Having Daughters Increases Parents’ Identification with the Republican Party

[antigone382]

63. Well, they took that data set because it was all that was available.

The GSS is one of our most reputable survey instruments and the methodology of data collection for that survey is very well understood by everyone in the sociological community. It is carried out by the U.S. Census Bureau. The only year in which questions on the gender and birth order of respondents' children was asked was in 1994, so that was the most recent data they could use. Data collected in the GSS tends to be so rich and to cover so many dimensions of American social life that it doesn't surprise me at all that twenty years later, compelling statistical analysis is still being carried out from one year of the survey. It would certainly be interesting and ideal for a more recent survey to be carried out on this topic, but it is almost certainly beyond the funding capacity of two university researchers to implement a randomly sampled national survey asking these questions. They have to use the best and most current data available to them, acknowledging its limitations. Perhaps they can use this published report to acquire funding for such a study.

Moreover, it's pretty standard when writing a research report to offer a potential reason for a finding, without asserting that that reason is correct, or that the finding is necessarily established. The fault for misinterpreting such speculation rests with the journalists reporting on the study.