researchers measure divorce in
their studies. Axinn and
In this paper, I address the above limitations using a sample of ever-married respondents. I measure the actual occurrence of divorce as opposed to the subjects’ attitudes about it. Finally, I only use data pertaining to the current or first spouse. This will control for the two experiences being relationship specific to the spouse. I explicitly test to see if parental divorce explains away the association between pre-marital cohabitation and divorce.
PAST RESEARCH
Increasing Cohabitation Rates: Explanations and Resulting Consequences
According to Cherlin (2002), the majority of first marriages are preceded by cohabitation. Remarriages are even more likely than first marriages to be preceded by a period of cohabitation. There are three primary explanations for the increase of people choosing to cohabitate before marriage: technological advancements, economic changes, and different cultural norms. The introduction of birth control in 1960 and the legalization of abortion in 1973 have greatly decreased the risk of unintended pregnancies and births. The increasing number of women emphasizing their own careers and education may also lead them to delay marriage for fear it will interfere with these personal pursuits. Men may also opt for a union with less responsibility and commitment due to varying opportunities for employment. Changing cultural norms with the greater acceptance of sex outside of marriage, the greater emphasis on personal fulfillment, and the decline in religious activity have negatively impacted the marriage rates.
Thomson and Colella (1992) found that cohabiting couples may face several consequences later in their marriage due to this first transition. They were said to experience less commitment to the institution of marriage and report lower relationship quality. The women in cohabiting couples tended to have more individualistic, less traditional views. An increased likelihood of divorce is also a threat to such couples. The researchers also found that these correlations were strengthened with longer periods of cohabitation.
Brown and Booth (1996) did further research into the quality of cohabiting relationships. They found similar results to those of Thomson and Colella (1992) in that married couples reported a higher quality of relationship than did those cohabitating. They speculated that marriage plans might have an effect on the quality of the relationship amongst cohabitors. Results showed that couples intending to marry showed no lower quality of relationship than did their married counterparts; only the couples living together without plans to marry reported lower relationship quality.
Cohabitation
and Divorce
Divorce rates have raised along with pre-marital cohabitation rates (Thomson & Colella, 1992). Many view pre-marital cohabitation as a way to discover if they are compatible with their mate without the obligations of marriage. The idea is that pre-marital cohabitation should lead to lower divorce rates. To the contrary, researchers have found that pre-marital cohabitation actually leads to an increase in divorce.
Two explanations have been offered to explain this contradiction: selection and causation. The selection argument is that people who choose to cohabitate are selected from those already pre-disposed to divorce. Those more committed to the institution of marriage are more likely to move directly into marriage without cohabitating. The causation argument is that there is something inherent in living with someone that increases the likelihood of divorce. This argument also includes the idea that a dissolved cohabiting relationship may reinforce the fact that intimate relationships are fragile and can erode if not attended to properly (Axinn & Thornton, 1992).
Axinn and Thornton (1992) conducted a study to investigate the relationship between cohabitation and divorce. The data used was collected from interviews given in the years 1962, 1963, 1966, 1977, 1980, and 1985. The factors they directly considered were the importance placed on marriage, expectations of marital stability, and the acceptance of divorce. They used a two-wave study collecting attitude measures five years apart. They also identified any union formation that may have occurred during this time. Some of the questions included were those asking if married people were generally happier in life and also if divorce was the best solution when couples could not work things out. Mothers were also questioned about their views towards the marriage of their children and attitudes towards divorce.
Axinn and Thornton (1992) found results indicating support for both the argument of selection and causation. The analyses showed that cohabitation occurs with the selection of those having greater acceptance of divorce and a low commitment to marriage. The results also showed important influences from the mothers’ attitudes as well. Daughters cohabitated at significantly lower rates when their mothers stated that they would be bothered if their daughters did not marry. Sons however were not influenced as heavily by their mothers’ opinions. For sons, it was their personal commitment to marriage that reduced the likelihood of their entrance into a cohabiting union. The way these personal and parental attitudes influence who cohabitates and who opts to only marry help to support the argument of selection. Increasing acceptance of divorce from those couples cohabiting between the two waves of data collection lends support to the causation argument (Axinn & Thornton, 1992).
Influences of Parental Divorce
In a separate line of research, scholars have
been interested in whether parental divorce influences a child’s union
formation and dissolution patterns.
Others have suggested that a parental divorce will increase the likelihood of a child divorcing in adulthood. Wallerstein et al (2000) conducted a longitudinal study of children over a 25-year period. Her main hypothesis for this study was that children may experience negative consequences of their parents’ divorce as they attempt to develop and maintain their own adult relationships. Wallerstein called this delayed manifestation “the sleeper effect”. One of the greatest consequences suffered by such children is due to the fact they never had role models from which to learn how to have a lasting loving marriage. This deficit in turn leads to higher divorce rates from children of divorced parents.
CURRENT INVESTIGATION
My purpose is to understand how parental divorce impacts the lives of their children later in life as they begin to form their own relationships. I hope to further the understanding of any associations between pre-marital cohabitation and divorce, as well as any associations that exist with parental divorce. Based on previous literature about pre-marital cohabitation, divorce, and parental divorce, I have developed the following hypotheses (see Figure 1):
Hypothesis
1: There will be a positive association
between pre-marital cohabitation and divorce.
Hypothesis 2: There will be a positive association between parental divorce and the adult child’s likelihood of cohabiting.
Hypothesis 3: There will be a positive association between parental divorce and the adult child’s likelihood of divorcing.
Hypothesis 4: The association between pre-marital cohabitation and divorce will “disappear” once parental divorced is controlled.
[FIGURE 1 ABOUT HERE]
To test these hypotheses I will estimate a binary logistic regression model 1 using all independent variables, excluding parental divorce. Then, for regression model 2 I will re-run all independent variables except this time including parental divorce.
DATA
I based this study on data taken from the National Survey of Families and Households (NSFH). The NSFH is a longitudinal study consisting of three waves of data. Wave 1 was collected 1987-1988, Wave 2 in 1992-1994, and Wave 3 2001-2002. Both interviews and self-administered questionnaires were used to collect data. The dataset consists of 13,007 people coming from a cross-section of 9, 367 households. A disproportionate sampling approach was used to better include Blacks, Puerto Ricans, Mexican Americans, single-parent families, families with step-children, cohabitating couples, and newlyweds. The primary respondent was randomly selected from each household.
To obtain my sample I used Wave 1 data collected from 1987-1988. My sample contains 9,911 people, 3, 924 male and 5,987 female (see Appendix A for sample selection criteria). All respondents had to have married at least once due to the fact that I am only testing for the occurrence of an actual divorce, not the likelihood of one. I excluded cases with the age at marriage below 16 or greater than 72 (likely due to coding error). Cases with any missing values for the following variables were then excluded: respondent’s parents’ marital status, whether or not and how long the respondent cohabited with his/her current/first spouse prior to marriage, whether any cohabitation was continuous until marriage, how the respondent’s first marriage ended, race, religion, sex, highest grade level completed, number of children, date of first child birth, and the date of current/first marriage.
MEASURES
Dependent
Variable: How Respondent’s First Marriage Ended: Divorce or Not
I used an NSFH question asking how the respondent’s first marriage ended to create a dummy variable called ‘divorce’. Respondents were given the choices of divorce, separation, death of partner, or no answer. I created the ‘divorce’ variable using only those who selected divorced. I chose to be conservative and not include those who were separated, and thus not officially divorced, due to the potential of reconciliation. I acknowledge this could under-estimate the number of marital dissolutions.
Independent
Variables
Parents’ Marital History
NSFH did not directly ask for the respondent’s parents’ marital status. I created the parental divorce measure based on a series of indirect questions. Subjects were first asked if there were any time periods (up to 4 separate times) with which they did not live with their parents before the age 18. They were then asked what the reason was for the separation for each time period given. I then created four variables from this question to identify whether the reason for not living with their parents at time period 1 through 4 was due to “parents separated or divorced”. I combined these four variables to create a variable indicating whether the child ever stopped living with their parents at any time before age 18 due to a separation or divorce. I used this created variable to measure whether or not the child’s parents divorced before the child reached age 18. I expect that parental divorce will be associated with an increase in the occurrence of an adult child’s divorce.
I acknowledge possible limitations due to the creation of the above variable. It would have been ideal to have a question that simply asked the parents’ marital status, but there was not one available. One possible limitation is that a parental divorce may have occurred after the child reached 18 which would not be captured by my measure. I do not however believe this is entirely limiting. I think the effects of divorce would be greatest before the child reached age 18 as they would still be living at home where the divorce would have an impact on their everyday lives. Another possible complication is that the question did not distinguish between separation and divorce, so the parental divorce rates used could possibly be over-estimated.
Cohabitation
The cohabitation variable was coded using the following question: “Nowadays, many unmarried couples live together; sometimes they eventually get married and sometimes they don't. Did you and your (first) (husband/wife) live together before you were married?” This question was ideal because it allowed me to focus on only the respondent’s current/first marriage. Focusing only on the current/first spouse was important because it allowed me to further look into Teachman’s (2003) idea that when cohabitation is limited to one’s spouse there is no increase in divorce. I expect that pre-marital cohabitation will be associated with an increase in the occurrence of an adult child’s divorce.
I created a variable to measure the duration of the cohabitation period by subtracting the date that the couple married from the date the cohabitation began. I did this to account for the time the couple spent cohabitating, instead of limiting the union period to only that time spent in marriage. The logic behind this comes in part from DeMaris and Vaninadha Rao (1992) in which they offer the idea that the greater risk of divorce could be from the greater exposure to instability during the cohabitation period. As previously stated, Thomson and Colella (1992) found that the correlation between cohabitation and divorce were strengthened for longer periods of cohabitation. I expect duration to be positively related to divorce in that the longer the duration the higher the rate of divorce
I used a variable asking whether or not cohabitation was continuous until the time of marriage. I recoded this variable equal to one if the respondent lived with their spouse continuously without taking a break, and zero if the respondent did not live continuously with their spouse and took a break. I used this question to test if a couple having taken a break would be correlated to their later divorce. I think taking a break would reinforce the instability of a relationship leaving them more likely to experience a divorce of their own. I expect to find a negative correlation between cohabiting continuously until marriage and divorce.Marriage and ChildrenThe age at which people get married has been found to be correlated to the likelihood of divorce. Cherlin (2002) suggests that people marrying for the first time as teenagers or over age 35 have higher divorce rates than those marrying in their twenties. To test this idea I created a variable to measure the respondent’s age at marriage. I did this by subtracting the current/first marriage date from the birth date, and dividing by 12 to represent it in years. I am limiting my test to see if delaying marriage into one’s twenties is associated with less divorce. I am not testing to see if there is any association between marrying after 35 and divorce. I expect to find a negative correlation between age at marriage and divorce.
Wallerstein et al (2000) and others have suggested that the children of divorce carry the consequences of a parental divorce with them into adulthood as they try to form their own relationships. I believe most parents and other adults are aware of potential negative repercussions a divorce may have on their children when deciding how to handle marital conflicts. Therefore, it can be inferred that some couples may attempt to stay together for the sake of the children. To test this idea I created a variable to account for any children that may be present in the marriage, including both biological and those from previous relationships. I expect to find a negative association between children being present in a marriage and divorce. To further test the effects of children in a marriage, I created a second variable. Indicating whether or not a child was born under “less than ideal” conditions, which I identified as either prior to or less than 5 months after the marriage date. I included this variable to account for both the stress of step-families as they adjust to their new roles, as well as marriages that may have occurred due to pregnancy. I chose 5 months after marriage with the idea that any less passage of time would have allowed the mother ample time to know before the marriage took place that she was pregnant thus placing her into the “less than ideal” category. Respondent’s Demographic Characteristics Researchers find that divorce rates are higher for African Americans compared to Whites (Cherlin 2002; Strong et al. 2001). Due to varying rates of divorce by race, I include a dummy variable equal to one if the respondent is White, and equal to zero otherwise. I have also included the respondent’s sex as a control variable (equal to 1 if male and 0 of female). Researchers (Cherlin 2002; Strong et al. 2001) find that low education levels and low-income are associated with higher rates of divorce among Whites. One explanation is that lower family income levels are associated with higher levels of stress, potentially increasing the likelihood of divorce. I chose to include only a variable of educational attainment because I did not think the NSFH measure of income was representative of the financial situation at the beginning of the current/first marriage and educational attainment is more likely to be steady over a lifetime. Therefore, I include a measure of the highest grade level completed, measured as years of education. I recognize not using income as a possible limitation. I expect to find a negative association between education attainment and divorce. Religion has been found to be related to divorce as well. Cherlin (2002) states that Roman Catholics have lower divorce rates than other religious groups. To account for this I have included a series of dummy variables indicating the respondent’s religion (i.e. Roman Catholic, Baptist, Methodist, and Protestant). I expect any domination of religion to be negatively correlated with divorce. Strong et al (2001) presents that it is one’s involvement in religious activities, as opposed to only one’s denomination that is negatively related to divorce. I did not include a measure of religious attendance as it was measured at the present time, not the time of the first marriage. I chose to be conservative and not include this measure because of the variability of religious attendance over one’s life span. Like the income variable, I did not think it would necessarily represent the respondent’s level of religious activity at the time of his/her current/first marriage. I recognize not having this variable to be a limitation. RESULTSDescriptive Statistics The Effect of Parental Divorce on Cohabitation To investigate my second hypothesis, I have created Table 1a from my analyses. It implies that there is a correlation between ‘Parents divorced’ and ‘Cohabitated with current/1st spouse’. This can be seen by comparing the percentage of respondents with divorced parents who cohabited (28.8%) with those with parents who are not divorced yet chose to cohabit (14.3%). This suggests that parental divorce is positively related to the adult child’s likelihood of experiencing a cohabiting union. It should be noted that the cross-tab does not allow any controls so results only imply there is a relationship. [TABLES 1a & 1b ABOUT HERE]
The Effect of Parental Divorce on the Adult Child’s Divorce From my analyses I have created Table 1b to investigate my third hypothesis. From this it is implied that a parental divorce is associated with the adult child’s later divorce. This can be seen by comparing the percentage of respondents with divorced parents who later themselves divorced (40.9%) with those with parents who are not divorced yet the respondent is divorced (30.3%). This result suggests that parental divorce is positively related to a child’s later divorce. It should again be noted that the cross-tab does not allow any controls so results only imply there is a relationship. Differences Between Cohabitors and non-Cohabitors Table 2 shows differences between respondents choosing to cohabit before marriage and those who do not. First, I will discuss the dependent variable, whether or not the respondent ended their first marriage in divorce. We can see that of cohabitors 30.4% of respondents are divorced while 31.6% of those not having cohabited are divorced. From the t-test significance column we can see that those who had chosen to cohabit are not significantly different from the respondents had chosen not to cohabit (p-value > 0.10). TABLE 2 ABOUT HERE Looking at demographic variables next, we can see more differences and similarities between respondents who have cohabited before marriage and those who have not. With respect to having divorced parents, there was a significant difference; 18.5% of respondents who cohabited had divorced parents while only 8.5% of those non-cohabitors had divorced parents. With respect to race, measured as White or non-White, there is no significant difference between those having cohabited (74.6%) and those who have not (77.0%). Cohabitors are comprised of slightly more males than non-cohabitors (43.5% vs. 38.9%). Non-cohabitors are comprised of slightly more females than cohabitors (61.1% vs. 56.5%). Education was also significantly correlated with cohabitation. Cohabitors have an average of 12.74 years of school completed while non-cohabitors have only 12.15 years. Various correlations were found amongst the religious affiliations of cohabitors and non-cohabitors. Cohabitors included significantly higher numbers of Roman Catholics (27.4%) than did non-cohabitors (24.2%). Cohabitors and non-cohabitors were also significantly different in the number of Baptists included - 24.3% of non-cohabitors were Baptist while 18.8% of cohabitors were Baptist. Non-cohabitors included slightly higher numbers of Methodists than did cohabitors where (11.1% vs. 7.5%). With respect to being Protestant, there was no significant difference; 6.6% of cohabitors were Protestant while 5.9% of non-cohabitors were Protestant.
Differences among those who cohabitated and those who did not were evident with respect to age at marriage and two variables involving children. The age at which the respondent got married was significantly related to cohabitation status. The average age at marriage of those cohabiting was 24.00 years while 22.31 years was the average age at marriage of those who did not cohabit. Cohabitors and non-cohabitors were significantly different in their parenthood status. Of non-cohabitors, 85.9% had children while only 76.0% of cohabitors were parents. The involvement of children in the respondent’s relationship either prior to marriage or less than 5 months after was also significantly correlated to cohabitation. 34.9% of cohabitors had children prior to or less than 5 months after marriage while only 15.5% of non-cohabitors had children prior to or less than 5 months after marriage.
Regression Results Table 3 shows the results of the regression analyses. I used binary logistic regression models because my dependent variable was measured as a zero or one. The tables show the coefficients as either being positive or negative. A negative coefficient can be interpreted as an increase in the independent variable creates a decrease in divorce. A positive coefficient can be interpreted that as the independent variable increases divorce increases as well. [TABLE 3 ABOUT HERE]
I find a positive correlation (p<.001) between educational attainment and divorce. This means that as education increases the chances of divorce within a marriage also increase. This does not support what Strong et al (2001) stated about educational attainment being negatively correlated to divorce. I find a negative correlation (p<.005) between being Roman Catholic and divorce. This further supports Cherlin (2002) in his statement that Roman Catholics have lower divorce rates than other religious groups. My results supported Cherlin’s (2002) statement that people marrying in their twenties are less likely to get divorced than those marrying as teenagers. I find that age at marriage is negatively associated (p<.001) with divorce. This implies that there may be some advantage to waiting until the twenties to get married. Again, I did not test to see if there was an association between respondents marrying beyond 35 and divorce. I used a variable accounting for any children born prior to or less than 5 months after marriage to measure potential added stress on the marriage from children. I find that having children before or less than 5 months after marriage is positively related (p<.001) to divorce. This is supportive of the idea that bringing children into a marriage in ‘less than ideal’ circumstances has a positive impact on the chances of divorce. The data does not support the hypothesis that pre-marital cohabitation would have a significant positive relationship to divorce (p-value > 0.10). Contrary to what was expected, these results suggest that living together before marriage does not significantly increase the risk of marital dissolution when the cohabitation is limited to the current/first spouse. Due to the fact that cohabitation was found not to be associated with divorce as hypothesis 1 states, the spurious relationship of hypothesis 4 could not be tested. Regression models 1 and 2 in Table 3 were intended to show results for hypothesis 4. Should my hypothesis have been supported I would have found there to be a positive relationship (p<.010) between cohabitation and subject divorce in model 1 but not in model 2 when parental divorce was tested as well. CONCLUSION and DISCUSSION This paper analyzes the relationship between pre-marital cohabitation and divorce. It also assesses the relationship between parental divorce and the adult child’s entry into a cohabiting union and the likelihood of later divorce. This study was also designed to test the effects of parental divorce on the association between pre-marital cohabitation and divorce. Previous studies have been limited by the methodology of assessing the relationship between pre-marital cohabitation and divorce. Axinn and Thornton (1992) used two waves of data to test the probability of divorce despite the fact that the two waves were only five years apart. Teachman (2003) presented the idea that past research had been limited by the measures of cohabitation. He found that when taking into account only cohabitation to one’s spouse, divorce was no more likely than their non-cohabiting counterparts and suggested further research look into this finding. I have improved upon both the above limitations. I have measured the current marital status and retrospectively looked back at cohabiting and marital unions. I have also addressed the second limitation by limiting all my measures to only the current/first spouse. Results show that pre-marital cohabitation does not have a positive correlation to divorce. This is not consistent with past research that cohabitation is somehow selective of couples more prone to divorce or that cohabiting somehow lessens one’s commitment to marriage (Axinn and Thornton, 1992). My results do, however, lend support to Teachman’s (2003) idea that when cohabitation is limited to one’s spouse there is no increase in divorce. Do these results imply that people should choose to cohabit before marriage to ‘test out’ their relationship? Although my results indicate it would not necessarily have a positive impact on the likelihood of divorce, I must also note it does not implicate it has a negative impact either. There was no significant relationship found between pre-marital cohabitation and divorce. This instead suggests that further research should be done to investigate if cohabiting with a particular spouse is associated with divorcing that particular person. Further research would ideally address all structural limitations as well. I find that parental divorce has a positive correlation to the child’s later divorce. This finding is consistent with previous research that parental divorce likely increases the probability that the child will experience a divorce of their own (Wallerstein et al, 2000). I also find that parental divorce is positively correlated with a child entering into a pre-marital cohabiting union. This too is consistent with previous research (Thomson & Colella, 1992). A cautionary note about this study is that not all measures were ideal. I was forced to create measures with the questions that were asked by NSFH. Despite these limitations, this study further advances the knowledge of factors leading to divorce. My study suggests that cohabiting before marriage does not lead to higher divorce rates. The experience of a parental divorce appears to be more closely associated with the child’s divorce. In a time where divorce rates are higher than most would prefer, it is important to discover potential predictive factors. Only after people are aware of what circumstances and experiences may lead to divorce can anyone hope to avoid it. REFERENCES
Axinn, William G and Thornton, Arland (1992). The Relationship Between Cohabitation
and Divorce: Selectivity or Causal Influence? Demography, 29, 3, 357-374.
Brown, Susan L and Booth, Alan (1996). Cohabitation Versus Marriage: A Comparison
of Relationship Quality. Journal of Marriage and the Family, 58, August, 668-678.
Cherlin, Andrew J.
(2002). Public and Private
Families 3rd Ed.
Hill. 254-267.
DeMaris, Alfred and Vaninadha Rao, K. (1992). Premarital Cohabitation and
Subsequent Marital Stability in the
Divorce Magazine:
http://www.divorcemag.com/statistics/statsUS.shtml
Strong,
Wadsworth/Thomson Learning. 506-508.
Teachman, Jay (2003). Premarital Sex, Premarital Cohabitation, and the Risk of
Subsequent Marital Dissolution Among Women. Journal of Marriage and Family, 65, May, 444-455.
Thomson, Elizabeth and Colella, Ugo (1992). Cohabitation and Marital Stability:
Quality or Commitment? Journal of Marriage and the Family, 54, May, 259-267.
Thornton, Arland (1991). Influence of the Marital History of Parents on the Marital and
Cohabitational Experiences of Children. American Journal of Sociology, 96, 4, 868-894.
Wallerstein, J. S., Lewis, J. M., & Blakeslee, S. (2000). The Unexpected Legacy of
Divorce.
Figure 1.
Hypothesized Relationship between Parental Divorce, Pre-Marital
Cohabitation, and Divorce
|
Table 1a.
Parental Divorce and Cohabitation |
|||
|
|
Subject’s parents divorced |
Subject’s parents did not divorce |
Total |
|
Cohabitated with current/1st
spouse |
289 28.8% |
1271 14.3% |
1560 15.7% |
|
Did not cohabitate with
current/1st spouse |
713 71.2% |
7638 85.7% |
8351 84.3% |
|
Total |
1002 100.0% |
8909 100.0% |
9911 100.0% |
|
Table 1b.
Parental Divorce and Subject’s Divorce |
|||
|
|
Subject’s parents divorced |
Subject’s parents did not divorce |
Total |
|
Subject divorced |
410 40.9% |
2703 30.3% |
3113 31.4% |
|
Subject did not divorce |
592 59.1% |
6206 69.7% |
6798 68.6% |
|
Total |
1002 100.0% |
8909 100.0% |
9911 100.0% |
|
Table 2. Descriptive Statistics |
|||||
|
Did subject (R) cohabitate
with current/1st spouse? |
|||||
|
|
Variable |
Yes |
No |
T-Test Significance |
Hypothesized Relationship
with Divorce |
|
Dependent
Variable |
|
|
|
|
|
|
|
R
is divorced |
30.4
% |
31.6
% |
|
|
|
Independent
Variables |
|
|
|
|
|
|
|
R’s
parents are divorced |
18.5
% |
8.5
% |
*** |
+ |
|
Race |
White |
74.6
% |
77.0
% |
|
|
|
Sex |
Male (1) |
43.5
% |
38.9
% |
*** |
|
|
|
Female
(0) |
56.5
% |
61.1
% |
||
|
Education |
Highest
level of Education |
12.74 |
12.15 |
*** |
- |
|
Religion |
Roman
Catholic |
27.4
% |
24.2
% |
* |
- |
|
|
Baptist |
18.8
% |
24.3
% |
*** |
- |
|
|
Methodist |
7.5
% |
11.1
% |
*** |
- |
|
|
Protestant |
6.6
% |
5.9
% |
|
- |
|
Age
|
At
time of marriage |
24.00 |
22.31 |
*** |
- |
|
Kids |
Current/1st
Spouse had kids coming into marriage or 1st child was born less
than 5 months after marriage |
34.9
% |
15.5
% |
*** |
+ |
|
|
Currently
a parent |
76.0
% |
85.9
% |
*** |
- |
|
Length
of Cohabitation |
How
long did R cohabitate with spouse? (years) |
1.52 |
0 |
|
+ |
|
|
Did
R live with spouse continuously until marriage? |
15.5
% |
0
% |
|
- |
|
n = 9,911 |
|||||
|
*** p<.001 **p<.005 *p<.010 |
|||||
|
Table 3.
Logistic Regression Results |
|||||||
|
|
Model 1 |
|
Model 2 |
||||
|
Variable |
B |
S.E. |
Exp (B) |
|
B |
S.E. |
Exp (B) |
|
R’s parents are divorced |
|
|
.326 *** |
.072 |
1.386 |
||
|
R cohabitated with
current/1st spouse |
-.024 |
.084 |
.976 |
|
-.054 |
.084 |
.947 |
|
White |
.154 * |
.059 |
1.167 |
|
.160 * |
.059 |
1.173 |
|
Sex |
.147 ** |
.049 |
1.158 |
|
.147 ** |
.049 |
1.159 |
|
Highest level of education |
.071 *** |
.009 |
1.074 |
|
.072 *** |
.009 |
1.074 |
|
Roman Catholic |
-.193 ** |
.061 |
.825 |
|
-.188 ** |
.061 |
.829 |
|
Baptist |
.040 |
.061 |
1.040 |
|
.038 |
.061 |
1.039 |
|
Methodist |
-.164 |
.080 |
.849 |
|
-.161 |
.080 |
.852 |
|
Protestant |
.164 |
.096 |
1.179 |
|
.151 |
.096 |
1.163 |
|
Age at marriage |
-.148 *** |
.007 |
.863 |
|
-.146 *** |
.007 |
.864 |
|
Current/1st
spouse had kids coming into marriage or 1st child was born less
than 5 months after marriage |
.874 *** |
.062 |
2.396 |
|
.861 *** |
.062 |
2.366 |
|
Currently a parent |
-.047 |
.069 |
.954 |
|
-.038 |
.069 |
.963 |
|
How long did R cohabitate
w/spouse? (years) |
-.028 |
.037 |
.972 |
|
-.031 |
.037 |
.969 |
|
Did R live w/spouse
continuously until marriage? |
.140 |
.157 |
1.150 |
|
.135 |
.157 |
1.145 |
|
Constant |
1.338 *** |
.191 |
3.811 |
|
1.253 *** |
.192 |
3.502 |
|
n = 9911 |
|||||||
|
Appendix A.
Sample Size |
|
|
Original / Total |
13,007 |
|
Exclude those who have
never married |
10,596 |
|
Exclude those with missing
data |
9,911 |
|
Sex of R (of 9,911) |
|
|
Male |
3,924 |
|
Female |
5,987 |