Sample Weights by Race
Figure 1 below shows the relationship between sample weights and race. Because the survey oversampled black women, we can expect that the sample weights for black women are smaller than that of white women, meaning that each black woman in the sample represents on average fewer black women in the population compared to white women.
Sample Weights by Weight
Figure 2 below shows the relationship between sample weights and physical weight. Although the sample design did not use physical weight as a factor in sampling the women, the graph shows us that women under 150 pounds have smaller ranges of sampling weights compared to women over 150 pounds. The ranges of sampling weights for women over 150 pounds show us that overall, women over 150 pounds tend to more frequently have lower sampling weights, meaning that they tend to more frequently represent fewer women in the population. Note that while weight is not one of the variables analyzed in our Results section, the distribution of weight provides us with the information that the women in the sample are distributed as expected. Therefore, we might expect that the health characteristics we do analyze are representative of the population.
Sample Weights by Income
Figure 3 below shows the relationship between sample weights and household income levels. The sample weights for income levels of $50,000 and greater contain quite a few outliers, especially concentrated in the lower sampling weight values. This shows that a greater number of households earning $50,000 or greater tend represent fewer households in the population, which is expected because the proportion of high earners in the population decrease with increasing levels of income. Note that while household income is not one of the variables analyzed in the Results section, the income distribution provides us with the information that the households in the sample are distributed similarly to the population. Household income heavily correlates with homeownership, which we do analyze by categories of women’s attitudes towards women working.
The analysis looks at the relationship between attitudes of women towards women working and homeownership, life satisfaction, and healthrelated characteristics. All estimates took into account the sample weights provided by the data set.
Table 4 shows the estimated proportions of “types” of women in the sample. The categories split into “careerminded” women, housewife type women, and women who value their husbands’ opinions. The classification of women in the data set comes from the responses to two questions about how the women felt about women working if their husbands agree/disagree.16 We classify the women who answered “Definitely Alright” or “Probably Alright” to both questions as “careerminded” women because they felt that women should be able to work if they wanted to despite their husbands’ opinions and family obligations. We classify the women who answered “Definitely Not Alright” and “Probably Not Alright” to both questions as “housewife type” women because they felt that women should not work if they wanted to due to family obligations and their perceived traditional women’s roles in society. We classify the women who answered “Definitely Alright” or “Probably Alright” in the case when the husband agrees and “Definitely Not Alright” or “Probably Not Alright” in the case when the husband disagrees as “values husband’s opinion” because these women changed their attitudes about women working depending on their husbands’ opinions, showing that these women believe a woman should work if she wants to only if her husband agrees that she should work.
Table 5 shows the estimated proportions of the three “types” of women by racial group. The “other” racial group is left out because of too few observations. Note that the estimated proportions of the three “types” of women for the white population is reflective of the distribution of the estimates from the full sample. However, the estimated proportions for the black population differs quite a bit from the population estimate. Approximately 6.25% of black women are “careerminded”, which is greater than the estimated 4.3% percent of “careerminded” women in the full sample and the 97.5 percentile estimate of 5.26%. Only approximately 18% of black women are “housewife type” women, which is only half of the estimate of the population proportion of “housewife type” women, suggesting that black women are less confined to the idea that a woman’s place is in the home. However, roughly 76% of black women believe that a woman should only work if her husband agrees to it, which is significantly greater than the population estimate for the full sample of roughly 60%.
These results show that black women’s attitudes towards working are quite different from white women’s attitudes towards working. This difference in attitudes may be a reflection of cultural values and norms or the higher poverty rates amongst the black community and the higher need for work. This may also be a reflection of black history, where the majority of the black population often worked in captivity or for very low wages, leading to lower rates of the idea that a woman’s role is in the home and not in the workforce.
Mosaic Plot for Levels of Happiness
Significance Test for Homeownership
Tables 6 and 7 show the expected and observed number of women who are homeowners. We analyze whether or not there exists a significant difference in homeownership based on women “type”. We perform a chisquared test to determine if a difference exists and the test gives us a value of 6.99 and a pvalue < 0.05 (0.0303). “Careerminded” women had a 2 percentage point increase in homeownership compared to the expected. “Housewife type” women had a 4.7 percentage point decrease in homeownership compared to the expected. Women who valued their husbands’ opinions had a 2.7 percentage point increase in homeownership compared to the expected. These results are not surprising as these results reflect how likely the respondent is to work or has worked in her lifetime. Because “careerminded” women most likely worked and some of the women in the category of women who values their husbands’ opinions worked, the two groups are expected to have greater rates of homeownership compared to “housewife type” women who likely never worked.
Figure 5 shows that the estimated proportions of women who smoke in the population by “type” are 20% for careerminded women, 22% for housewife type women, and 27% for women who value their husbands’ opinions. Estimated proportions of women who had high blood pressure were similar to estimated proportions of women who smoked.
Our analysis showed that “careerminded” women were more likely to be homeowners, less likely to report being happy, less likely to smoke and have high blood pressure, and more likely to have had a drink in the past year. “Housewife type” women on the other hand, had lower rates of homeownership than expected, had higher levels of reported happiness and fell in the middle in terms of estimated proportions of women who smoke, drink, and have high blood pressure. Women who believe that women should only work if their husbands agree to it had greater rates of homeownership than expected as well, had higher levels of reported happiness, had greater rates of smoking and high blood pressure, and had the lowest estimated proportion of women who drink.
These results reflect the attitudes and choices of women in 1995. Although the data is somewhat dated, our analysis brings importance to the idea that women’s attitudes may be an obstacle in the goal of steering women into the workforce, and even encouraging women to enter predominantly males industries. Differences in homeownership between the three “types” of women is just one example of how women’s attitudes towards women working can be a factor in determining wealth accumulation during one’s life. There definitely exist other obstacles in bringing gender equality in the workforce, but this analysis emphasizes the important role of women’s attitudes towards women working as a factor in determining women’s labor force participation rate and consequently, future financial security and wealth.