Full Citation
Title: Partner choice and parameter estimates: modelling the effect of preferences
Citation Type: Working Paper
Publication Year: 2019
ISBN:
ISSN:
DOI:
NSFID:
PMCID:
PMID:
Abstract: The simulations have a number of implications for the practice of assortative mating research. First (and this is not a particularly new observation but has perhaps new force), different mechanisms can generate quite similar patterns of assortative mating, so we need to use information additional to that in square tables of marriages to differentiate between the mechanisms. Second, while loglinear models are clearly necessary and far superior to simpler summaries of the structure of tables, we need to be cautious about making inferences from parameters to preferences. Processes happen through time, while loglinear models are one-shot. If at all possible, we should use data with a time dimension, and incorporate time in the modelling. In the absence of properly longitudinal data, this can be read as a renewed injunction to focus on recent marriages where at all possible.
Url: http://www.ulsites.ul.ie/sociology/sites/default/files/wp2019-01.pdf
User Submitted?: No
Authors: Halpin, Brendan
Series Title: University of Limerick Department of Sociology Working Paper Series
Publication Number: WP2019-01
Institution: University of Limerick Department of Sociology
Pages: 1-33
Publisher Location:
Data Collections: IPUMS USA
Topics: Family and Marriage, Other
Countries: