (1967): a. Antisocial behavior b. Lively social presence vs. social apprehension c. Lack of order vs. adherence to order d. Aggressive sociability e. Cynicism vs. moralism f. Excitement seeking g. Marlowe-Crowne Social Desirability Scale (Crowne and Marlowe, 1960) h. The Socialization and Sociability Scales from the California Personality Inventory (Gough, 1957) i. Criticism of youth scale (assesses estrangement from peers) j. Family tension scale k. Resentment of parental dependency scale Scales i-k were taken from Moore and Holtzman (1965). The parents, who were paid ten dollars for participating in the study, completed the questionnaire materials at home with the interviewer present. The mothers and fathers worked independently. Each parent filled out the Bronfenbrenner Parent Activity Inventory for himself as applied to the adolescent in the study, the six Kalin and Williams scales included in the adolescent data collection, the Marlowe-Crowne Social Desirability Scale, a background questionnaire, and measures of parental drinking behavior. The drinking questions were 19 taken from Cahalan and Cisin (1969). A series of questions regarding "implicative drinking" taken from Haberman and Sheinberg (1967) was added to assess possible drinking problems. Statistical Analysis Selection of dependent variables: Three dependent variables all of which have some bearing on drinking behavior were investigated in terms of which of the independent variables best predicted a high score on each of these dimensions. The dependent variables are (1) Quantity-frequency which is a pure intake measure, (2) Alcoholic Anti- social Behavior which is a combined problem intake and delinquent or antiSocial behavior measure, and (3) Social Reasons for Drinking which is an external motivational measure having to do with drinking related to social pressures, and is presumed to be a less problem-oriented measure and more closely associated with normal drinking patterns than are the other two. . Missing data: Although we had complete information from our adolescent subjects, there were problems of subject attrition based on missing data from parents. For our seventy-five adolescent subjects, only'forty-five mothers and thirty-six fathers fully completed parental questionnaires. In order to assess the impaCt of this missing data, an analysis of variance was calculated to determine whether the adolescent subjects differed in their responses according to their category of parental questionnaire completion. Next a Duncan Multiple Range Analysis was performed on those variables with an F score at the .05 level of Sig- nificance to assess the magnitude of the differences between categories. RESULTS AND DISCUSSION This study utilized a multivariate approach to assess the combinatorial effects of the variables under consideration. (a) As a preliminary step, simple correlations were calculated to establish the relationship of the independent variables with the three dependent ones. (b) Next, stepwise multiple regressions were performed separately on each of the following classes of variables: adolescent personality, parental personality, parental drinking, child rearing practices, family atmosphere and demographic characteristics. This was done so as to determine which of the variables within each class contributed appre- ciably to increasing the prediction of problem drinking. (c) In order to gain a broader view of the relative importance of each class of vari- ables as well as some idea of their interaction with other classes, a second multivariate regression analysis was performed on the best predictor variables from (b) to produce the optimum pattern which would predict problem drinking in this group of adolescent girls. A third multivariate analysis was also run, consisting of stepwise multiple regressions performed on variables selected from child data only, i.e., adolescent personality, family atmosphere, adolescent perceptions of child rearing practices, and demographic variables. This analysis omitted the parent data in order to maximize the number of subjects and to provide information asnto whether the inclusion of 20 21 subjects whose parents did not submit parental questionnaires would change the nature or direction of predictions. The following is intended as a brief theoretical discussion of our results, patterned after the one other set of multivariate analy- ses in this area (Cahalan, 1970) that attempts to set the findings within a field theory (i.e., personality + family influences + social influences) framework. Table 1 indicates those variables to which adolescent sub- jects differed in their responses according to their category of paren- tal questionnaire completion. Table 1. Report of missing data analyses. 7*: “A A Categories of questionnaire completion: No father questionnaire-father at home. No father questionnaire-father divorced or separated. No mother questionnaire-mother at home. No mother or father questionnaire-both at home. . No mother or father questionnaire-divorced or separated. Mother questionnaire but father dead. Father and mother questionnaire-both at home. \IC‘U'l-hOON-I' O O O O O 0 Variables differing_according to category of questionnaire: Frequency of religious attendance: Category 4 significantly (p<.05) higher than category 7. Relevance of religion: Category 6 significantly higher than categories 1, 4 and 7. Age of first smoke: Category 3 significantly younger than categories 4, 6 and 7. Need for money: Category 6 Significantly higher than categories 5 and 7 Child's view of paternal absence: Categories 5 and 2 significantly higher than all others. Child's view of maternal affection: Categories 6 and 7 significantly lower than categories 3 and 4. Delinquent role: Category 1 significantly higher than categories 2, 3, 4, 6 and 7. 22 The impact of our missing data appears to center primarily upon the loss of information from parents and our resultant difficulty in establishing the relative importance of parental factors upon adoles- cent drinking patterns as well as the direction of their influence. For all child data variables, we can feel relatively confident that our final analysis is not unduly affected by the missing data. This is because we have separate analyses utilizing all adolescent subjects and, in comparing one analysis with another, there appears to be no measurable shift in either direction or importance of the variables. In the case of the parent data, however, we are missing information from the neglectful or largely absent parents and the parents of our more delinquent subjects. We also appear to be missing data from the parents of subjects likely to be light drinkers and abstainers so that both ends of the subject drinking continuum are underrepresented in our parent sample. This is undoubtedly the reason that we have so few significant findings with respect to parental variables as well as the reason that parental variables are not salient in our final multivariate analysis. ‘The bulk of our parental data appears to come from that great middle group who maintains and promotes "normal" patterns of alcohol consumption. Appendix A contains a more complete discussion of those variables measurably affected by our missing data and the implications of this loss for our analysis. The simple correlations between the three dependent variables suggeSt that there is a greater direct relationship between Quantity- frequency and Alcoholic Antisocial Behavior than there is between 23 either of these variables and Social Reasons for Drinking. The simple correlations between Quantity-frequency and Alcoholic Antisocial behav- ior is .59, whereas the simple correlation between Quantity-frequency and Social Reasons for Drinking is .30; and that between Alcoholic Antisocial Behavior and Social Reasons for Drinking is .34. From the simple correlations reported in Table 2 on demographic factors associated with drinking, we can see that higher alcohol con- sumption is associated with less dependence upon school, church and home as sources of financial and affiliative needs. This is in line with the suggestion in the studies by Jones (1971) and Zucker (1970) that adolescent problem drinkers are withdrawn and set themselves apart from the normal activities of adolescence. Alcoholic Antisocial Behavior correlates significantly with variables indicating an emphasis on economic activities and a lack of participation in religious or school-related activities together with an earlier and heavier use of other substances (e.g., tobacco) and a greater lifestyle of deviance. This category clearly supports Zucker's (1970) finding that heavy drinking in adolescence is associated with antisocial behavior and indicates that socially deviant behavior by problem drinkers begins in adolescence rather than in adulthood as Jones' (1971) data suggest. The earlier incidence of alcoholic acting out for our sample may reflect changes in the societal structure with ‘regard to parental supervision of teenagers since the time of Jones' data collection (i.e., circa 1938). The combination of low educational expectations and relatively poor grades indicates that the alienation from school is not limited to extracurricular activities and includes 24 :1. Table 2. Simple correlations of adolescent drinking measures with demographic variables.a Adolescent Drinking Variables Quantity Alcoholic Social frequency ASB reasons Demogrgphic Variables (N=75) Drinking rating .41** .65** .23* Age of first smoke -.19 .49** .03 Total religion -.06 .37** .28** Religious attendance -.13 .36** .08 Religious org. activity -.23* .35** .41** Relevance of religion -.02 .33** .19 Parttime job .25* .34** .03 Weekly income .18 .33** .02 Age of first drink —.11 .32** .07 Offices held -.14 .31** .28* Smoking frequency .19 .29* .03 Educational future -.08 .28* .18 Work as source of friends .27* .19 .21 Need for money -.25* .22 .27* Grade average -.10 .26* .20 Repeat courses .00 .23* .24* Member of organizations -.02 .24* .10 aOnly those variables achieving a significance level of p .05 are reported above (40% of the total). dix B. * p<.05 ** p<.01 For a complete listing, see Appen- 25 reduced performance and involvement with schoolwork. The distance that these girls appear to place between themselves and the adult figures to whom the average adolescent turns for guidance supports the implica- tion in the earlier studies that they have difficulty in establishing and accepting dependency relationships. Those girls who stress Social Reasons for Drinking display a similar independence from traditional supports, but do not exhibit the antisocial behavior found in the Alcoholic Antisocial Behavior category. They appear to be more like the problem drinking adolescents described by Jones (1971) in that they appear withdrawn and asocial, but do not display early rebellious behaviors to the same degree. The religious affiliation which she describes is lacking, however, in this as in the other two drinking classes. There is the suggestion from all three drinking measures that the premature use of alcohol may reflect a more general posture of premature adulthood. These children appear to be experiencing an early independence together with their use and abuse of adult substances. With respect to the adolescent personality and family atmos- phere variables (see Table 3), the picture is clearly one of deviant, perhaps rebellious behavior on the part of these girls, some of which appear to be directed against the parents. The Alcoholic Antisocial .Behavior measure is further correlated with a lack of order and an asocial role suggesting that these girls may be more out of control than those scoring high on the other two drinking measures. The pessimism, leaving the field, and physical aggression suggest that these girls are neither confident nor capable in their 26 Table 3. Simple correlations of adolescent drinking measures with adolescent personality variables.a Adolescent Drinking Variables Quantity Alcoholic Social frequency ASB reasons Adolescent Personality Variables (N=74) Leave the field .54** .74** .35** Excitement seeking .50** .72** .41** Delinquent role .27** .69** .26** Cynicism .34** .58** .44** Antisocial behavior .36** .57** .35** Physical aggression .44** .52** .38** Sex behavior .28* .51** .23* CPI Socialization -.40** -.45** —.37** Drug use .49** .36** .18 Stein Stability ‘ -.37** -.39** -.42** Parental defiance .27** .37** .19 Stein Optimism -.30** -.25* -.37** Stein Control -.11 -.37** -.15 Aggressive sociability .24* .33** .11 Family tension .27* .11 .15 Orderliness -.22 -.25* -.16 aOnly those variables achieving a Significance level of p<.05 are geported above (78% of the total). For a complete listing, see Appen- ix B. * p<.05 ** p<.01 27 approaches to problems, and supports Jones' (1971) depiction of the problem drinker as lacking in confidence and social skills. The family atmosphere appears to be tense and strife laden, following that reported in the literature. The acting out on the part of girls scoring high on Social Reasons for Drinking appears to be less marked and there are fewer indications of a negative perception of the family or rebellion against ' it. Nonetheless, it is present and all three measures corroborate Zucker's (1970) portrayal of the adolescent problem drinker as impul- sive, undercontrolled and deviant. Turning to the child rearing practices of fathers as perceived by their daughters, reported in Table 4, those girls scoring high on Quantity-frequency and Alcoholic Antisocial Behavior present their male parents in the negative light so frequently found in the literature on fathers of female alcoholics. Fathers are seen as absent and lacking in emotional or instru- mental support of their daughters. This is particularly the case for the Alcoholic Antisocial Behavior category and suggests that the male parent is a more important factor, at least when he does not fulfill the usual responsibilities of fatherhood, than is often assumed. The negative evaluation of father-daughter interaction is missing from the Social Reasons for Drinking correlations and the only variable to achieve significance is the use of social isolation as a means of dis- cipline. These fathers may be performing the usual paternal role to a greater extent, and their impact may lie in the shaping of conformity rather than deviant behavior. 28 Table 4. Simple correlations of adolescent drinking measures with child's perception of child rearing practices.a ..A . ¥ “E Adolescent Drinking Variables Quantity Alcoholic Social frequency ASB reasons Child's View of Mother Practices (N=73) Neglect .50** .46** .41** Instrumental companionship -.28* -.42** .14 Expressive rejection .37** .25* .35** Affiliative companionship -.24* -.33** .17 Affective reward -.20 —.31** .19 Nurturance -.24* -.31** .24* Parental absence .25* .26* .20 Deprivation of privileges .10 .00 .25* Affection -.21 -.24* .22 Affective punishment .16 .02 .24* Child's View of Father Practices (N=73) Parental presence -.23* -.37** .12 Instrumental companionship -.26* .34** .17 Affiliative companionship -.25* -.33** .06 Affection -.19 -.33** .09 Affective reward -.32** -.29* . .10 Nurturance -.20 -.28* .02 Social isolation .20 .19 .24* Paternal absence ,. .24* .03 .09 Affective punishment .13 .23* .04 A— aOnly those variables achieving a significance level of p<.05 are reported above (45% of the total for fathers and 50% of the total for mothers). For a complete listing, see Appendix B. * p<.05. ** p<.01. 29 The adolescents' perceptions of their mothers' child rearing practices conforms to the highly negative image of a cold and rejecting parent reported throughout the literature on female alcoholics. The friction between mother and daughter during the problem drinker's ado- 1escence, which is reported by Karpman (1948), Kinsey (1966) and Wid- seth (1971) is not surprising in view of these results nor is the frequently found rejection of the mother by the daughter. One would expect these girls to have difficulty in identifying with and assuming the usual female role with such deviant and distant models. While the Quantity-frequency and Social Reasons for Drinking measures stress the neglectful and punishing aspects of the mother's child rearing practices, those girls scoring high on Alcoholic Anti- social Behavior seem more acutely aware of the lack of affection and support than do the others. These girls appear to experience the most perceived difficulty with both parents, while those adolescents falling in the Social Reasons for Drinking category report negative interactions with the mother only. Our teenage subjects' perceptions of their parents indicate that they are unlikely to expect support and guidance from adults. This may in part explain their premature assumption of adult responsi- bilities and prerogatives, i.e., their relative financial and moral independence and their adoption of the sexual, drinking and smoking habits of maturity. It may be that the relationship and identification with deviant, neglectful mothers play a primary role in producing this state of affairs while the father's behavior toward his daughter results in increasing her alienation or ameliorating the effects of her 3O interaction with the mother. Viewed from this vantage point, the impulsivity and poor judgment displayed by problem drinking teenagers is understandable as the behavior of those assuming responsibilities beyond their level of competency. Turning to Table 5, we see that aggressive sociability on the part of the mother is associated with all three drinking measures. If one considers this as a prosocial form of acting out, it implies some identification with the mother by our adolescent drinkers. The only drinking measure which is associated with the mother's intake is Social Reasons for Drinking, whereas both Alcoholic Antisocial Behavior and Social Reasons for Drinking are related to the father's use of alcohol. There is some support here for Jones' (1971) suggestion that the drinking pattern of the opposite sexed parent is more influential in determining the child's drinking pattern than is that of the same sexed parent. The relationship of health problems on the part of both parents with the Alcoholic Antisocial Behavior measure may indicate that these parents are the most disturbed. Perhaps the alcoholic and psychiatrically ill fathers so often reported in the literature on female alcoholics are concentrated in this group. The low number of jobs held by these fathers may reflect period of unemploy- ment related to such problems. As Table 6 demonstrates, the only variable relating to the father's perception of child rearing practices which correlates signi- ficantly with any of the three dependent variables is that of achieve- ment demands which is associated with Social Reasons for Drinking. This is in line with the earlier suggestion of less inadequate role 31 Table 5. Simple correlations of adolescent drinking measures with parental personality and drinking indices.a LA L ‘_ AL L A Adolescent DrinkinggVariables Quantity Alcoholic Social frequency ASB reasons Paternal Personality and Drinking Variables (N=36) Combined temperate and intem- perate use .07 .08 .49** Temperate use .04 .OO .49** Usual D F .13 .12 .34* Implicative drinking .29 .30* .33* Hospital visits .20 .32* .09 Number of jobs held -.19 -.31* .02 Aggressive sociability -.16 .05 .31* Maternal Personality and DrinkingVariables (N=45) Aggressive sociability .35* .35* .38** Usual D F .17 .07 .34* Intemperate use .19 .26 .30* Excitement seeking .24 .03 .29* Health -.12 -.29* -.15 aOnly those variables achieving a significance level of p<.05 are reported above (35% of the total for fathers and 25% of the total for mothers). See Appendix B for a complete listing. * p<.05. ** p<.Ol. 32 Table 6. Simple correlations of adolescent drinking measures with a parental child rearing practices (as reported by parents). Adolescent Drinking Variables Quantity Alcoholic Social frequency ASB reasons Paternal Child Rearing Practices (N=39) Achievement demands .00 -.22 .37* Maternal Child Rearing Practices (N=47) Social isolation .36* .11 ,.O4 Affective reward -.30* -.35* -.21 Affiliative companionship -.18 -.14 -.34* aOnly those variables achieving a significance level of P<.05 are reported above (5% of the total for fathers and 15% of the total for mothers). See Appendix B for a complete listing. * P<.05. 33 performance on the part of these fathers. It is interesting to note that the adolescents stressed the lack of paternal interaction while the fathers appear to focus upon their behavior when present. It may be that fathers have relatively greater impact when they are deviant than when they conform to the expected paternal role. The differences noted thus far between girls scoring high on Social Reasons for Drinking and those scoring high on the other two drinking measures reside in the greater deviancy reported for the latter rather than in the more posi- tive attributes of the former. The mothers' perceptions of child rearing practices presented in Table 6 do not deny the emotional distance between mother and daughter indicated in the adolescent reports, but certainly suggest a more benign pattern of interaction. In comparing the perceptions of child rearing practices by parents and adolescents, it must be remembered that the parents most likely to corroborate reports of neglect and abuse are among those who failed to complete parental questionnaires. In looking over the relative contributions of each class of variables, it is clear that the parental perceptions of child rearing practices were the least salient. How much this reflects the influence of our missing data as compared to the possibility that the adolescent's perception of her childhood experiences is more crucial than either her parents' perceptions or the reality of what occurred cannot be decided on the basis of this data. Parental personality and drinking indices were next in ascending importance, but again our missing data precludes definitive statements. Parental factors did appear to be relatively more salient for girls scoring high on Social Reasons for Drinking. 34 This may reflect less friction between parents and children, with the result that there may be greater identification with and influence by parents. Adolescent perceptions of child rearing practices and demo- graphic variables appear about equally influential, but the adolescent personality variables bore the greatest significance to the three drink- ing measures. This suggests that the adolescent personality is the single most important factor in producing and describing drinking patterns, but the considerable impact of social and familial influences cannot be overlooked. The importance of the child rearing variables in predicting Alcoholic Antisocial Behavior is to be contrasted with the utility of the parent information in predicting Social Reasons for Drinking; this contrast implies that as deviance becomes more extreme, the family has less influence upon behavior. This does not deny the importance of the family in promoting the development of deviancy, but suggests that with time a deviant lifestyle increasingly becomes a closed system. Multivariate Analyses Table 7 presents the equations produced by the multivariate analyses performed within groups of subject data. These results paral- lel the simple correlations in stressing the importance of economic factors. While the score on the Alcoholic Antisocial Behavior measure is predicted primarily by purchasing power and relative economic inde- pendence, the Social Reasons for Drinking category presents a broader spectrum of alienation from social institutions and mores. These 35 Table 7. Stefiwise multiple regressions of adolescent drinking measures wit demographic var1ables* Quantityrfrequency R_ R? +Sociometric drinking rate .40 .16 Alcoholic Antisocial Behavior +Sociometric drinking rate .64 .41 +Parttime job .68 .47 -Parent money .71 ‘ .51 Social Reasons for Drinking -Religious organization activity .39 . .16 -Need for money .49 . .24 +Repetition of courses _ .54 .29 +Smoking frequency 1 .58 ' .34 +Sources of close friends .62 .38 *Significant at the p<.05 level. 36 heavier social drinking girls not only avoid commitment to church and school, but also move beyond the school and the family as a source of friends, so that it is likely that their peer group rather than any particular social setting is the focus of their loyalty and energies. The Social Reasons for Drinking category appears to be affected by a greater number of demographic factors. This may repre- sent a relatively broader influence of environmental and situational variables in producing this pattern of alcohol consumption, as compared to other drinking configurations. As Table 8 demonstrates, the adolescent personality variables predictive for all three drinking behavior categories are those which reflect an antisocial and impulsive mode of social interaction. While the Alcoholic Antisocial Behavior category appears to exhibit a more extreme version of this general pattern, the simple correlations between the independent variables reported above suggest that the appearance of any one of them signals the existence to some degree of the others. The extent to which deviant behavior is expressed would appear to be the primary differentiating factor between categor- ies. Personality influences seem to be relatively less important in predicting scores on Social Reasons for Drinking. This analysis corroborates Zucker's (1970) contention that problem drinking is closely related to other forms of antisocial behavior. Primary coping mechanisms for these drinking patterns appear to be withdrawal, escape, and striking out. Impulse control is poorly developed and there seems to be a low level of tolerance of frustra- tion. 37 Table 8. Stepwise multiple regressions of adolescent drinking measures with adolescent personality variables*. anntity-frequency R. B? +Leave the field .56 .32 +Drug use .66 .44 -Stein Control .70 .49 +Physica1 aggression .72 .52 Alcoholic Antisocial Behavior +Delinquent role .71 .50 +Cynicism ' .79 .63 +Drug use .81 .67 +Excitement seeking .84 .70 Social Reasons for Drinking +Cynicism .49 .24 +Physica1 aggression .56 .31 *Significant at the p<.05 level. 38 Neglect by the mother and lack of affective reward by the father, at least as perceived by girls, characterizes the heavier in- take subjects (see Table 9). The Alcoholic Antisocial Behavior category is predicted not only by the perceived lack of positive interactions but also by the presence of affective punishfiEnt on the part of both parents. A nega- tive perception of treatment received by parents appears important for acting out behavior. Whether this perception is causal for the behav- ior or the result of parental reaction to overt deviance is not clear. Delinquent behavior would be likely to increase existing problems of parent-child interaction. No aversive child rearing practices on the part of the father appear in the equation predicting Social Reasons for Drinking. This follows the simple correlations in suggesting that these fathers are not perceived in negative terms. The literature on adult female a1co- holics reports that many of them describe their fathers more favorably than they do their mothers. There also tends to be more conflict with the same sexed parent. The simple correlations between the independent variables presented above indicate that these parents mirror each other's neglect of and negative interaction with the daughter. This is best portrayed by the Alcoholic Antisocial Behavior measure and may be most true for the parents of these girls. While the girls scoring high on Quantity- frequency report a lack of positive interaction with both parents and those adolescents giving Social Reasons for Drinking report neglect by mothers, the Alcoholic Antisocial Behavior measure reflects both 39 Table 9. Stefiwise multiple regressions of adolescent drinking measures wit adolescent perceptions of child rearing practices.* anntity-frequency R_ R? +Neglect by mother .50 .25 -Affect reward by father .55 .30 Alcoholic Antisocial Behavior +Neglect by mother .46 .21 -Mother's presence .55 .30 +Affective punishment by mother . .62 .38 -Affiliative companionship with father .59 .34 +Affective punishment by father .66 .44 Social Reasons for Drinking +Neglect by mother .41 .17 *Significant at the p<.05 level. 4O neglectful and negative interactions with both parents. It may be that neglect precipitates increased alcoholic intake, but that anti- social behavior germinates in a more actively felt rejection of the child by one or both parents. Turning to Tables 10 and 11, we see that Social Reasons for Drinking is the only drinking measure for which the father's person- ality, drinking pattern, or perception of child rearing practices is predictive. Again, there is an implied special relationship between these fathers and their daughters. The influence of the paternal drinking pattern together with the emphasis upon achievement demands suggests the possibility of identification with the father by girls scoring high on Social Reasons for Drinking. The correlation between achievement demands made by the father and this drinking measure fur- ther indicates some interest in and interaction with daughters as well as reflecting one of the modes by which conformity behavior may be induced;‘ If love is conditional upon achievement, there is increased likelihood of efforts by the child to please others in order to gain this important reward. ‘ Aggressive sociability on the part of the mother was equally predictive of all three drinking patterns, suggesting a similarity among mothers parallel to the similarity of personality factors corre— lated with the three drinking measures. The addition of poor health for mothers of girls scoring high on Alcoholic Antisocial Behavior probably indicates a relatively poorer adaptation by these mothers-- in line with Jones' (1971) finding that the mothers of problem drinkers were less adequate in general. Nonetheless, it appears likely that 41 Table 10. Stepwise multiple regressions of adolescent drinking measures with parental perceptions of child rearing practices.* Mother Perceived Child Rearing_Practices: Qggntityefrequency R_ R? +Social isolation .36 .13 -Nurturance .47 .22 Alcoholic Antisocial Behavior -Affective reward .35 .12 +Maternal absence .46 .21 Social Reasons for Drinking_ -Affiliative companionship .34 .11 Father Perceived Child Rearing Practices: Social Reasons for Drinking +Achievement demands .38 .13 ‘*Significant at the p<.05 level. 42 Table 11. Ste ise multiple regressions of adolescent drinking measures wit parental personality and drinking variables.* Mother's Personality and Drinking: Quantity-frequency B_ R? +Aggressive sociability .35 .12 Alcoholic Antisocial Behavior tAggressive sociability .35 .12 -Hea1th .45 .20 Social Reasons for Drinking: +Aggressive sociability .38 .15 Father's Personality-and Drinking: Social Reasons for Drinking +Combined temperate and intemper- ate use of alcohol .49 .24 *Significant at the p<.05 level. 43 the differences among mothers reported in Table 10 are more quantitative than qualitative in nature. This again parallels the findings for personality factors reported for their daughters. The mothers' perceptions of their child rearing practices support the claims made by adolescent subjects that their mothers are emotionally distant and unresponsive. Mothers of daughters scoring high on Social Reasons for Drinking report a somewhat less negative picture of their child rearing practices while mothers of adolescents scoring high on Alcoholic Antisocial Behavior appear the least respon- sive of all. The mothers of girls characterized by all three drinking measures would seem likely to discourage help seeking or other dependent behavior on the part of their daughters. Without the learning that normally takes place within the mother-child relationship, one would expect these children to have difficulty in achieving successful depen- dency relationships with adults and peers outside the family. The premature and frustrated adulthood indicated in the earlier correlations with relative economic independence, a turning from the family, church and school, and the frustrated and maladaptive coping mechanisms of leaving the field and physical aggression may represent attempts to cope with this vacuum of adult responsibility and guidance. While the father's use of social isolation reported in the simple correlations for Social Reasons for Drinking may influence conformity behavior, the mother's use of social isolation together with a lack of nurturance suggests that her use of this child rearing technique represents an attempt at distancing rather than influencing her child's behavior. 44 Table 12 presents the correlation coefficient for each class of variables with the three drinking measures. The Social Reasons for Drinking measure is the most diversely influenced dependent variable with contributions from all classes of independent variables.‘ Demographic and adolescent personality factors appear most important-for Alcoholic Antisocial Behavior and Social Reasons for Drinking, while child rearing practices as perceived by mother and child and adolescent personality variables are most salient for Quantity-frequency. It appears that social and situational factors ‘may be less influential in determining alcoholic intake than they are in determining its pattern of expression. The import of the father's lack of contribution to the Quantity-frequency and Alcoholic Antisocial Behavior categories and his salience for the Social Reasons for Drinking measure may be that the more absent and deviant fathers did not complete ’parental questionnaires, whereas the more present and perhaps more ‘adequate fathers of daughters scoring high on Social Reasons for Drink- ing did. ‘There was evidence in the missing data analyses that those fathers who were perceived by their daughters as absent were less likely to have completed parental questionnaires. ' Table 13 presents the results of the multivariate analysis performed across Child Data groups. Stein control is present in the equation for Quentity-frequency only because of its relationship with ‘ leaving the field and drug usage. Its presence signifies that part of the importance of these two variables for predicting Quantity-frequency resides in their common basis of a lack of control. The coping mechan- isms of escape, withdrawal, and striking out are apparent, reiterating 45 Table 12. Multiple correlation coefficients obtained through stepwise multiple regressions of adolescent drinking measures with each class of independent variables.* Quantity Alcoholic Social frequency ASB reasons Classes of Independent Variables Demographic variables .40 .71 ' .62 Adolescent personality .72 .84 .56 Adolescent perceptions of child rearing practices .55 .66 .41 Maternal perceptions of child rearing practices .47 .46 .34 Paternal perceptions of child . rearing practices .38 Maternal personality and drinking .35 .45 .38 Paternal personality and drinking ‘, .49 *Significant at the 5 per cent level. 46 Table 13. Stepwise multiple regressions of adolescent drinking measures across classes of adolescent variables.* Quantity-frequency R_ R? +Leave the field ' .54 .29 +Drug usage .62 .38 +Child's perception of father absence .66 .44 +Physical aggression .70 .49 +Stein control .74 .54 Alcoholic Antisocial Behavior +Sociometric drinking rating .64 .41 +Cynicism .70 .49 +Parttime job .74 .55 +Child's perception of maternal neglect .76 .58 -Child's perception of affective punish- ment by mother .79 .64 -Child's perception of affiliative com- ‘ panionship with father ' .81 .67 +CPI Sociability .83 .71 +Parent money .84 1.71 Social Reasons for Drinking +Physical aggression .50 .25 +Child's perception of paternal presence .57 .33 -Re1igious organization activity _' .63 .40' -Stein Stability .67 .45 -Resentment of dependency .69 .47 . +Child's perception of expressive rejec- ' tion by mother .72 .52 *Significant at the 5 per cent level. 47 the strong relationship between asocial or antisocial behavior and high consumption of alcohol for this group. The presence of drug usage suggests that premature and excessive drinking is only one of several substance consumption indicators, and supports Riester and Zucker's (1968) finding that drug and alcohol usage are positively correlated. The child's perception of her father's absence again points to the importance of his influence with regard to drinking intake. The combination of antisocial personality factors and father absence strongly ' suggests a prominent role for fathers in the development of impulse control and prosocial behavior. Only two classes of variables are represented, adolescent personality and child perception of child rear- ing practices. 'The implication is that alcohol intake is less dependent upon environmental and social influences than it is upon the child's ' personality make-up and perception of child rearing experiences. This mirrors the relative status of the correlation coefficients reported 'for the multiple regression analyses performed within groups of subject data. 'Asking parents for money is present in the equation for Alco— holic Antisocial Behavior because of its negative relationship with having a parttime job. ‘Its presence underlines the importance of a ' parttime job for this category. The same is true for the sociability variable. It was pulled into the equation because of its negative relationship to cynicism and the child's perception of the mother as ‘neglectful. 'The combination of the child's perception of the mother as neglectful and as not using affective punishment suggests that these girls are largely without supervision and that their mothers 48 resort to physical means of discipline. ‘ThiS‘together with the child's Aperception'of her father as not providing affiliative companionship indicates that both conformity training and sex-role socialization may be neglected in these families. ‘Love withdrawal is generally considered an effective instrument for inducing conventional behavior and there is evidence that paternal interaction is important in promoting appro- 'priate female role identification and behavior. The salience of a peer perceived drinking rating (sociometric drinking rating) supports an hypothesis of more overt misbehavior on the part of these girls. It ‘appears'that'they are more completely identified with an asocial or ‘ antisocial role and more visibly so than are other teenagers. In terms ‘of having achieved a set identity and financial and moral independence from parents, the Alcoholic Antisocial Behavior category would appear ' to contain the most prematurely adult of our adolescent sample. This 'is in accord with findings by Jessor and Jessor (1975) Which indicate ' that onset of drinking in adolescence is closely associated with the ~ ‘ occurrence of other "transition" behaviors which mark the movement from ' adolescence to adulthood. "'For the Social Reasons for Drinking measure, there is a lack ' of resentment of parental intrusiveness together with a high incidence of waywardness. On a simple correlational basis, waywardness and resentment of dependency are positively related. Also on a simple 'correlational basis, waywardness and adolescent perception of the father as gbsent are positively related. This seeming contradiction suggests the possibility that waywardness on the part of these girls expresses a desire to call attention to this felt lack of parental intrusiveness 49 (or concern). If the father's presence represents demands for maturity and responsibility and the mother‘s expressive rejection an unwilling- ness to allow closeness or dependency, then these girls may be reacting against a premature and forced independence. ’The lack of'a stable adjustment at school or home (wayward- ness) together with the lack of participation in religious organizational activities supports the earlier suggestion that girls scoring high on Social Reasons for Drinking reject traditional social institutions. Some rebellion against authority figures is implied which the lack of resentment of dependency would seem to belie. Of course, the emphasis on antisocial behavior is paradoxical for girls who stress social and conforming reasons for drinking. To suggest that this drinking pattern represents a reaction formation'to unmet dependency needs does not seem unreasonable. Personality and family atmosphere factors seem to be the most salient in this analysis, followed by the child's perception of child rearing practices and lastly demographic factors. Again, this pattern appears to be the most broadly affected of the three and clearly requires a field theory approach. Turning to Table 14 for the results of the multivariate analy- sis performed across groups and including only subjects with complete data, we find that the mother's aggressive sociability score is predic- tive of her daughter's Quantity-frequency score in conjunction with an antisocial behavior orientation and the daughter's perception of neglect by her mother; Some identification by the daughter with prosocial acting out on the part of the mother is indicated. The importance of 50 Table 14. Stepwise multiple regressions of adolescent drinking measures across all classes of independent variables.* Quantityyfrequency .R R? "+Physical aggression ' .67 .45 +Leave the field .79 .62 +Child's perception of neglect by mother .84 .71 +Mother's aggressive sociability .86 .75 Alcoholic Antisocial Behavior +Cynicism .63 .40 -Age of first smoke .74 .55 -Chi1d's perception of affective reward by father .80 .65 +Father's hospitalizations ‘ .84 .71 'Social Reasons for Drinking ‘+Physical aggression .62 .38 +Child's perception of paternal presence .68 .47 -Stein Stability .76 .58 * Significant at the 5 per cent level. 51 the child's perception of her father's absence in the Child Data ' analysis is not apparent here.' Since this analysis included only sub- jects whose parents completed questionnaires, it may be that deviance I on the part of fathers must be more marked than that of mothers before it becomes salient for'increased alcoholic consumption. 'The relative importance of each class of variables present in the predictive equation for Quantity-frequency (adolescent person- ality, child's perception of child rearing practices and mother's personality) repeats the emphasis reported in Table 12 on adolescent personality factors. ' For Alcoholic Antisocial Behavior, the adolescent perception of a lack of affect reward'from the father suggests the importance of warmth and support from the father in developing conformity behavior. The cynicism may reflect a lack of rewards for positive behavior. The relationship of the father's hospital stays with this drinking measure suggests the possibility that these fathers are the most disturbed and probably the least adequate of our sample. The implication here is that fathers are important in the development of adequate impulse con- trol while mothers may be more influential in determining the direction of deviant behavior. An early onset of antisocial and deviant behavior for the Alcoholic Antisocial Behavior category is suggested by the presence of early smoking. 'As with Quantity-frequency, child data rather than parent data accounts for the majority of the variance. While one "'assume5'some impact upon child development by parental personality. "drinking'and child rearing practices, and there is a good deal of 52 support for this in our data; the parents"drinking patterns and "perceptionS'of'their'child'rearing’practices seem least predictive among our classes of variables. "The importance of the child's perception of the father's "presence*for the Social Reasons for Drinking measure is reiterated in 'this analysis.‘ Acting out on the part of these girls is evidenced by their use of physical aggression and their tendency toward waywardness. This may also indicate some masculine identification with the father as well as the possibility of rebellion against a strict and distant paternal figure. 'If there were any suggestion in the data that these fathers are harsh and punishing, one might conclude that the wayward- ness is in reaction to paternal abuse, but these fathers are the least ' negatively evaluated for all three drinking measures. GENERAL DISCUSSION In taking an overview of the results of our various analyses, the most striking and consistent feature is the relationship between asocial or antisocial behavior and the consumption of alcoholic bever- ages. The difficulty in this study in ferreting out distinctions between problem drinking and delinquency behavior suggests the former as one variant of the latter. The behavior may be more prosocial than overt delinquency in some drinking patterns and simply one mode of antisocial expression in others. While there is a general commonality in the relationships found between the independent variables and each of the three dependent variables, some differences exist in the degree to which certain behav— iors are exhibited on these threelneasures as well as in the relative salience of the independent variables for each of them. In addition, there is an implied special relationship between the fathers of girls scoring high on Social Reasons for Drinking and their daughters which is positive or neutral in distinction to the negative father-daughter interactions reported for the Quantity-frequency and Alcoholic Anti- social Behavior measures. Girls scoring high on Alcoholic Antisocial Behavior appear to be more completely and visibly identified with an asocial or antisocial role and to have the most perceived difficulty with both parents. Girls scoring high on Social Reasons for Drinking appear to exhibit less antisocial behavior and to report less friction 53 54 with parents than do girls scoring high on the other two measures. There appears to be a relatively broader spectrum of influence from the various classes of independent variables upon the Social Reasons for Drinking category while there is the implication that Quantity- frequency is less dependent upon environmental and social influences than it is upon the child's personality make-up and perception of child rearing practices. Impulse control is clearly a central issue in the development of drinking behavior as is undersocialization in general. Jones (1971) found some similarities between abstainers and problem drinkers. Among the important differences were the abstainers' emotionally controlled,' responsible, and conventional behavior in contrast to the undercontrolled and impulsive behavior displayed by problem drinkers. Our data suggest that the development of undercontrolled and impulsive behavior may ema— nate from the identification by our teenage subjects with mothers who are themselves deviant and impulsive, together with a lack of adequate sex-role socialization through instrumental or affiliative companionship with either parent and the lack of appropriate conformity or impulse 4 control training on the part of the father. In the case of the girls scoring high on Social Reasons for Drinking, the emphasis upon rebel- lious behavior in the face of father presence suggests that the father's influence may reside in his role as an authority figure rather than as an absent parent. One may speculate that the girls high in the Alcoholic Anti- social Behavior category show the absence of the father's socializing influence while adolescents high in the Social Reasons for Drinking 55 category Show the effects of a present but inflexible, paternal socializing agent. While overcontrol on the part of both parents or identification with an overcontrolled parent in reaction to an impulsive or alcoholic one (as Wood and Duffy suggest) may facilitate light drink- ing or abstention, identification with a deviant mother and over- socialization by a controlled father may encourage drinking as one expression of rebellion against the felt combination of neglect and rigid authority. The incompleteness of our parental data and the absence of findings with respect to the fathers of girls scoring high on Social Reasons for Drinking unfortunately limits us to speculation concerning the relationship of paternal variables to the three drinking measures. However, in his study of familial antecedents of responsi- bility and leadership in adolescents, Bronfenbrenner (1961) found that "It is the parent of the opposite sex who is likely to be relatively more affectionate and oversolicitous toward the child." He also noted that "It follows that girls, in contrast to boys, are in greater risk of overprotection from fathers than from mothers. In line with this fact, our data show that it is paternal presence rather than absence that correlates negatively_with the development of responsibility in girls, and it is the absence of the mother rather than the father which impairs the leadership status of the girl." A second major issue and one which is continually debated in the literature on alcoholism is that of the relationship between depen- dency strivings and the high consumption of alcohol. Another of the differences which Jones (1971) found between problem drinkers and abstainers was the abstainers' ability to accept dependency relationships. 56 Judging from the relative alienation from parents portrayed in our three drinking categories, and the further alienation from church and school associated with the Alcoholic Antisocial Behavior and Social Reasons for Drinking measures, it does not appear that these girls are able to form or maintain dependency relationships with adults. While their posture is one of independence, it does not seem to be a mature one based upon self-confidence and competency in coping with conflict. Their impulsivity and maladaptive coping mechanisms suggest that they missed the training as well as the socialization that secure and appro- priate dependence upon adults provides for the normal adolescent. Theirs is rather a premature adulthood and their posturing of mature freedoms and habits most likely the replacement and over-compensation for unmet dependency needs. Our findings from the analyses including parental variables suggest that these parents would be unlikely to encourage or, in some cases, tolerate appropriate dependency behavior in their children. No wonder, then, the cynicism toward society's offerings and the premature reliance upon oneself and one's peers for warmth and guidance. The absence of any relationship between drinking measures and indices of socio-economic status deserves some comment here. In the study by Bronfenbrenner cited above, there were four scales which failed to show any significant association between the level of education and child rearing practices of either parent. These were affection, nur- turance, affective reward and affective punishment. Given the promi- nence which these variables achieved in the present study, it is not surprising that socio-economic factors were unremarkable. While social 57 class status may have some impact upon more specific aspects of drinking, it does not appear salient for the development of problem drinking in this adolescent sample. Our findings suggest that parental training and treatment are more important in determining how the child later behaves in reference to the consumption of alcohol than are the parents' person- alities or their own drinking patterns. This may explain why research on the impact of parental behavior so often is confusing. It may be that parental drinking is influential to a marked degree only when matched with certain child rearing practices, so that particular drink- ing patterns on the part of the parents do not necessarily result in the similar use of alcohol by their children. Also, the child rearing practices which influence the abuse of alcohol do not have to come from alcohol abusing parents, but are available to all. With respect to all of the variables analyzed in the present study, there is some likelihood that our findings apply only to certain types of problem drinking development. Those girls who will begin their drinking careers following adolescence are not represented in our three drinking measures and among them may be subjects whose present behavior is dependent, compliant and conventional and whose parents may be overly attentive and protective. The necessary access routes to alcoholic beverages such as the ability to purchase beer, wine, or liquor and the unsupervised opportunities in which to imbibe may be missing during adolescence for some of those who will later become problem drinkers. We also do not know that the subjects high on our three drinking meas- ures will continue their present patterns of consumption and behavior. 58 However, our results are logically consistent with much of the literature in the area and would appear to describe some of the condi- tions surrounding the development of what may be the more visible seg- ment of the problem drinking population among women. If factors of impulse control and dependency conflicts are as important for problem drinking as this study indicates, the teenage years would seem a particularly appropriate focus for future research since the life tasks of achieving independence from parents and learning to c0pe with the larger society are central during adolescence. Inter- vention at this time, which is aimed at providing alternative adult support and guidance and which attempts to help the floundering adoles- cent cope with her early maturity rather than focusing upon strict conformity to conventional standards, appears crucial to avoid the con- solidation of antisocial and maladaptive attitudes and behaviors. Evidence of the children in need of such intervention appears to be clearly evident in our sample through their general alienation, pessim- ism, and tendencies toward leaving the field and physical aggression. Parental child rearing practices are best viewed as preventative issues. Our findings suggest that continued discussion of personality .yer§g§_familial yer§g§_environmental determinants is rather useless. It is more pertinent to turn from the contemplation of one-faceted explanations or rebuttals and emphasize the utilization of the vast research bank on adult alcoholism for preventative and treatment oriented brainstorming. Long term research which involved a preventa- tive approach of teaching parents methods of impulse control training and encouragement of assertive, self-enhancing coping mechanisms 59 followed by a later assessment of rates of problem drinking in the subject population as opposed to the general population would seem particularly appropriate. BIBLIOGRAPHY BIBLIOGRAPHY Bleuler, M. Familial and personal background of chronic alcoholics. In Diethelm, 0. (Ed.) Etiology of Chronic Alcoholism. Spring- field: Thomas, 1955. Boss, M. Zur frage der erbbiologischen bedeutung des alkohols. Mschr Psychiat Neurol 72:264, 1929. Cited in Schuckit et a1., I969. Bronfenbrenner, U. Some familial antecedents of responsibility and leadership in adolescents. In: Petrullo, L. and Bass, B. M., Leadership_and Interpersonal Behavior. New York: Holt, Rine- hart and Winston, 1961, pp. 239-271. Cahalan, D., Cisin, I. H., and Crossley, H. M. American drinking prac- tices: A national study of drinking behavior and attitudes. New Brunswick, N. J.: Publications DiViSTBn, Rutgers Center of Alcohol Studies, 1969. Cahalan, Don. Problem drinkers. San Francisco: Jossey-Bass, 1970. Cramer, C. and Blacker, E. Early and late problem drinkers among female prisoners. Journal of Health and Human Behavior, 4:282, 1963. Crowne, D. P. and Marlowe, D. A new scale of social desirability inde- pendent of psychopathology. Journal of Consulting Psychology, 1960, 24. PP. 349-354. Gough, H. G. Manual for the California Psycholggical Inventory. Palo Alto, California, Consulting Psychologists Press, 1957. Grossman, P. H. Drinking motivation: a cluster analytic study of three samples. Unpublished doctoral dissertation, Department of Psychology, University of Colorado, Boulder, 1965. (Mimeoed Research Report #28, Tri-Ethnic Research Project, University of Colorado.) Haberman, P. W. and Sheinberg, J. Implicative drinking reported in a household survey: A corroborative note on subgroup differ- ences. Quarterly_Journal of Studies on Alcohol, 1967, 28, pp. 538-543. Jackson, J. and Connor, K. Attitudes of parents of alcoholics, moderate drinkers and non-drinkers toward alcoholism. Quarterly Journal of Studies on Alcohol, 14:596-613, 1953. 60 61 Jessor, Richard, Jessor, Shirley L., and Finney, John. A social psychology of marijuana use: Longitudinal studies of high school and college youth. Journal of Personality_and Social Psychology, 26:1-15, 1973. Jessor, R. and Jessor, S. L. Adolescent development and the onset of drinking: A longitudinal study. Journal of Studies on Alco- hgl, 36:27-51, 1975. Johnson, M. W. Physician's views on alcoholism with special reference to alcoholism in women. Nebraska State Medical Journal, 50: 378-384, 1965. Jones, M. C. Personality correlates and antecedents of drinking pat- terns in adult-males. Journal of Consulting and Clinical Psychology, 32:2-12, 1968. Jones, M. C. Personality antecedents and correlates of drinking pat- terns in women. Journal of Consulting_and Clinical Psycholggy, 36:61-69, 1971. Kalin, R. and Williams, A. F. Personality self-descriptions of heavy drinking college students. Quarterly Journal of Studies on Alcohol, in press. Karpman, B. The alcoholic woman. Washington, D.C.: Linacre Press, 1948. Kinsey, B. A. The female alcoholic: a socia1_psychological study. Springfield: Charles C. Thomas, 1966. ' Kulik, J. A., Stein, K. B., and Sarbin, T. R. Dimensions and problems of adolescent antisocial behavior. Journal of Consultingand Clinical Psychology, 1968, 32:375-382. Lisansky, E. Alcoholism in women: Social and psychiatric concomitants. Quarterly Journal of Studies on Alcohol, 18:588-623, 1957. Lolli, G. Alcoholism in women. Connecticut Review of Alcoholism, 5: 9-11, 1953. ‘ ' Massot, Hamel and Deliry. Alcoolisme féminin; données statistiques et psychopathologiques. Journal of Medicine, Lyon, 37:265-269, 1956. (Abstract, Quarterly Journal of Studies on Alcohol, 18:144, 1957.) McClelland, D. C., Davis, W. N., Kalin, R., and Wanner, E. The drinking man: Alcohol and human motivation. New York: Free Press, 1972. 62 Moore, 8. M. and Holtzman, W. H. Tomorrow's Parents. Austin: University of Texas Press, 1965. Park, P. Problem drinking and role deviation. In Pittman, D. J. and Snyder, C. (Eds.), Society, culture and drinking patterns. New York: Wiley, 1962. Parker, F. B. Sex role adjustment in women alcoholics. Quarterly Jour- nal of Studies on Alcohol, 33:647-657, 1972. Rathod, N. H. and Thomson, 1. G. Women alcoholics. Quarterly Journal of Studies on Alcohol, 32:45-52, 1971. Riester, A. E. and Zucker, R. A. Adolescent social structure and drink- ing behavior. Personnel and Guidance Journal, 33:304-312, 1968. Rosenbaum, B. Married women alcoholics at the Washingtonian Hospital. Quarterly_Journal of Studies on Alcohol, 19:79-89, 1958. Schuckit, Marc, Pitts, Ferris N., Reich, Theodore, King, Lucy J., and Winokur, George. Alcoholism: 1. Two types of alcoholism in women. Archives of General Psychiatry, 20:301-306, 1969. Sherfey, M. J. Psychopathology and character structure in chronic a1co- holism. In Diethelm, 0. (Ed.), Etiology of Chronic Alcoholism. Springfield: Thomas, 1955. Straus, R. and Bacon, S. Drinking in College. New Haven: Hillhouse Press, 1958. ' Sutherland, E. and Tordella, C. Personality traits and the alcoholic: a critique of existing studies. Quarterly Journal of Studies on Alcohol, 11:559-1950. Syme, L. Personality characteristics and the alcoholic: a critique of current studies. Quarterly Journal of Studies on Alcohol, 18:288-301, 1957. Wall, J. A study of alcoholism in women. American Journal of Psychia- tr , 93:943-952, 1937. Widseth, J. C. and Mayer, J. Drinking behavior and attitudes toward alcohol in delinquent girls. International Journal of Addic- tion, 6:453, 1971. Winokur, G. and Clayton, P. Family history studies: II. Sex differences and alcoholism in primary affective illness. British Journal of Psychiatry, 113:973-979, 1967. Wood, H. P. and Duffy, E. L. Psychological factors in alcoholic women. American Journal of Psychiatry, 123:341-345, 1964. 63 Zucker, R. A. Genesis of_problem and normal drinking in adolescents. Mimeo, Psychology‘Department, Michigan State UniverSity, East Lansing, Michigan, 1970. Zucker, R. A. and Barron, F. H. Parental behaviors associated with problem drinking and antisocial behavior among adolescent males. Chafetz, M. E. (Ed.), Reprinted from the Proceedings of the First Annual Alcoholism Conference of the National Institute on Alcohol Abuse and Alcoholism. Reprinted by U.S. Department of Health, Education, and Welfare Alcohol, Drug Abuse, and Mental Health Administration - Publication # (NIH) 74-675, pp. 276-296, 1973. APPENDICES APPENDIX A MISSING DATA MISSING DATA As can be seen from Table A.l, girls whose parents reside at home and did not complete the parental questionnaires have a signifi- cantly higher Frequency of Religious Attendance than do girls whose parents reside at home and completed parental questionnaires. This is just the opposite of what one might expect assuming that church-going youngsters have parents who are concerned about their children's leisure time activities. However, it may be that these parents are relatively more traditional and conservative and that this is reflected in a resis— tance toward participation in social science research. The same relationship appears for the variable Devoutness. Also, girls whose mothers completed the parental questionnaire but whose fathers are dead are significantly more devout than are girls whose fathers reside at home and did not complete a parental questionnaire, girls whose fathers are divorced or separated and did not complete a parental questionnaire and girls whose parents reside at home and com- pleted parental questionnaires. . The variable Relevance of Religion follows the same pattern with daughters whose mothers completed the parental information but whose fathers are dead scoring significantly higher. The relationship between these three religious variables and completion of parental infonmation suggests that the parents of our more religious subjects may be underrepresented. Since there is a negative correlation between 64 65 Table A.1. Missing data analyses. Categories of questionnaire completion: No father questionnaire--father at home. No father questionnaire--father divorced or separated. No mother questionnaire--mother at home. No mother or father questionnaire--both at home. No mother or father questionnaire--divorced or separated. Mother questionnaire but father dead. Father and mother questionnaire--both at home. \IOIm-FDWNd o o o o o o 0 Variables differing_according to category of questionnaire: 1. Frequency of religious attendance Groups 1 2 3 4 5 6 7 F P X 3.00 4.00 4.25 4.58 3.00 4.57 3.41 2.30 .04 SD 2.83 1.41 1.50 .93 1.41 .79 1.62 N 2 2 4 24 2 7 34 Significantdifferences betweenygroups 4 and 7. 2. Devoutness Groups 1 2 3 4 5 6 7 F P X 2.00 2.00 2.75 2.75 3.00 3.14 2.38 2.62 .02 SD 0.00 0.00 .50 .53 0.00 .69 .70 N 2 2 4 24 2 7 34 Significant differences between groups 1, 2, 7 and 6; and between groups 4 and 7. j 3. Relevance of religion Groups 1 2 3 4 5 6 7 F P K 2.50 3.00 3.75 3.58 3.50 4.43 3.24 2.22 .05 SD .71 0.00 .50 1.06 .71 .53 .92 N 2 2 4 24 2 7 34 Significant differences between groups l,_4, 7 and 6. 66 Table A.l (Continued) 4. Occupational class Groups 1 2 3 4 5 6 7 F P X 6.5 4.5 5.5 4.33 4.0 4.29 3.44 3.34 .006 SD .71 .71 1.0 1.37 0.0 1.89 1.26 N 2 2 4 24 2 7 34 Significant differences betweengroups l and 4. 5. Age of first smoke Groups 1 2 3 4 5 6 7 F P X 15.00 15.00 11.75 17.17 15.00 17.86 16.00 3.47 .005 SD 1.41 1.41 5.06 2.20 2.83 2.34 2.42 N 2 2 4 24 2 7 34 Significant differences betweenygroups 4, 6, 7 and 3. 6. Socio-economic status Groups 1 2 3 4 5 6 7 F P X 63.5 47.5 56.5 48.2 37.0 45.6 38.7 2.93 .013 SD 2.1 4.9 8.5 11.0 19.8 16.7 13.6 N 2 2 4 24 2 7 34 Significant differences between groups 1, 3,_4 and 7. 7. Need for money Groups 1 2 3 4 5 6 7 F P K 3.5 2.5 3.5 2.9 2.0 3.6 2.6 2.44 .03 SD .7 .7 .6 .8 0.0 .8 .9 N 2 2 4 24 2 7 34 Significant differences between groups 5,y7 and 6. 67 Table A.l (Continued) 8. Child's perception ofypaternal absence Groups 1 2 3 4 5 6 7 F P X .5 12.0 3.3 3.6 11.5 2.9 5.2 3.81 .003 SD .7 4.2 3.6 2.3 10.6 1.9 4.2 N 2 2 4 24 2 7 34 Significant differences between groups 1, 3, 4, 6, 7 and 5; and between groups 1, 3, 4, 6,,7 and 2. 9. Child's perception of maternal affection Groups 1 2 3 4 5 6 7 F P X 10.0 12.0 16.8 12.8 13.5 8.3 10.4 2.97 .012 SD 5.7 1.4 2.4- 3.9 4.9 3.8 4.0 N 2 2 4 24 2 7 34 Significant differences between groups 3, 4 and 6; and between groups 3,_4 and 7. 10. Child's perception of maternal presence Groups 1 2 3 4 5 6 7 F P X 15.0 14.5 8.8 10.7 5.0 12.4 13.8 2.32 .043 SD 5.7 3.5 6.2 4.9 5.7 5.0 4.5 N 2 2 4 24 2 7 34 Significant differences betweenygroups 4, 5 and 7. 11. Delinquent role Groups 1 2 3 4 5 6 7 F P X 7.5 .5 1.8 1.5 3.5 2.7 1.8 2.47 .032 SD 9.2 .7 2.4 1.8 3.5 2.9 2.0 N 2 2 4 24 2 7 34 Significant differences between groups 2, 3, 4, 6, 7 and 1. 68 the religious variables and our drinking measures when all adolescent subjects are included in the analysis, it appears likely that we are missing data from parents of subjects at the lower end of the quantity- frequency continuum, i.e., light drinkers and abstainers. None of the above variables appeared in our final multivariate equations so that their overt relationship to our missing data is not salient. Girls were more likely to state a Need for Money when their fathers were dead and their mothers completed parental questionnaires than when their parents were divorced or separated and neither parent supplied information or when both parents were at home and completed the parental questionnaires. This may represent the relatively lower earning power of women in this country. In terms of our parent sample, it may indicate that we are missing information from parents of light drinkers and abstainers since the variable Need for Money is negatively related to our drinking categories. Those subjects whose fathers reside at home but did not com- plete a parental questionnaire scored significantly higher on the vari- able, Occupational Class, than did those subjects whose parents are both at home but did not supply parental data. Those subjects whose parents are both at home and filled out parental questionnaires scored significantly lower on Occupational Class than did girls whose parents reside at home but did not complete parental questionnaires, girls whose fathers reside at home but did not complete parental questionnaires and girls whose mothers reside at home but did not complete parental ques- tionnaires. This latter pattern was also true for the variable, Socio- economic Status, and the clear implication is that we are missing data 69 from parents with higher social class status. Neither of these variables was related to our drinking categories in any of the analyses. However, the absence of these parents from our sample would preclude the emergence of a relationship between social class and any of the other parental variables, such as child-rearing practices. Subjects were more likely to rate their mothers as present (Child's Perception of Mother's Presence) when both parents were at home and completed the parental questionnaires than when both parents were at home but did not complete questionnaires or when the parents were divorced or separated and neither parent supplied a questionnaire. While it is not at all surprising to find this association, the nega- tive correlation between Child's Perception of Mother's Presence and the drinking category, Alcoholic Antisocial Behavior, suggests that in missing data from mothers perceived as absent, we are missing informa- tion from mothers of problem drinking subjects. Subjects whose parents were both at home and completed parental questionnaires or whose fathers were dead but whose mothers completed questionnaires scored significantly higher on the variable, Affection from Mother, than did subjects whose parents were both at home but did not complete questionnaires or whose mothers were at home but did not complete the parental data. The apparent relationship between lack of perceived maternal affection and missing maternal data and our finding that the child's perception of affection from the mother is negatively correlated with problem drinking suggests that the mothers of problem drinking subjects are underrepresented in our sample. 70 Girls whose fathers are divorced or separated and did not complete a parental questionnaire and girls whose parents are divorced or separated and did not complete parental questionnaires are more likely to report Paternal Absence than are girls whose parents fall in any other category of parental questionnaire completion. Since the children of divorced parents customarily reside with the mother, this is not surprising. However, it does indicate that we are missing data from just those parents most likely to have problem drinking daughters. Child's Perception of Paternal Absence is an important variable for predicting problem drinking in our study and the survey of the litera- ture supports this finding. Adolescents whose mothers were at home but did not complete a parental questionnaire scored significantly younger on Age of First Smoke than did other subjects. Since Age of First Smoke was negatively related to the problem drinking categories as well as to other anti- social behavior variables, this indicates that we are missing parental information from mothers of problem drinking subjects. Girls whose fathers are at home but did not supply parental data score significantly higher on the variable, Delinquent Role, than do other subjects. This suggests that a neglectful or disinterested father living at home may be more influential in promoting antisocial behavior in his daughter than is the father who is altogether absent. Again, it would appear that our missing data includes material from just those parents about whom we would most like to have information. In summary, the impact of our missing data appears to center primarily upon the loss of information from parents and our resultant 71 difficulty in establishing the relative importance of parental factors upon adolescent drinking patterns as well as the direction of their influence. For all child data variables, we can feel relatively con- fident that our final analysis is not unduly affected by the missing data. This is because we have separate analyses utilizing all adoles- cent subjects and, in comparing one analysis with another, there appears to be no measurable shift in either direction or importance of the variables. In the case of the parent data, however, we are missing information from the neglectful or largely absent parents and the parents of our more delinquent subjects. We also appear to be missing data from the parents of subjects likely to be light drinkers and abstainers so that both ends of the subject drinking continuum are underrepresented in our parent sample. This is undoubtedly the reason that we have so few significant findings with respect to parental vari- ables as well as the reason that parental variables are not salient in our final multivariate analysis. The bulk of our parental data appears to come from that great middle group who maintains and promotes “normal“ patterns of alcohol consumption. APPENDIX 8 SIMPLE CORRELATIONS Table 8.1. Simple correlations of adolescent drinking measures with all demographic variables. Adolescent Drinking Variables Quantity Alcoholic Social frequency ASB reasons Demngraphic Variables (N=75) Educational future .08 .28* .18 Extended family .03 .11 .05 Nonrelatives at home .04 .09 .08 Times away from home .08 .09 .16 Months away from home .07 .03 .03 Religious discrepancy .Ol .07 .22+ Religious attendance .13 .36** .08 Religious org. activity .23 .35** .41** Devoutness .Ol .18 .07 Relevance of religion .02 .33** .19+ Father's education .13 .02 .04 Mother's education .00 .10 .08 Occupational class .01 .06 .00 Smoking frequency .19+ .29* .03 Age of first smoke .19+ .49** .03 Liquid intake .01 .11 .14 Total religion .06 .37** .28* SES .Ol .06 .06 Grade .19+ .07 .11 Drinking rating .41** .65** .23* Know peers .13 .02 .07 Peer know .07 .15 .00 72 Table 8.1 (Continued) 73 _‘ Ad01escent Drinking Variables Quantity Alcoholic Social frequency ASB reasons Age of first drink -.11 -.32** -.O7 Driver's license .14 .12 .02 1' Use of car .14 .11 .09 ‘ Own car -.07 .ll -.05 Summer job .04 .04 .20+ Parttime job .25* .34** .03 ' Money from parents -.17 -.05 -.01 Income weekly .18 .33* .02 Need for money -.25* -.22+ -.27* Grade average -.10 -.26* -.20+ Repeat courses .00 .23* .24* Varsity sports .14 -.02 -.05 Member organizations -.02 -.24* -.10 Offices held -.14 -.31** -.28* School friends -.15 -.ll -.07 Neighborhood friends .15 .14 .18 Family friends .06 -.Ol .16 Work friends .27* .19+ .21+ Other friends .02 -.l9+ -.02 Close friends .15 .19+ -.01 + p ‘<.10 * p <:.05 **p < .01 74 Table 8.2. Simple correlations of adolescent drinking measures with all adolescent personality variables. Adolescent Drinking_Veriables Quantity Alcoholic Social frequency ASB reasons Adolescent Personality Variables (N=74) Parental defiance .27* .37** .19+ Drug use .49** .36** _ .18 Sex behavior .28* .51** .23* Delinquent role .27* .69** .26* Leave the field .54** .74** .35** Physical aggression .44** .52** .38** Excitement seeking .50** .72** .41** Antisocial behavior .36** .57** .35** Social apprehension .08 -.O3 .18 Orderliness -.22+ -.25* -.16 Aggressive sociability .24* .33** .11 Cynicism .34** .58** .44** Excitement (KW) .20+ .20+ .05 CPI sociability -.05 -.12 -.15 CPI socialization -.40** -.45** -.37** Criticism of youth -.02 .06 .06 Family tension .27* .11 .15 Resent dependency .17 .14 -.02 Stein stability -.37** -.39** -.42** Stein optimism -.30** -.25* -.37** Stein control -.11 -.37** -.15 + p < .10 * p<.05 **p< .01 75 Table 8.3. Simple correlations of adolescent drinking measures with all child's perception of child rearing practices. Adolescent Drinking Variables Quantity Alcoholic Social frequency ASB reasons Child's View of Mother Practices (N=73) Physical punishment .06 -.O9 .18 Deprivation of priv. .10 .OO .25* Material reward .11 .07 .10 1 Power .05 -.ll .08 Achievement demands .11 .05 .15 Social isolation .02 -.10 .09 Instrumental companion .28* -.42** .14 Principled discipline .14 -.O9 .04 Neglect .50** .46** .41** Paternal absence .25* .26* .20+ Nurturance .24* -.31** .24* Affection .21+ -.24* .22+ Protection .01 -.11 .04 Affective punishment .16 .02 .24* Expressive rejection .37** .25* .35** Indulgence .06 .15 .01 Intercession .11 -.17 .09 Affiliative companion .24* -.33** .17 Affective reward .20+ -.31** .19+ Paternal presence .05 -.22+ .10 Table 8.3 (Continued) Adolescent Drinking Variables Quantity Alcoholic Social frequency ASB Reasons Child's View of Father Practices (N=73) Physical punishment .08 .17 .12 Deprivation of priv. .12 .16 .19+ Material reward .00 .00 .01 Power .02 .06 .02 Achievement demands .02 .03 .02 Social isolation .20+ .19+ .24* Instrumental companionship -.26* .34** -.17 Principled discipline -.21+ .10 .05 Neglect .14 .22+ .03 Paternal absence .24* .03 .09 Nurturance -.20+ .28* .02 Affection -.19+ .33** -.09 Protection .09 .04 .06 Affective punishment .13 .23* .04 Expressive rejection .20+ .13 .06 Indulgence -.09 .02 -.11 Intercession .02 .17 -.1O Affiliative companionship -.25* .33** -.06 Affective reward -.32** .29* -.1O Paternal presence -.23* .37** .12 + p<.10 * p < .05 ** p < .01 77 Table B.4. Simple correlations of adolescent drinking measures with all parental personality and drinking indices. Adolescent Drinking Variables Quantity Alcoholic Social frequency ASB reasons Paternal Personality and Drinking Verigbles (N=36) "‘ Usual 0F .13 .12 .34* : QFV class -.21 -.28+ -.27+ 3 Drinking worries -.02 -.02 -.16 i Implicative drinking .29+ .30* .33* Number of jobs -.19 -.31* -.02 Health .09 -.07 .27+ Hospital .20 .32* -.09 Frequency of religious attendance -.19 -.14 .01 Religiosity -.13 -.08 -.07 Religion total -.16 -.O9 .06 Temperate use -.04 -.OO .49** Intemperate use -.05 -.13 .03 Combined use .07 .08 .49** Antisocial behavior .01 -.O7 -.02 Social apprehension .09 .05 -.08 Orderliness .09 .23 .01 Aggressive sociability -.16 .05 .31* Cynicism .07 .02 .13 Excitement seeking -.26+ -.06 -.06 M-C social desirability .Ol -.08 -.12 78 Table 8.4 (Continued) Adolescent Drinking Variables Quantity Alcoholic Social frequency ASB reasons Maternal Personalityeend Drinking Veriables (N=45) Usual DF .17 .07 .34* QFV class -.27+ -.16 -.27+ Drinking worries .02 -.O9 .07 Implicative drinking .08 .09 .05 Number of jobs .04 -.14 -.27+ Health -.12 -.29* - -.15 Hospital -.09 -.O4 .15 Frequency of religious attendance .04 -.08 -.04 Religiosity .05 .02 -.03 Religion total .07 .01 -.09 Temperate use .10 .03 .14 Intemperate use .19 .26+ .30* Combined use -.03 -.14 -.06 Antisocial behavior .18 .15 .19 Social apprehension .11 .18 -.ll Orderliness -.O4 -.O7 .01 Aggressive sociability .35* .35* .38** Cynicism .06 .03 .03 Excitement seeking .24+ .03 .29* M-C social desirability .01 .03 -.11 + p < .10 * p<.05 ** p < .01 fl 79 Table 8.5. Simple correlations of adolescent drinking measures with parental child rearing practices (as reported by parents). Adolescent Drinking Variables Quantity Alcoholic Social frequency ASB reasons I: Paternal Child Rearing Practices (N=39) Physical punishment -.09 .04 .00 Deprivation of priv. .08 .21 .17 E Material reward .02 .08 .22 ' Power -.01 -.02 .12 Achievement demands .OO -.22 .37* Social isolation .18 .15 .12 Instrumental companion -.11 -.14 -.02 Principled discipline -.18 -.25 .10 Neglect .17 .08 .23 Paternal absence .10 .03 .20 Nurturance -.08 -.05 .24 Affection —.17 -.15 .15 Protection .03 .04 .25 Affective punishment .26+ .21 .25 Expressive rejection .05 .06 .01 Indulgence .07 -.10 .27+ Intercession -.O4 -.13 .19 Affiliative companion -.23 -.29+ .01 Affective reward .OO .05 .22 Paternal presence -.07 -.00 .04 80 Table 8.5 (Continued) Adolescent Drinking Variables ' Quantity Alcoholic Social frequency ASB reasons Maternal Child Rearing Practices (N=47) Physical punishment .12 .18 .13 r- Deprivation of priv. .10 .15 .08 1 Material reward .03 -.01 .09 ' Power -.01 -.04 -.1O ? Achievement demands .10 -.OO .14 t Social isolation .36* .11 .04 Instrumental companion -.04 -.20 -.ll Principled discipline -.04 -.25+ -.02 Neglect -.O9 .14 -.O9 Paternal absence -.04 .22 -.17 Nurturance -.19 -.17 -.20 Affection -.17 -.05 -.25+ Protection .07 .27+ -.10 Affective punishment .24+ .00 .09 Expressive rejection -.01 -.Ol -.07 Indulgence .07 -.10 .27+ Intercession -.04 -.13 .19 Affiliative companion -.23 -.29+ .01 Affective reward -.OO .05 .22 Maternal presence -.07 -.00 .04 + p < .10 * p < .05 ** p < .01 APPENDIX C DESCRIPTION OF MEASURES USED IN DATA COLLECTION Table 0.1. Description of measures used in data collection. Variable Brief Description Source A. Adolescent Data Quantity-frequency Park Problem Drink- ing Scale Alcoholic Antisocial Behavior Social Reasons for Drinking Overall alcoholic beverage consumption-standard drinks per year Questions about social com- plications of drinking and perceived alcohol depen- dency. 11 items concerned with drunkenness and antisocial behavior while drinking. An external motivational measure having to do with drinking related to social pressures and presumed to be a less problem oriented measure and more closely asSociated with normal drinking patterns than are Quantity-frequency or the Park Problem Drinking Scale. Questionnaire meas- ures of usual quan- tity and frequency; modified from Caha- lan, Cisin, and Crossley (1969). Scale developed by Park (1962). Some items from Kulik, et a1. (1968); most written for this study. Attitude scales taken from Grossman's clus- ter analytic study (1965). Antisocial or Asocial Activity Variables Drug Use 4 items from antisocial be- havior checklist concerned with barbiturates, mari- juana, etc. Essentially the same as the Kulik et a1. "Drug Use" cluster. 81 Most items from Kulik et a1. (1968). 85.nh Table C.l (Continued) 82 Variable Brief Description Source Sexual Behavior Delinquent Role Leaving the Field Serious Physical Aggression Excitement and Sen- sation Seeking C.P.I. Socializa- I tion 4 items concerned with het- erosexual activity;.e.g., petting, intercourse.. Sim- ilarity of item content determined inclusion. 5 items from antisocial be- havior checklist that ass- essed Kulik, et al.'s cluster of this name (ex- cept that all alcoholic ASB items were omitted). 5 items from antisocial be- havior checklist that ass- essed running away, quit- ting job, breaking dates. Similarity of item content determined inclusion. 6 items from antisocial be- havior checklist that ass- essed physical violence to humans and animals. Simi- larity of item content determined inclusion. 7 items from antisocial be- havior checklist that ass- essed thrill seeking, gamb- ling, joy-riding, etc. Similarity of item content determined inclusion. Personality Variables California Personality Inventory SO Scale; high scores indiEate social maturity; low scores, under-socialization. Some items from Kulik, et a1. (1968). Some especially written for this study. Items from Kulik, et a1. (1968). Items from Kulik, et a1. (1968). Items from Kulik, et a1. (1968). Items from Kulik, et a1. (1968). Gough (1956, 1957). ‘EF‘: 1.1.— . . ' Table C.1 (Continued) 83 Variable Brief Description Source C.P.I. Sociability M.C. Social Desira- bility Kalin and Williams Scales Antisocial Behavior Social Apprehension vs. Lively social Presence Adherence to Order vs. Lack of Order Aggressive Socia- bility Moralism vs. Cyni- cism Stein Cluster 1 Home Stability vs. Waywardness Stein Cluster 2 Optimism vs. Distrust Stein Cluster 3 Observance of Convention vs. Asocial Role Adolescent's Percep- tion of Father and Mother Child Rearing Practices California Personality Inventory SY Scale; high scores indTEate outgoing sociability. Assesses need for social approval. 6 factor analytically de- rived scales that all show strong relationships to heavy drinking. Cluster analyses of the Gough C.P.I. Socialization Scale; differentiates de- linquents from non- delinquents. 20 scales measuring expres- sive, disciplinary, and instrumental child-rearing practices. Separate rat- ings for the mother's and father's activities during the time the child was grow- ing up. Gough (1956, 1957). Crowne and Marlowe )3 (1960). Kalin and Williams (1966); McClelland, et a1. (1972). Stein, et a1. (1966). Bronfenbrenner (1961). Table C.1 (Continued) 83 Variable Brief Description Source C.P.I. Sociability M.C. Social Desira- bility Kalin and Williams Scales Antisocial Behavior Social Apprehension vs. Lively social Presence Adherence to Order vs. Lack of Order Aggressive Socia- bility Moralism vs. Cyni-' cism Stein Cluster 1 Home Stability vs. Waywardness Stein Cluster 2 Optimism vs. Distrust Stein Cluster 3 Observance of Convention vs. Asocial Role Adolescent's Percep- tion of Father and Mother Child Rearing Practices California Personality Inventory SY Scale; high scores indTEate outgoing sociability. Assesses need for social approval. 6 factor analytically de- rived scales that all show strong relationships to heavy drinking. Cluster analyses of the Gough C.P.I. Socialization Scale; differentiates de- linquents from non- delinquents. 20 scales measuring expres- sive, disciplinary, and instrumental child-rearing practices. Separate rat- ings for the mother's and father's activities during the time the child was grow- ing up. Gough (1956, 1957). Crowne and Marlowe (1960). Kalin and Williams (1966); McClelland, et a1. (1972). Stein, et a1. (1966). Bronfenbrenner (1961). Table C.1 (Continued) 84 Variable Brief Description Source Parental Defiance Family Tension Resentment of Depen- dency Criticism of Youth B. Parent Data Q-F-V (Quantity- frequency- variability Index of Drink- ing) Implicative Drink- ing FamilyAAtmosphere Variables 5 items from antisocial be- havior checklist that ass- essed ag ression against parents (shouting, cursing, rule breaking). One of Kulik, et al.'s clusters of delinquent activity. True-False scale of family 'quarreling, lack of famil- ial problem solving. 19 items. Used to assess lack of family solidarity. True-False scale concerned with the child's resentment of parental intrusiveness- distrust. 6 items. True-False scale that ass- essed liberalism-conserva- tism of attitudes about youth. Chosen to assess estrangement from peer group. Assesses uSual amount (quantity-frequency) of alcoholic beverage con- *sumption, with some weight- ing based on consumption of larger amounts at more in- frequent intervals. Low numbers mean high consump- tion. An index of problem drink- ing, based on questions relating to social compli- cations of alcohol use. Items from Kulik, et a1. (1968). Developed by Moore and Holtzman (1965). Developed by Moore and Holtzman (1965). Developed by Moore and Holtzman (1965). Developed by Caha- lan, et a1. (1969). Haberman and Shein- berg (1967). m FL-L‘. “73“" Table C.1 (Continued) 85 Variable Brief Description Source Drinking Worries Parent Perceptions of Child Rearing Practices Parent Personality Variables A one item index question asking how much the subject worries about his drinking (irrespective of amount consumed). Included as a measure of anxiety about drinking. 20 scales with items iden- tical to those of the adol- escents. Filled out by each parent for his or her own behavior during the time the child was growing up. See relevant adolescent variable descriptions. From the Midtown Mental Health Study; Srole, et a1. (1961). Adapted from Bron- fenbrenner (1961). Typed and Printed in the U.S.A. Proteuionel Thesis Preparation Cliff and Paula Hauhey 144 Maplewood Drive East Lansing, Michigan 48823 Telephone (517) 337-1527 HICHIGRN STATE UNIV. LIBRARIES Ill 1 312930101 9257