JPSM Seminar

'Why Machines Matter for Social Science and Survey Research'

With increases in technology and big data streams comes increased opportunities for
researchers to access information about members of diverse populations. And as the
number of sources of information continues to increase, our data sets become wider
and richer. This new information offers social scientists and survey researchers and
practitioners, alike, the potential for improved designs and better predictive models.
In an era of increasing costs, decreasing participation and the proliferation of
information available about individuals from a multitude of sources, the race is on for
developing and applying techniques that can leverage the various potential in the
masses of big data for improved survey designs, more efficient fielding and better
analytic models for imputation and estimation. This presentation discusses some of
the new ways in which both supervised and unsupervised machine learning methods
are applied to all aspects of the survey research process.