7 responses to “Kids these days really know how to throw off a narrative on gender and families

  1. homunq

    Shouldn’t you be using a hierarchical model, with random effects for the terms with years? That would reduce the size of apparent trends, correcting for sampling error.

    If your code were in R, I’d post a fix.


  2. Another thing I worry about is changes in GSS coverage over time. GSS has had a Spanish language version since 2006, so there’s better coverage of the Spanish-speaking population since then. How does that figure in? And how sensitive are trend analyses to how sampling weights are used?

    And how about adding quaratic and cubic year terms and interactions and showing that marginsplot (and significance tests)?


  3. So, I poked around the GSS site and looked at the questions list. I found the one re: men should work and women should stay home, but I didn’t see any related question either stated in a gender-neutral way or where the genders are reversed. I’m the breadwinner in my family and my husband is a SAHD because we think that everyone is better off if one of the adults in the household is able to be available for the kids full-time.

    But if I were being surveyed, and I was only given the option of the question as stated, I would probably answer “agree” but not because I think women should stay home. I think SOMEONE should stay home if possible, but that doesn’t mean you’re bad or wrong if that doesn’t work for your family (for whatever reason).

    TL/DR: I think they need to update the question, or at least ask a couple different variations of the question in order to capture some useful information.

    P.S. Let me know if I’m missing something.


    • The issue is that it’s good to be able to track questions over time; that’s why the old questions are important. New questions don’t mean as much because you don’t have the context of the trend, so you have to make assumptions or guess about what they mean. The hope is that the meaning of the questions changes more slowly than the responses, so that changes in the responses tells you something valuable.


  4. A postscript on the problems of getting new GSS data.
    For the AJS article Philip notes, my memory is that we had it written and accepted with one “knot” in the time series in the mid 1990s, but they delayed so long in publishing it that we had to add more GSS data. The new data continued the mild post-2000 rebound in gender attitudes, so we had to add a second knot at 2000 and analyze it — which really muddied the story (even thought the post-2000 rebound is very mild and almost entirely cohort replacement).

    The question that still fascinates me is that in both Joanna’s MTF high school seniors’ data, and the GSS national data, the turnaround in the mid-1990s appears very abrupt and enduring. So, what was so dramatic then that could turn around a decades-long trend?


  5. Pingback: Family Inequality year-end review | Family Inequality

Comments welcome (may be moderated)

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s