Chapter 5 Quantitative study
5.5 Quantitative study on interethnic couples
In the previous sections, all the variables (i.e. love, satisfaction, couple cultural identity, and personal commitment) and their measurements for the newly proposed cultural model have been looked at in detail. In this section, a quantitative study on Chinese/non-Chinese couples using the cultural model will be presented.
Hypothesis
In section 5.1, the cultural model was introduced, in which depicted the
hypothesis of the quantitative study. The hypothesis is that love, satisfaction (i.e. dyadic adjustment), and couple cultural identity (i.e. acculturation to the partner and similarity of couple’s individualism/collectivism) will predict personal commitment and each will account for unique variance in personal commitment of Chinese/non-Chinese
interethnic couples. The relationships between love, satisfaction, couple cultural identity and personal commitment are: the more love, the more personal commitment; the more satisfaction (i.e. the more dyadic adjustment), the more personal commitment; the more congruent couple cultural identity (more acculturation to the partner and more similarity of couple’s individualism/collectivism), the more personal commitment.
In order to test this hypothesis, the following methods will be used on the dyadic data from both partners. First, a correlation analysis will be carried out to see whether the correlations between love, satisfaction (dyadic adjustment), couple cultural identity and personal commitment are significant and in the right direction for men and women respectively.
Second, a standard multiple regression will be carried out to see how much variance love, satisfaction (dyadic adjustment) and couple cultural identity account for in personal commitment, whether the regression model is significant, and whether each variable significantly predicts personal commitment for men and women, respectively. In order to see whether each variable accounts for unique variance in personal
commitment, the significance of the regression model and the coefficients of each variable with personal commitment will be examined. If the model is significant and each coefficient is significant, then each variable accounts for unique variance in personal commitment.
176
Third, the sample of the current research study consists of both partners, so partner effect will also be looked at through standard multiple regression to determine whether partner variables (the partner’s love, satisfaction, and couple cultural identity) significantly predicted one’s own personal commitment together with one’s own love, satisfaction and couple cultural identity. Fourth, demographic variables might have effects on personal commitment, so these will be tested using t-tests and ANOVA. Whether men’s and women’s horizontal and vertical individualism and collectivism tendencies have effects on personal commitment will also be tested by ANOVA. Hierarchical multiple regressions will be carried out in order to see how well love, satisfaction (dyadic adjustment) and couple cultural identity predict personal commitment with the demographic variables being controlled for men and women respectively (the demographic variables are those which have been found having significant effect on personal commitment). In order to have more knowledge of the sample, t-tests and descriptive statistics will be carried out to see the difference between partners’ scores on each variable in the cultural model and both partners’
individualism/collectivism tendencies. All above analyses will utilise Statistical Package for the Social Sciences (SPSS) Version 19 (IBM Corp., New York, USA).
Finally, structural equation modelling (SEM) will be used to test the fitness of women’s and men’s commitment with all their significant predictors. This analysis will utilise Linear Structural Relationships (LISREL) 8.80 Student (Scientific Software International, Inc).
Method
Method of finding the participants
The participants of this research, Chinese married to/cohabit with non-Chinese individuals, were approached in the United Kingdom first. The response rate was not very high, so participants were then approached in the United States, as there were more Chinese/non-Chinese heterosexual couples in the United States than in the United Kingdom. Couples in the interview study were approached first (five couples from the interview study have completed the questionnaires), and then couples were approached in the China towns and Chinese communities in the United Kingdom (only a few questionnaires back). Couples were also approached from all the occasions that the
177
researcher was attending. Couples from personal contacts (i.e. through the researcher and the researcher’s social networks) had a significant high response rate, which means that if there was a person who knew both the researcher and the couples, either directly or indirectly, would encourage the couples to participate. Couples in the United States were approached through the researcher’s personal contacts either living in the United States or having contacts living in the United States. The researcher much relied on the contacts in the United States to actually approach the couples in the United States, as the researcher was in the United Kingdom. In order to increase the sample size of the study, all the participants were asked to approach or recruit other eligible couples (snowball method), however the message could only pass onto one further step in a few occasions.
Participants
The participants were 25 Chinese/non-Chinese heterosexual couples in the United Kingdom and 12 Chinese/non-Chinese heterosexual couples in the United States. Of the 37 couples, 84% were married and 16% were cohabiting. For the married couples, the marriage duration ranged from 0.5 to 34 years with a mean of 9.87 years. 70.3% of women were in their first marriage and 13.5% of women were in their second marriage; 64.9% of men were in their first marriage, 8.1% of men were in their second marriage and 10.8% were in their third marriage. For the cohabiting couples, the cohabiting duration ranged from 1 to 5 years with a mean of 2.66 years. To consider the marriage duration and cohabiting duration (for cohabiting couples and for married couples who had cohabited before marriage) together, the total duration couples living together ranged from 1 to 34 years with a mean of 9.51 years. Women’s age ranged from 21 to 64 with a mean age of 39.32 years. Men’s age ranged from 20 to 70 with a mean age of 44.02 years. The duration of living in the United Kingdom or United States for women ranged from 1 to 41 years with a mean of 12.28 years. The duration of living in the United Kingdom or United States for men ranged from 11 to 63 years with a mean of 36.66 years. Thirty-four women and 15 men were immigrants to the United Kingdom or United States, and they were all first generation immigrants. The ethnicities of the women consisted of 86.5% Chinese, 10.8% White, and 2.7% other ethnicities. The parents of women were all from the same ethnic group as the women. All of the female participants grew up in the country of their birth. The ethnicity of the men consisted of 13.5% Chinese, 83.8% White, and 2.7% Asian (Indian, Pakistani, Bangladeshi or other).
178
With one exception, all the men were from the same ethnic group as their parents. Almost all of the male participants (91.8%) grew up in the country of their birth. Of all the couples, 67.6% were middle class, 18.9% were upper middle class, and 13.5% were lower middle class. For the socio-economic status of the family that each partner came from (i.e. their parents’ socio-economic status in the country that they were living), 61.1% couples in which the partners had the same family socio-economic background, 19.4% couples in which women had a higher socio-economic background than the partner’s, and 19.4% couples in which women had a lower socio-economic background than the partner’s. For women’s education, 54.1% had postgraduate degrees, 35.1% had first degrees, and 10.8 % had less than the first degree. For men’s education, 43.2% had postgraduate degrees, 32.4% had first degrees, and 24.3% had less than the first degree. 48.6% couples in which partners had similar levels of education, 32.4% couples in which women had a higher educational level than the partner’s and 18.9% couples in which women had a less educational level than the partner’s. 62.2% of women and 40.5% of men did not have any religion, 24.3% of women and 32.4% of men were Christians, and 13.5% of women and 27% of men had other religions.
Measurements (APPENDIX 2 and APPENDIX 3) and scoring
Couples were asked to fill in the measurement of romantic love (Rubin, 1970), Commitment scale by Rusbult et al.(1998), Dyadic Adjustment Scale (Spanier, 1976), couple cultural identity scale (questions extracted from acculturation scales and Triandis’s (1996) horizontal and vertical individualism and collectivism scale) and demographic questions. All measures were translated into Chinese by the researcher, and then the Chinese questionnaire was back-translated into English separately by two other people who were fluent in both languages. The English questionnaire and the back-translated English questionnaire were then compared. During this process, unsuitable expressions in Chinese were revised. Participants completed the following measures:
1. Romantic love scale (Rubin, 1970): This measure consists of 13 questions such as “If ___ were feeling badly, my first duty would be to cheer him (her) up” and “I feel that I can confide in ___ about virtually everything” and rates on a 9- point scale from “Not at all true; disagree completely” (1) to “Definitely true; agree completely” (9) (p. 267-268). The score of romantic love is the sum of each item in this scale. High scores in this scale would reflect high levels of love.
179
2. Commitment scale by Rusbult et al. (1998): This scale has been identified as being effective in measuring personal commitment (Johnson et al., 1999; Adams & Jones, 1997; Stanley & Markman, 1992). The scale includes seven items such as “I want our relationship to last for a very long time” and “I would not feel very upset if our relationship were to end in the near future” and rates on a 9- point scale from “Do Not Agree At All” (0) to “Agree Completely” (8) (Rusbult et al., 1998, p. 390-391). The score of commitment is the sum of each item in this scale. High scores in this scale would reflect high levels of personal commitment.
3. Dyadic Adjustment Scale (DAS: Spanier, 1976): The DAS has 32 questions, such as the degree of agreement/disagreement on “Handling family finances”, “In general, how often do you think that things between you and your partner are going well?”, and “Do you and your mate engage in outside interests together?” (p. 27-28). In the current study, a single score of DAS will be used based on the definition of satisfaction by Johnson (1991; Johnson et al., 1999), which was “attraction to the relationship”. One of Johnson’s (Kapinus & Johnson, 2003) studies used six questions to measure satisfaction within his framework of personal commitment, which were all reflected in different aspects of DAS. Therefore, all items in DAS have been added together to form a single score of dyadic adjustment. High scores in dyadic adjustment would reflect high levels of satisfaction.
4. Couple cultural identity scale: This will be measured by two scales,
acculturation to the partner scale and Triandis’ (1996) horizontal and vertical individualism and collectivism scale. The acculturation to the partner scale (see Appendix 2 and Appendix 3) was created specifically for this study. Items of the acculturation to the partner scale have been added together to form one part of “couple cultural identity”. High scores would reflect high levels of couple cultural identity. Triandis’ (1996) horizontal and vertical individualism and collectivism scale included 29 questions on a 9-point scale ranging from strongly disagree (1) to strongly agree (9). Eight questions on vertical
individualism such as “It annoys me when other people perform better than I do”, seven questions on horizontal individualism such as “Being a unique individual is important to me”, eight questions on vertical collectivism such as “I would do what would please my family, even if I detested that activity”, and six questions
180
on horizontal collectivism such as “The well-being of my co-workers is important to me” (p. 415). The scores for the couple cultural identity from Triandis’ scale have been calculated as: calculating each item’s absolute value of the subtraction between partner’s scores, and then adding this absolute value for each item in vertical individualism, horizontal individualism, vertical
collectivism, and horizontal collectivism together to form another part of “couple cultural identity”. Lower score means highly congruent values on individualism and collectivism between partners, which would reflect a higher couple cultural identity.
5. Demographic information: Participants completed questions assessing socio- demographic characteristics and relationship variables. Questions asked for information including participant’s age, gender, length of current marriage, whether it was the first marriage (if not, how many prior marriage(s)), number of children of the current marriage, current social status, parents’ social status, educational level, religion, ethnicity, and parents’ ethnicities.
Procedure
The researcher obtained informed consent from the participants who then completed either an English or Chinese version of the questionnaire. As there were some private questions in the questionnaire, participants were assured of anonymity and confidentiality. The respondents were free to choose English or Chinese version
questionnaire to answer. The respondents were told that there was no right or wrong answer to each question and they were encouraged to choose the answers that best reflected their own personality and situation. All the participants were asked to answer the questionnaire separately from and not to compare answers with their partner. Some participants emailed their completed questionnaires back, some filled in the printed questionnaire, and one couple answered the questionnaires through telephone. Questionnaires from both partners were necessary for the study (completed
questionnaires from only one partner were not considered). After the data collection was completed, participants were thanked and debriefed about the study.
181
Results
Reliability analysis was carried out to test the consistency of the questions on each scale for the current sample. Table 5.1 shows the reliabilities of Rubin’s measurement of romantic love, Rusbult’s commitment scale, Spanier’s DAS, acculturation to the partner scale, and four categories of Triandis’s horizontal and vertical individualism and collectivism scale for women and men respectively. In general, the reliability alphas were acceptable. There was one exception: the reliability of the vertical individualism scale for women was very low (alpha=.41). However when one item, “Competition is the law of nature” (Triandis, 1996, p. 415) was deleted, the reliability rose to .74 according to item-total statistics. For men, the reliability for the scale dropped to .63 after the same item was deleted. Therefore the item was deleted from the scale in the following analysis.
Table 5.1 Reliabilities of scales
Horizontal and vertical individualism and collectivism tendencies (see Table 5.2) were examined through computing variables and descriptive statistics. Men and
women’s horizontal and vertical individualism and collectivism tendencies scores were taken from the maximum average score among the four categories. These results are partly consistent with Triandis (1995) who noted that individualists tend to be horizontal individualists while collectivists tend to be vertical collectivists. This was
Women (alpha) Men (alpha) Rubin’s measurement of romantic love .85 .78
Rusbult’s commitment scale .88 .73
Spanier’s DAS .91 .91
acculturation to the partner scale .63 .60
vertical individualism .41 .70
vertical individualism (after dropping
one item) .74 .63
horizontal individualism .85 .90
vertical collectivism .76 .66
182
true for men in the current sample, but not for women. The most frequent value tendency for men was horizontal individualism, but for women was horizontal
collectivism. There were only a few vertical individualists in the sample. Both men and women had higher percentages of horizontal characteristics than vertical characteristics. In the current sample, women were more collectivistic and men were more
individualistic. This result might reflect the fact that most women participants were Chinese immigrants and most men participants were Westerners.
horizontal individualism (%) vertical individualism (%) vertical collectivism (%) horizontal collectivism (%) Men 67.6 5.4 10.8 16.2 Women 37.8 2.7 16.2 43.2
Table 5.2 Men’s and women’s horizontal and vertical individualism and collectivism tendencies in percentage.
Correlation analysis was carried out for all the scales; Table 5.3 describes the correlations between all the variables used in this study. Although there were some significant correlations, the coefficients were not particularly high, indicating that the scales were not measuring the same constructs. There were some significant
correlations between personal commitment and love, dyadic adjustment, acculturation to the partner, and couple’s individualism/collectivism for men and women respectively. However, couple’s individualism/collectivism did not correlate significantly to men’s personal commitment. Note that couple’s individualism/collectivism significantly correlated with both men and women’s acculturation to the partner, and women’s dyadic adjustment, which means the greater similarity of couple’s
individualism/collectivism tendency, the greater acculturation to the partner for both men and women, and the greater dyadic adjustment for women. However, couple’s individualism/collectivism was significantly correlated to women’s, but not to men’s, personal commitment, love, and dyadic adjustment.
ANOVA of men’s and women’s horizontal and vertical individualism and collectivism tendencies on men and women’s personal commitment did not show significant differences. However men’s horizontal individualism yielded the highest men’s personal commitment, followed by men’s vertical collectivism and horizontal
183
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
1. women’s personal commitment
2. men’s personal commitment .45**
3. women’s love .60** .20
4. men’s love .22 .61** .27
5. women’s acculturation to the partner .59** .09 .23 -.02
6. men’s acculturation to the partner .43** .29* .49** .32* .52**
7. women’s dyadic adjustment .61** .44** .59** .39** .52** .56**
8. men’s dyadic adjustment .31* .56** .39* .52** .20 .53** .71**
9. couple’s individualism/collectivism -.33* .02 -.29* .09 -.50** -.39** -.30* -.19
10. women’s horizontal individualism .03 .01 .05 .25 .05 -.00 -.05 .08 -.21
11. men’s horizontal individualism .04 .14 .09 .31 -.10 .20 .24 .23 .09 .25
12. women’s vertical collectivism .41* .36* .25 .27 .30 .27 .39* .33* -.21 -.27 -.15
13. men’s vertical collectivism -.12 -.05 -.01 .02 .-09 .07 .09 .23 -.07 -.30 -.15 .06
14. women’s horizontal collectivism .22 .07 -.04 .14 .17 -.09 .02 -.12 .00 .08 -.24 .31 .02
15. men’s horizontal collectivism -.03 -.17 .06 -.07 -.12 .04 -.01 -.08 -.32 -.24 -.03 .15 .52** .15
16. women’s vertical individualism -.31 -.18 -.16 .20 -.34* -.21 -.19 .03 .23 .22 .13 -.28 -.12 -.09 -.32
17. men’s vertical individualism -.26 -.07 -.08 .05 -.23 -.13 -.08 -.05 .10 .17 .20 -.27 -.39* -.16 -.34* .33*
**. Correlation is significant at the 0.01 level. *. Correlation is significant at the 0.05 level.
184
collectivism. Women’s horizontal collectivism yielded the highest women’s personal commitment, followed by women’s vertical collectivism and horizontal individualism. Both men’s and women’s vertical individualism yielded the lowest personal
commitment.
Within-sample T-tests were carried out to examine the difference between the couples on personal commitment, love, dyadic adjustment, acculturation to the partner, vertical individualism, horizontal individualism, vertical collectivism, and horizontal collectivism. Table 5.4 describes the means and standard deviations (SDs) and the t scores/significance level. Women’s love scores were significantly lower than men’s, women’s horizontal individualism scores were significantly lower than men’s, and women’s horizontal collectivism scores were significantly higher than men’s. The other t-tests did not show significant differences between women and men.
**. Correlation is significant at the 0.01 level. *. Correlation is significant at the 0.05 level.
Table 5.4 Means, SDs, and the t scores/significance level between men and women.
ANOVA and t-tests were carried out to find out which demographic variables had significant effects on personal commitment. It was found that the number of previous marriages for both women and men significantly differentiated men’s personal
commitment. According to Table 5.5, men had the highest personal commitment scores if they and their partner were in a first marriage, followed by men and women in a second marriage, and men in their third marriage. Cohabiting couples had the lowest
Women (Mean) Women (SD) Men (Mean) Men (SD) T score Personal commitment 52.10 6.04 52.73 4.83 -0.65 Love 86.30 14.53 94.73 10.50 -3.31** Dyadic Adjustment 114.11 18.16 112.31 15.68 -0.84
Acculturation to the partner 42.00 6.51 41.49 6.58 0.49
Vertical individualism 28.46 9.20 29.62 8.73 -0.68
Horizontal individualism 42.24 11.57 48.35 11.31 -2.65*
Vertical collectivism 44.24 9.61 42.76 9.06 0.71
185
personal commitment scores. Having child/children had significant impact on both men’s and women’s personal commitment. According to Table 5.6, couples who had child(ren) had significantly higher personal commitment scores than couples who did not have a child.
Cohabit First marriage Second marriage Third marriage F-ratio
Women 47.33 53.85 53.40 5.63**
Men 47.33 54.29 53.33 51.00 4.55**
**. Correlation is significant at the 0.01 level.
Table 5.5 Mean scores, F-ratios and the significance of men’s personal commitment on the number of marriage of both women and men.
No child Having child(ren) T score
Women 49.65 55.00 3.19**
Men 50.85 54.94 3.01**
**. Correlation is significant at the 0.01 level.
Table 5.6 Mean scores, t scores and the significance of men and women’s personal commitment on having/not having children.
Less than college More than college T score
Women 56.00 51.64 -4.02***
Men 50.00 53.60 1.57 ns
***. Correlation is significant at the 0.001 level.
Table 5.7 Mean scores, t scores and the significance of men’s personal commitment on men’s educational level and women’s personal commitment on women’s educational
level.
Women’s educational level had significant effect on women’s personal commitment, but men’s educational level did not have significant effect on men’s