Krosnick presentation on data quality and survey mode
Nov 06 2009
I had the good fortune last Friday to be able attend a presentation by Jon Krosnick on the work he and some of his grad students and faculty colleagues have been pursuing during the past decade on the effect of survey mode on data quality. In the late 1990s when Jon and I both were professors at Ohio State, Jon used the OSU survey center, which I directed, to gather RDD datasets that he would later compare to internet samples based on probability sampling. Having invested a fair amount of my own time supporting the rigor of some of Jon's RDD surveys, I was especially interested in what he has learned.
In the past few months, AAPORnet has contained some exchanges between Jon, Doug Rivers, and others regarding whether internet surveys that are not based on probability sampling can be as accurate as surveys based on probability samples. As you also might know, AAPOR currently has a Task Force chaired by Reg Baker on internet panels and survey error.
Today I was made aware that a video of Jon\\'s presentation from last Friday is available on a website put up by Knowledge Networks (http://www.knowledgenetworks.com/insight), which sponsored the presentation . The website also contains other materials related to the presentation that can be downloaded.
The video runs about an hour, but I strongly recommend it to all interested in survey mode effects on data quality, and encourage employers to recommend it to their statistical and methodological research staff.
This article has been read 3593 times