Perceived Causes of Divorce: An Analysis of Interrelationships MARGARET GUMINSKI CLEEK University of Wisconsin Center,Washington County T.ALLAN PEARSON* Comprehensive Services of Ozaukee County,Wisconsin Interrelationships between perceived causes of divorce were investigated utilizing factor analysis.In a sample of 275 males and 336 females,seven dimensions of divorce,underly- ing 18 possible contributing causes,were revealed.Significant differences were found be- tween the sexes both in the frequencies with which causes were identified and in the com- position of the seven factors.This suggests the need to look beyond single causes,exploring constellations of problems separately for each sex. Since many interacting factors are likely to con- demographic data provided by county,state,and tribute to a divorce,multicausal descriptions that federal bureaus which rarely include reasons for focus on the interrelationships between these fac- divorce,other than the legal ones. tors should be more illuminating than unitary Smaller scale studies have examined the cause- conceptions of divorce (Price-Bonham and of-divorce question more directly through infor- Balswick.1980). mation provided by interviews and surveys.They The literature contains many large-scale studies have tended,however,to concentrate on a fairly that have focused on the demographic character- limited number of divorce causes and,more im- istics of divorcing individuals,such as age,educa- portantly,have examined these causes in isolation tional level,income,length of marriage,and so on rather than in interaction. (Kop,1976;Spanier and Glick,1981;Thomson, The major objective of the current study is to 1976;Vigderhaus and Fishman,1978;Weed, combine the information from these two types of 1974).These studies,while supplying valuable in- research in order to investigate the interrelation- formation on the probabilities of divorce for ships between perceived divorce causes on a scale groups of individuals sharing particular character- comparable to that of some demographic studies. istics,generally do not provide data on the per- sonal reasons for divorce.This is understandable METHODS given that such studies usually rely on aggregate The Wisconsin no-fault divorce law,which went into effect in 1978,requires petitioners and The data reported in this paper were collected in 1980 requests respondents to attend a divorce counsel- and 1981.An earlier version of this paper was presented ing session conducted by the county or by a pri- at the annual meeting of the Midwestern Psychological vate therapist. Association in Chicago,May 1983. At the counseling sessions conducted for indi- viduals with minor children in Ozaukee County, Department of Psychology,University of Wisconsin Center-Washington County,400 University Dr.,West Wisconsin in 1980 and 1981,single-page surveys Bend,WI 53095. including demographic data items as well as cause-of-divorce items were distributed to the par- *Mental Health,Alcohol,and Drug Abuse Coordi- ticipants.Demographic data items on the survey nator,Comprehensive Services of Ozaukee County, concerned sex,age,length of marriage,years of Ozaukee County Courthouse,121 W.Main,Port education,number of previous marriages,and Washington,WI 53074. number of children.Eighteen possible divorce February 1985 JOURNAL OF MARRIAGE AND THE FAMILY 179 This content downloaded from 211.80.94.134 on Mon,19 Dec 2016 05:27:44 UTC All use subject to http://about.jstor.org/terms
Perceived Causes of Divorce: An Analysis of Interrelationships MARGARET GUMINSKI CLEEK University of Wisconsin Center, Washington County T. ALLAN PEARSON* Comprehensive Services of Ozaukee County, Wisconsin Interrelationships between perceived causes of divorce were investigated utilizing factor analysis. In a sample of 275 males and 336females, seven dimensions of divorce, underly- ing 18 possible contributing causes, were revealed. Significant differences were found be- tween the sexes both in the frequencies with which causes were identified and in the com- position of the seven factors. This suggests the need to look beyond single causes, exploring constellations of problems separately for each sex. Since many interacting factors are likely to con- tribute to a divorce, multicausal descriptions that focus on the interrelationships between these fac- tors should be more illuminating than unitary conceptions of divorce (Price-Bonham and Balswick, 1980). The literature contains many large-scale studies that have focused on the demographic character- istics of divorcing individuals, such as age, educa- tional level, income, length of marriage, and so on (Kop, 1976; Spanier and Glick, 1981; Thomson, 1976; Vigderhaus and Fishman, 1978; Weed, 1974). These studies, while supplying valuable in- formation on the probabilities of divorce for groups of individuals sharing particular character- istics, generally do not provide data on the per- sonal reasons for divorce. This is understandable given that such studies usually rely on aggregate The data reported in this paper were collected in 1980 and 1981. An earlier version of this paper was presented at the annual meeting of the Midwestern Psychological Association in Chicago, May 1983. Department of Psychology, University of Wisconsin Center-Washington County, 400 University Dr., West Bend, WI 53095. *Mental Health, Alcohol, and Drug Abuse Coordi- nator, Comprehensive Services of Ozaukee County, Ozaukee County Courthouse, 121 W. Main, Port Washington, WI 53074. demographic data provided by county, state, and federal bureaus which rarely include reasons for divorce, other than the legal ones. Smaller scale studies have examined the cause- of-divorce question more directly through infor- mation provided by interviews and surveys. They have tended, however, to concentrate on a fairly limited number of divorce causes and, more im- portantly, have examined these causes in isolation rather than in interaction. The major objective of the current study is to combine the information from these two types of research in order to investigate the interrelation- ships between perceived divorce causes on a scale comparable to that of some demographic studies. METHODS The Wisconsin no-fault divorce law, which went into effect in 1978, requires petitioners and requests respondents to attend a divorce counsel- ing session conducted by the county or by a pri- vate therapist. At the counseling sessions conducted for indi- viduals with minor children in Ozaukee County, Wisconsin in 1980 and 1981, single-page surveys including demographic data items as well as cause-of-divorce items were distributed to the par- ticipants. Demographic data items on the survey concerned sex, age, length of marriage, years of education, number of previous marriages, and number of children. Eighteen possible divorce February 1985 JOURNAL OF MARRIAGE AND THE FAMILY 179 This content downloaded from 211.80.94.134 on Mon, 19 Dec 2016 05:27:44 UTC All use subject to http://about.jstor.org/terms
TAble 1.MEAN AGE.YEARS OF EDUCATION. demographic measures.Compared with Wiscon- AND LENGTH OF MARRIAGE FOR MALES AND FEMALES IN 1980 AND 1981 sin as a whole,Ozaukee County has a population with a higher median income and a larger per- Survey Mean Mean Years of Mean Length centage of its inhabitants involved in manufactur- Date Age Education of Marriage ing.As can be seen in the table,Wisconsin,and 1980 Ozaukee County as well,have a somewhat lower Males 35.1 12.8 113 divorce rate than the United States average. Females 322 11.9 11.5 1981 Perceived Divorce Causes Males 34.8 12.2 11.7 Females 32.3 12.2 11.7 Table 3 consists of the survey list of 18 per- ceived divorce causes,the percentages of males and females who checked each problem,and the rank of each problem by sex.Males checked an causes generated from clinical interaction with average of 3.2 problems and females 4.2 prob- divorcing persons also were listed.Any combina- lems.Other researchers also have found that tion of the 18 causes could be indicated as being important in an individual's divorce action. women tend to make more marital complaints than men (Levinger,1966;Kitson and Sussman, Names were not requested on the forms. 1982). In 1980,508 individuals were asked to attend the sessions;335 attended.In 1981,429 persons As can be seen in Table 3,communication problems was the most frequently indicated cause were invited,and 304 attended.Although comple- tion of the surveys was not required,96%%of those for both sexes:69.7070 of females and 59.30%of males selected it.Basic unhappiness and incom- attending in 1980-147 males and 176 females- patibility were ranked second and third by both did so.In 1981,surveys were completed by 95 sexes.There were,however,a number of sex dif- of those attending-128 males and 160 females. ferences in problems indicated.Females checked FINDINGS basic unhappiness (10.63;p<.002),in- compatibility (X=8.46;p <.004),emotional Demographic Information abuse (X2 59.51;p .001),alcohol abuse- Table I shows the mean age,years of educa- spouse (x?21.76;p <.001),infidelity- tion,and length of marriage for the male and spouse (x=10.63;p<.002),and physical female participants in 1980 and 1981.Of those abuse (x2=42.28;p <.001)as contributing fac- participants filing for divorce in 1980,90.7%%had tors to their divorce more often than did males not been married previously;in 1981,80.30 were Males checked alcohol abuse-self (x'=24.38; previously unmarried.Survey respondents in 1980 p<.001)and women's lib (x=26.83;p< had an average of 2.3 children;in 1981 this .001)more often than did females.It should be average was 2.2. noted that,although all participants had minor Since same-sex participants did not differ sig- children,children were mentioned as a problem nificantly on these demographic characteristics by only 4.4%%of the males and 8.9%of the for the two years of data collection,the data for females:this difference between the sexes was the two years were combined and analyzed sepa- statistically significant (x=5.01;p <.02). rately for each sex. Problems rarely mentioned by either sex were Ozaukee County,located in southeastern Wis-drug abuse-self and spouse,infidelity-self, consin,is compared in Table 2 with the state of mental illness,and religious differences. Wisconsin and the United States on various These differences between the sexes as to the TABLE 2.COMPARISON OF OZAUKEE COUNTY WITH WISCONSIN AND THE UNITED STATES ON VARIOUS DEMOGRAPHIC MEASURES Ozaukee United Measures County Wisconsin States 1975 divorce rate per 1,000 population 2.5 2.9 4.9 1970 population urban 67.3 65.9 73.5 1974 per capita income ($ 5,608 4,468 4,572 1969 median family income (S) 12,620 10.065 9,586 1975 employees in manufacturing 47.9 37.9 30.3 1975%employees in wholesale and retail 21.9 26.5 27.4 1970%one-unit housing structures 82.6 70.5 69.1 Source:U.S.Department of Commerce,Bureau of Census,County and City Databook,1977. 180 JOURNAL OF MARRIAGE AND THE FAMILY February 1985 This content downloaded from 211.80.94.134 on Mon,19 Dec 2016 05:27:44 UTC All use subject to http://about.jstor.org/terms
TABLE 1. MEAN AGE, YEARS OF EDUCATION, AND LENGTH OF MARRIAGE FOR MALES AND FEMALES IN 1980 AND 1981 Survey Mean Mean Years of Mean Length Date Age Education of Marriage 1980 Males 35.1 12.8 11.3 Females 32.2 11.9 11.5 1981 Males 34.8 12.2 11.7 Females 32.3 12.2 11.7 causes generated from clinical interaction with divorcing persons also were listed. Any combina- tion of the 18 causes could be indicated as being important in an individual's divorce action. Names were not requested on the forms. In 1980, 508 individuals were asked to attend the sessions; 335 attended. In 1981, 429 persons were invited, and 304 attended. Although comple- tion of the surveys was not required, 96% of those attending in 1980-147 males and 176 females- did so. In 1981, surveys were completed by 95% of those attending-128 males and 160 females. FINDINGS Demographic Information Table 1 shows the mean age, years of educa- tion, and length of marriage for the male and female participants in 1980 and 1981. Of those participants filing for divorce in 1980, 90.7% had not been married previously; in 1981, 80.3% were previously unmarried. Survey respondents in 1980 had an average of 2.3 children; in 1981 this average was 2.2. Since same-sex participants did not differ sig- nificantly on these demographic characteristics for the two years of data collection, the data for the two years were combined and analyzed sepa- rately for each sex. Ozaukee County, located in southeastern Wis- consin, is compared in Table 2 with the state of Wisconsin and the United States on various demographic measures. Compared with Wiscon- sin as a whole, Ozaukee County has a population with a higher median income and a larger per- centage of its inhabitants involved in manufactur- ing. As can be seen in the table, Wisconsin, and Ozaukee County as well, have a somewhat lower divorce rate than the United States average. Perceived Divorce Causes Table 3 consists of the survey list of 18 per- ceived divorce causes, the percentages of males and females who checked each problem, and the rank of each problem by sex. Males checked an average of 3.2 problems and females 4.2 prob- lems. Other researchers also have found that women tend to make more marital complaints than men (Levinger, 1966; Kitson and Sussman, 1982). As can be seen in Table 3, communication problems was the most frequently indicated cause for both sexes: 69.7% of females and 59.3% of males selected it. Basic unhappiness and incom- patibility were ranked second and third by both sexes. There were, however, a number of sex dif- ferences in problems indicated. Females checked basic unhappiness (X2 = 10.63; p < .002), in- compatibility ( X2 = 8.46; p < .004), emotional abuse ( X2 = 59.51; p < .001), alcohol abuse- spouse (X2 = 21.76; p < .001), infidelity- spouse (X2 = 10.63; p < .002), and physical abuse ( X2 = 42.28; p < .001) as contributing fac- tors to their divorce more often than did males. Males checked alcohol abuse-self ( X 2 = 24.38; p < .001) and women's lib (x2 = 26.83; p < .001) more often than did females. It should be noted that, although all participants had minor children, children were mentioned as a problem by only 4.4% of the males and 8.9?% of the females; this difference between the sexes was statistically significant (X2 = 5.01; p < .02). Problems rarely mentioned by either sex were drug abuse-self and spouse, infidelity-self, mental illness, and religious differences. These differences between the sexes as to the TABLE 2. COMPARISON OF OZAUKEE COUNTY WITH WISCONSIN AND THE UNITED STATES ON VARIOUS DEMOGRAPHIC MEASURES Ozaukee United Measures County Wisconsin States 1975 divorce rate per 1,000 population 2.5 2.9 4.9 1970 % population urban 67.3 65.9 73.5 1974 per capita income ($) 5,608 4,468 4,572 1969 median family income ($) 12,620 10,065 9,586 1975 %o employees in manufacturing 47.9 37.9 30.3 1975 % employees in wholesale and retail 21.9 26.5 27.4 1970 % one-unit housing structures 82.6 70.5 69.1 Source: U.S. Department of Commerce, Bureau of Census, County and City Databook, 1977. 180 JOURNAL OF MARRIAGE AND THE FAMILY February 1985 This content downloaded from 211.80.94.134 on Mon, 19 Dec 2016 05:27:44 UTC All use subject to http://about.jstor.org/terms
TABLE 3.PERCENTAGES OF MALES AND variables for each sex are"pure"in that they load FEMALES INDICATING EACH OF 18 PERCEIVED DIVORCE CAUSES (RANK) on one factor only,while 7 variables in each analysis load on two factors.The only variable Perceived Divorce Cause Females Males that loads on more than two factors is sexual Communication problems,which in the male analysis appears in problems 69.7% 1) 59.3% 1) four factors.Unlike the other variables,then,the Basic unhappinessa 59.9 2) 46.9 (2) variable of sexual problems as perceived by males Incompatibilitya 56.4 (3) 44.7 (3) Emotional abusea 55.5 4 24.7 is not circumscribed in its effect but contributes to (6) Financial problems 32.9 (5) 28.7 (5) a number of different constellations of problems. Sexual problems 32.1 (6) 30.2 (4) This result is compatible with the findings of Lev- Alcohol abuse-spousea 30.0 (7) 58 (14) inger (1966)and Rhyne (1981). Infidelity-spousea 25.2 (8) 10.5 (9) Physical abusea 21.7 (15) A comparison of the seven factors for each sex (9) 3.6 In-laws 10.7 (10 11.6 (8 listed in Table 4 reveals clear similarities in the Childrena 8.9 (11) 4.4 (16) composition of two factors,although the relative Religious differences 8.6 (12) 6.5 (12) importance of each factor is different for each Mental illness 5.0 (13) 6.9 (11) Drug abuse-spouse 3.9 sex. (14) 1.4 (17) Infidelity一self 3.9 (15) 6.2 (13) Factor I for females,Interpersonal Interaction. Women's libb 3.0 (16) 14.5 (7) is quite similar to Factor 5 for males.In the Alcohol abuse- -selfb .9 (17) 9.4 (10) female analysis,this factor includes the variables Drug abuse-self 3 (18) 1.1 (18) of incompatibility and basic unhappiness,both of aProblems indicated significantly more often by which load on this factor only.Communication females. bProblems indicated problems,the most frequently indicated problem significantly more often by males. for both sexes,and sexual problems also load on Factor I for females.The comparable factor for males,Factor 5,is identical except that emotional perceived causes of divorce were not unexpected. abuse replaces communication problems as a per- Levinger(1966)reported that women were signifi- ceived divorce cause.The difference in the relative cantly more likely to cite physical abuse,financial importance of these factors may indicate that,for problems,alcohol abuse,and lack of love,while females,problems involving communication and husbands complained more often of in-law prob- affect have a salience that they do not have for lems or sexual incompatibility. males. Although Table 3 indicates the percentages of Another area of comparability can be found in subjects who perceived various problems as im- Factor 2 for females and Factor 3 for males.Each portant in their divorce,it does not reveal the in- of these factors has been named Abuse.Variables terrelationships between these divorce causes.In that constitute the female factor are emotional order to investigate reasons for divorce within a abuse and alcohol abuse-spouse,both of which multicausal framework,separate factor analyses load on this factor only;while physical abuse, were performed for each sex financial problems and alcohol abuse-self each load on one other factor as well.The male factor Factor Analysis Results of abuse is quite similar to this one.Alcohol Table 4 presents the factors extracted by these abuse-self (which loads only on this factor for factor analyses.A varimax rotation,preceded by males),financial problems,and physical abuse are a principal factor analysis,was used to provide a present in the male analysis as well.In the male simplified,orthogonal factor structure.Variables abuse factor,emotional abuse and alcohol or causes within a factor are listed in Table 4 in abuse-spouse are absent,and the variable of sex- descending order of the strength of their loading ual problems is present.In light of other research on that factor. on alcoholism,the composition of both of these Seven factors were extracted in each case.Each factors suggests that alcohol abuse by the male is factor can be considered an underlying dimension the key variable,with physical abuse (Gelles, or interacting combination of causes of divorce. 1974),financial problems,and sexual problems The identification of each of these underlying probably attributable to the male alcohol prob- dimensions stems from the examination of the lem.The attribution of the problem to male alco. variables-or perceived causes-that load signifi- holism and not to female alcoholism is supported cantly on the factor,as well as the factorial com- also by the composition of Factor 5 for females. plexity of the variables.Factorial complexity Female alcoholism appears again here but does refers to the number of factors on which a par- not occur with emotional abuse or physical abuse ticular variable loads significantly.In this study 10 and is negatively related to financial problems. February 1985 JOURNAL OF MARRIAGE AND THE FAMILY 181 This content downloaded from 211.80.94.134 on Mon,19 Dec 2016 05:27:44 UTC All use subject to http://about.jstor.org/terms
TABLE 3. PERCENTAGES OF MALES AND FEMALES INDICATING EACH OF 18 PERCEIVED DIVORCE CAUSES (RANK) Perceived Divorce Cause Females Males Communication problems 69.7% (1) 59.3% (1) Basic unhappinessa 59.9 (2) 46.9 (2) Incompatibilitya 56.4 (3) 44.7 (3) Emotional abusea 55.5 (4) 24.7 (6) Financial problems 32.9 (5) 28.7 (5) Sexual problems 32.1 (6) 30.2 (4) Alcohol abuse-spousea 30.0 (7) 5.8 (14) Infidelity-spousea 25.2 (8) 10.5 (9) Physical abusea 21.7 (9) 3.6 (15) In-laws 10.7 (10) 11.6 (8) Childrena 8.9 (11) 4.4 (16) Religious differences 8.6 (12) 6.5 (12) Mental illness 5.0 (13) 6.9 (11) Drug abuse-spouse 3.9 (14) 1.4 (17) Infidelity-self 3.9 (15) 6.2 (13) Women's libb 3.0 (16) 14.5 (7) Alcohol abuse-selfb .9 (17) 9.4 (10) Drug abuse-self .3 (18) 1.1 (18) aproblems indicated significantly more often by females. bproblems indicated males. significantly more often by perceived causes of divorce were not unexpected. Levinger (1966) reported that women were signifi- cantly more likely to cite physical abuse, financial problems, alcohol abuse, and lack of love, while husbands complained more often of in-law prob- lems or sexual incompatibility. Although Table 3 indicates the percentages of subjects who perceived various problems as im- portant in their divorce, it does not reveal the in- terrelationships between these divorce causes. In order to investigate reasons for divorce within a multicausal framework, separate factor analyses were performed for each sex. Factor Analysis Results Table 4 presents the factors extracted by these factor analyses. A varimax rotation, preceded by a principal factor analysis, was used to provide a simplified, orthogonal factor structure. Variables or causes within a factor are listed in Table 4 in descending order of the strength of their loading on that factor. Seven factors were extracted in each case. Each factor can be considered an underlying dimension or interacting combination of causes of divorce. The identification of each of these underlying dimensions stems from the examination of the variables-or perceived causes-that load signifi- cantly on the factor, as well as the factorial com- plexity of the variables. Factorial complexity refers to the number of factors on which a par- ticular variable loads significantly. In this study 10 variables for each sex are "pure" in that they load on one factor only, while 7 variables in each analysis load on two factors. The only variable that loads on more than two factors is sexual problems, which in the male analysis appears in four factors. Unlike the other variables, then, the variable of sexual problems as perceived by males is not circumscribed in its effect but contributes to a number of different constellations of problems. This result is compatible with the findings of Lev- inger (1966) and Rhyne (1981). A comparison of the seven factors for each sex listed in Table 4 reveals clear similarities in the composition of two factors, although the relative importance of each factor is different for each sex. Factor 1 for females, Interpersonal Interaction, is quite similar to Factor 5 for males. In the female analysis, this factor includes the variables of incompatibility and basic unhappiness, both of which load on this factor only. Communication problems, the most frequently indicated problem for both sexes, and sexual problems also load on Factor 1 for females. The comparable factor for males, Factor 5, is identical except that emotional abuse replaces communication problems as a per- ceived divorce cause. The difference in the relative importance of these factors may indicate that, for females, problems involving communication and affect have a salience that they do not have for males. Another area of comparability can be found in Factor 2 for females and Factor 3 for males. Each of these factors has been named Abuse. Variables that constitute the female factor are emotional abuse and alcohol abuse-spouse, both of which load on this factor only; while physical abuse, financial problems and alcohol abuse-self each load on one other factor as well. The male factor of abuse is quite similar to this one. Alcohol abuse-self (which loads only on this factor for males), financial problems, and physical abuse are present in the male analysis as well. In the male abuse factor, emotional abuse and alcohol abuse-spouse are absent, and the variable of sex- ual problems is present. In light of other research on alcoholism, the composition of both of these factors suggests that alcohol abuse by the male is the key variable, with physical abuse (Gelles, 1974), financial problems, and sexual problems probably attributable to the male alcohol prob- lem. The attribution of the problem to male alco- holism and not to female alcoholism is supported also by the composition of Factor 5 for females. Female alcoholism appears again here but does not occur with emotional abuse or physical abuse and is negatively related to financial problems. February 1985 JOURNAL OF MARRIAGE AND THE FAMILY 181 This content downloaded from 211.80.94.134 on Mon, 19 Dec 2016 05:27:44 UTC All use subject to http://about.jstor.org/terms
TABLE 4.RESULTS OF FACTOR ANALYSES ON 18 PERCEIVED DIVORCE CAUSES FOR MALES AND FEMALES Females Males 1.Interpersonal Interaction (12.7) 1.Drug Abuse (15.3%) Incompatibility Drug abuse-self Basic unhappiness Drug abuse-spouse Sexual problems Mental illness Communication problems Physical abuse Women's lib 2.Abuse (9.1%) 2.Various Differences (9.1%) Emotional abuse In-laws Alcohol abuse-spouse Emotional abuse Physical abuse Communication problems Financial problems Religious differences Alcohol abuse-self Financial problems 3.Infidelity (7.7%) 3.Abuse (7.9) innidy—38mue Alcohol abuse-self Infidelity- Financial problems Drug abuse一spouse Physical abuse Sexual problems 4.Religion7.6明o) 4.Independence (7.2) Religious differences Infidelity-spouse Children Women's lib Physical abuse (-)a Sexual problems Communication problems 5.Female Alcohol Abuse (6.7) 5.Interpersonal Interaction (6.5%) Alcohol abuse-self Basic unhappiness Financial problems (-)a Incompatibility Mental illness (-)a Sexual problems Emotional abuse 6.1n-laws(6.3%) 6.Female Alcohol Abuse (6.1) In-laws Children Infidelity-self Alcohol abuse-spouse Communication problems(-)a Sexual problems 7.Independence (6.2) 7.Infidelity (5.7%) Women's lib Infidelity-self Sexual problems Religious differences Children Basic unhappiness aThe different signs for divorce causes loading on a particular factor indicate that the causes are related to that factor in opposite directions;i.e.,for Factor 4,females who checked the problems of religious differences and chil- dren tend not to check physical abuse,and vice versa. For males,female alcoholism (Factor 6)occurs combination in the female analysis.Something to with the variables of children and sexual prob- note in the makeup of Factors 1 and 2 for males is lems. the predominance of external,concrete problems The factor that accounts for the most variance Mental illness,in-laws,religious differences, in the male data,Drug-Abuse (Factor 1),had no financial problems,women's lib,and drugs are comparable factor in the female data.Found in examples of these problems.Perhaps this indi- combination with the two drug-abuse components cates that some males look more to influences for males are mental illness,which loads on this outside of the relationship to explain their marital factor only,physical abuse,and women's lib. disruption than they do to problems involving af- Three of these divorce causes do not appear to be fective interaction with their wives. crucial in many relationships in this sample,as Several factors that have been given identical Table 3 indicates:the drug abuse and mental ill- factor names in the male and female analyses are ness causes were indicated by only a small per- nevertheless different in composition.The factor centage of individuals.However,in the cases named Infidelity,Factor 3 for females and Factor where they were indicated,they were found in 7 for males,is one of these.In both these factors, combination. infidelity by the male is indicated,but the similari- Factor 2 for males provides another example of ty ends there.Females tend to associate male in- a constellation of variables that does not appear in fidelity with male drug abuse and their own in- 182 JOURNAL OF MARRIAGE AND THE FAMILY February 1985 This content downloaded from 211.80.94.134 on Mon,19 Dec 2016 05:27:44 UTC All use subject to http://about jstor.org/terms
TABLE 4. RESULTS OF FACTOR ANALYSES ON 18 PERCEIVED DIVORCE CAUSES FOR MALES AND FEMALES Females Males 1. Interpersonal Interaction (12.7%) Incompatibility Basic unhappiness Sexual problems Communication problems 2. Abuse (9.1 %) Emotional abuse Alcohol abuse-spouse Physical abuse Financial problems Alcohol abuse-self 3. Infidelity (7.7%) Infidelity-spouse Infidelity-self Drug abuse-spouse 4. Religion (7.6%) Religious differences Children Physical abuse (-)a 5. Female Alcohol Abuse (6.7%) Alcohol abuse-self Financial problems (_)a Mental illness (-)a 6. In-laws (6.3 %) In-laws Infidelity-self Communication problems (-)a 7. Independence (6.2%) Women's lib Sexual problems Children 1. Drug Abuse (15.3%) Drug abuse-self Drug abuse-spouse Mental illness Physical abuse Women's lib 2. Various Differences (9.1%) In-laws Emotional abuse Communication problems Religious differences Financial problems 3. Abuse (7.9%) Alcohol abuse-self Financial problems Physical abuse Sexual problems 4. Independence (7.2%) Infidelity-spouse Women's lib Sexual problems Communication problems 5. Interpersonal Interaction (6.5o%) Basic unhappiness Incompatibility Sexual problems Emotional abuse 6. Female Alcohol Abuse (6.1%) Children Alcohol abuse-spouse Sexual problems 7. Infidelity (5.7%) Infidelity-self Religious differences Basic unhappiness aThe different signs for divorce causes loading on a particular factor indicate that the causes are related to that factor in opposite directions; i.e., for Factor 4, females who checked the problems of religious differences and chil- dren tend not to check physical abuse, and vice versa. For males, female alcoholism (Factor 6) occurs with the variables of children and sexual prob- lems. The factor that accounts for the most variance in the male data, Drug-Abuse (Factor 1), had no comparable factor in the female data. Found in combination with the two drug-abuse components for males are mental illness, which loads on this factor only, physical abuse, and women's lib. Three of these divorce causes do not appear to be crucial in many relationships in this sample, as Table 3 indicates: the drug abuse and mental ill- ness causes were indicated by only a small per- centage of individuals. However, in the cases where they were indicated, they were found in combination. Factor 2 for males provides another example of a constellation of variables that does not appear in combination in the female analysis. Something to note in the makeup of Factors 1 and 2 for males is the predominance of external, concrete problems. Mental illness, in-laws, religious differences, financial problems, women's lib, and drugs are examples of these problems. Perhaps this indi- cates that some males look more to influences outside of the relationship to explain their marital disruption than they do to problems involving af- fective interaction with their wives. Several factors that have been given identical factor names in the male and female analyses are nevertheless different in composition. The factor named Infidelity, Factor 3 for females and Factor 7 for males, is one of these. In both these factors, infidelity by the male is indicated, but the similari- ty ends there. Females tend to associate male in- fidelity with male drug abuse and their own in- 182 JOURNAL OF MARRIAGE AND THE FAMILY February 1985 This content downloaded from 211.80.94.134 on Mon, 19 Dec 2016 05:27:44 UTC All use subject to http://about.jstor.org/terms
fidelity.For males,their own infidelity is associ- REFERENCES ated with religious differences and basic unhappi- Gelles,R. ness. 1974 The Violent Home:A Study of Physical Ag- Another set of factors,Factor 7 for females and gression Between Husbands and Wives.Bever- Factor 4 for males has been named Independence ly Hills,CA:Sage Publications. In the female factor,women's lib is associated Kitson,G.C.and Sussman,M.B. with children,while in the male factor,infideli- 1982 "Marital complaints,demographic character- ty-spouse and communication problems are istics,and symptoms of mental distress in found in combination with women's lib.These divorce."Journal of Marriage and the Family two factors may reflect a somewhat different view 44(February):87-101. of women's liberation by males and females. Kop,P. 1976 "Age of marriage and divorce trends in CONCLUSION Amsterdam during the period 1911-71.''Jour- nal of Biosocial Science 8:137-143. In this paper interrelationships between 18 per- Levinger,G. ceived causes of divorce were investigated utilizing 1966 "Sources of marital dissatisfaction among ap- the technique of factor analysis.Significant dif- plicants for divorce."American Journal of ferences between the sexes were found in the fre- Orthopsychiatry 36(October):803-807. quencies with which 9 of the 18 causes were indi- Price-Bonham,S.and Balswick,J.O. cated,as well as in the composition of the 7 fac- 1980 "The noninstitutions:divorce,desertion and remarriage."Journal of Marriage and the tors for each sex. Family 42 (November):959-972. This study attempts to generate hypotheses con- Rhyne,D. cerning divorce.The factor analysis technique 1981 "Bases of marital satisfaction among men and allows the examination of perceived divorce women,"Journal of Marriage and the Family causes in a multicausal framework and so 43 November):941-954. broadens the understanding of divorce beyond Spanier,G.B.and Glick,P.C. simpler conceptions.It suggests that research in 1981 "Marital instability in the United States:some this area must investigate constellations of per- correlates and recent changes."Family Rela- tions31:329-338. ceived problems rather than single causes of divorce and that these sets of problems must be Thomson,K.S. 1976 "The divorce profile:differential social corre- examined separately for each sex.It is also possi- lates in 1952 and 1972.''International Journal ble that such constellations are different for of Sociology of the Family 6:253-263 groups of individuals sharing particular demo- U.S.Bureau of the Census graphic characteristics such as age,income,and 1977 County and City Databook.Washington,DC: years of marriage. U.S.Dept.of Commerce Vigderhaus,G.and Fishman,G. 1978 Social indicators of marital instability.USA. 1920-1969.'Social Indicators Research 5: 325-344. Weed,J.A. 1974 "Age at marriage as a factor in state divorce rate differentials."Demography 11(3):361- 375. February 1985 JOURNAL OF MARRIAGE AND THE FAMILY 183 This content downloaded from 211.80.94.134 on Mon,19 Dec 2016 05:27:44 UTC All use subject to http://about.jstor.org/terms
fidelity. For males, their own infidelity is associ- ated with religious differences and basic unhappi- ness. Another set of factors, Factor 7 for females and Factor 4 for males has been named Independence. In the female factor, women's lib is associated with children, while in the male factor, infideli- ty-spouse and communication problems are found in combination with women's lib. These two factors may reflect a somewhat different view of women's liberation by males and females. CONCLUSION In this paper interrelationships between 18 per- ceived causes of divorce were investigated utilizing the technique of factor analysis. Significant dif- ferences between the sexes were found in the fre- quencies with which 9 of the 18 causes were indi- cated, as well as in the composition of the 7 fac- tors for each sex. This study attempts to generate hypotheses con- cerning divorce. The factor analysis technique allows the examination of perceived divorce causes in a multicausal framework and so broadens the understanding of divorce beyond simpler conceptions. It suggests that research in this area must investigate constellations of per- ceived problems rather than single causes of divorce and that these sets of problems must be examined separately for each sex. It is also possi- ble that such constellations are different for groups of individuals sharing particular demo- graphic characteristics such as age, income, and years of marriage. REFERENCES Gelles, R. 1974 The Violent Home: A Study of Physical Ag- gression Between Husbands and Wives. Bever- ly Hills, CA:Sage Publications. Kitson, G. C. and Sussman, M. B. 1982 "Marital complaints, demographic character- istics, and symptoms of mental distress in divorce." Journal of Marriage and the Family 44 (February):87-101. Kop, P. 1976 "Age of marriage and divorce trends in Amsterdam during the period 1911-71." Jour- nal of Biosocial Science 8:137-143. Levinger, G. 1966 "Sources of marital dissatisfaction among ap- plicants for divorce." American Journal of Orthopsychiatry 36 (October):803-807. Price-Bonham, S. and Balswick, J. 0. 1980 "The noninstitutions: divorce, desertion and remarriage." Journal of Marriage and the Family 42 (November):959-972. Rhyne, D. 1981 "Bases of marital satisfaction among men and women." Journal of Marriage and the Family 43 (November):941-954. Spanier, G. B. and Glick, P. C. 1981 "Marital instability in the United States: some correlates and recent changes." Family Rela- tions 31:329-338. Thomson, K. S. 1976 "The divorce profile: differential social corre- lates in 1952 and 1972." International Journal of Sociology of the Family 6:253-263. U.S. Bureau of the Census 1977 County and City Databook. Washington, DC: U.S. Dept. of Commerce. Vigderhaus, G. and Fishman, G. 1978 "Social indicators of marital instability. USA. 1920-1969." Social Indicators Research 5: 325-344. Weed, J. A. 1974 "Age at marriage as a factor in state divorce rate differentials." Demography 11(3):361- 375. February 1985 JOURNAL OF MARRIAGE AND THE FAMILY 183 This content downloaded from 211.80.94.134 on Mon, 19 Dec 2016 05:27:44 UTC All use subject to http://about.jstor.org/terms