Friday, May 05, 2006

Traditional Mother Power

Hypothesis

In a graduating class of about 630 students, I could probably count the number of couples with both hands and feet with digits to spare. In a school with about a 70% Asian population (most first or just barely second generation), it would not be an exaggeration to say that many of them had conservative parents who stuck to fairly traditional gender roles. Despite the number of my classmates who had two working parents, it was evident that the mother's role had been much less of the "bring home the bacon" type. What are the affects of a traditional mom and the dating outcomes of their children? I am looking at women's responses to views on traditional gender roles and the subsequent influence that may have on the dating/sexual habits of their children.

The cliche of the "nature vs. nurture" argument stands this test as so many environmental influences in dating trends seem to push for more openness in the dating market. Despite the spirit of a new age of high-speed, hook-up culture, perhaps parents still have a stronge presence on the weighings-in of dating behavior of their children. I would like to believe that traditional parents can still influence the initial age their kids begin to explore dating and all that it entails.

Data/Empirics

Using the NLSY1979 and NLSY1979Children/Young Adults data, I look at the mother's family views on traditional roles of wives. I will end up only following along with matched sets of mothers and children from these surveys with mother respondents from 1979 and child respondents from 2002. The main regression will be traditionality on dating behavior. Responses to the family/wife attitude questions will be controlled by religious activity, presence of the father, closeness to the mother, gender, and race of the children when considering the regression. These factors only cover some of the variables that would ultimately affect the dating/relations outcomes of the children. There will probably be three similar regressions: dating, sexual activity, and cohabitation behavior from these children.

Concerns

Because many of the preferences toward dating have a lot to do with the endogeneity of preferences, peer effects, and other factors such as the attractiveness of the child/young adult respondent, I face a lot of internal validity issues. There needs to be serious considerations on how to control for environmental influences such as high school dating atmosphere. In the example of my high school, it is more than likely that the scarce numbers came from a peer effect that ended up generating a date-less culture. The attractiveness of the respondent is also a highly variable factor that could influence the behavioral results. If I can successfully find variables that somehow capture this effect, I can probably have a better idea of the true relationship as these omitted variables gets added to the regression. Another concern comes from the direct prohibition by parents to date before a certain age. This may skew data toward older ages if such restructions are followed through by the child.

Comments:
About the attractiveness variable, one of the problems we dealt with in 1123 this semester made me think a lot about this problem. It seems that the only way to measure attractiveness (nearly)objectively is to simply take an average of opinions. This is of course difficult, but even using 2 or 3 people's opinions would help eliminate subjectivity
 
Huilin,

You need to develop a more economic version of this hypothesis. That is, the decision that you seem interested in is the decision to date in adolescence. You seem interested in explaining variation in this choice using mother's attitudes, but you haven't connected the adolescents dating decision to the mother's attitudes. Why does an adolescent care what his/her mother's attitudes are when deciding to date.

Further, how do you interpret the coefficient in your regression? Assuming that you find results which suggest that certain attitudes in the mother affect the kid's dating decision, what does that mean? Specificly, does it mean that having a mother with certain attitudes CAUSES me to change my dating behavior? If so, how does this occur? Can you do anything to support your explanation?
 
Coming from a semi-traditional Indian family I can definitly understand where this project is coming from. However, I think that therein lies one problem. A lot of the affect on a mother's attitudes come from their culture. Asian parents might be more likely to have certain attitudes and also date less. However, the fact that they date less might now be due to these attitudes but instead due to being Asian. Therefore, we have a lurking race variable that is influencing both the mother's attitude variable and the dating variable. Therefore, I think it is imperative that you control for race and ethnicity in the regression. This will also make the regression more interesting. It will now be interpreted as the difference within culture of having a mom with certain values vs. another mom with other values. Essentially you will be comparing the "cool Americanized" asian moms to the "traditional" asian mom.
 
As Bryce notes, I think it is important to distinguish correlation and causation with tregard to mothers' attitudes and their children's dating decisions.
On a different note, depending on the data you are using, perhaps it would be easier and more concrete to try to compare the mother's religious piety to the dating decisions of their children. Through friends and acquaintances, I have learned that religious piety of the parents can affect the religious piety of their children, and, in turn, the children's dating decisions. Accordingly, perhaps religiosity might be worth exploring.
 
I agree with Mark about looking into the religious piety and if you don't use this as a variable that you want to test, perhaps consider controlling for it. Religion has a big effect, particularly on sexuality, so you might not want to leave it out of the study completely.
 
since posting, I've actually changed my project a lot. I am now looking more into the eagerness to marry as opposed to dating age. There is some very workable data on NLSY79 that I have been fiddling around with and it takes away a lot of the omitted variable bias with attractiveness. Race is still an issue to consider, but there is a lot of difficulty in finding significant results in isolating individual races, especially when Asian affects so little of the data.
 
Post a Comment

Subscribe to Post Comments [Atom]





<< Home

This page is powered by Blogger. Isn't yours?

Subscribe to Posts [Atom]