Friday, October 06, 2006

Strange Data Indeed

by Ken Houghton

As I mentioned below, the Yale Work-Life Balance Survey is out. While I leave it to Kim to do the heavy lifting, here is a curious comparison:
It’s financially difficult to support a family with one salary. (1=strongly disagree, 7=strongly agree)

Males = 4.50
Females = 5.02*

We can ignore, for the moment, that there is a statistically-significant gender difference in responses to that question. Working strictly with the data at face value, compare the scoring to
If I have children someday, I think it will be hard to financially support my family on one salary (either mine or my spouse’s). (1=strongly disagree, 7=strongly agree)

Males = 3.90
Females = 4.47*

Again, we can ignore that the gender difference is statistically significant and look only at the data, which was indicated on the same scale for both questions.
    dMales = 4.50 - 3.90 = 0.60
    dFemales = 5.02 - 4.47 = 0.55

I have no idea if the delta difference is statistically significant and, again, note that I'm not looking at that aspect of the data.

What is amazing is that, for both men and women, there appears to be a significant expectation that they will be more able to support a family on one salary than is currently possible.

Does this look to anyone else as if it's a serious disconnect, or am I missing something?
As a first reaction, I'd tend to view these as responses of the 'conditions suck in general but are not so bad for me' variety. Strictly speaking, it's hard if not impossible to identify from the survey whether the respondents are systematically more wrong about the former vs. the latter. Though every once in a while I'll hear about something that seems like evidence of individual 'money delusion' that would make me put my money on 'conditions suck in general.'
Note, too, that the sample for the Yale Work-Life survey is Yale students. The first (depersonalized) question reflects Yalies understanding of the world "out there;" the second, their perception of their own prospects. And Yale students rightly rate their chances of getting by on one income much higher than those of the "average" American family; not only will most Yale respondents make it through the educational system with an Ivy League degree in hand, but on average they have far greater familial resources to fall back on (e.g., in starting a business, buying a house, etc) than the average American BA recipient.

(Sorry for the delayed response -- coincidentally, I was giving a talk at Yale.)
Thanks, Kim. An all-Yalie sample would explain things... the interesting question is what you'd get from a more-or-less random sample.
Post a Comment

<< Home

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