Recap: Joint Statistical Meetings (JSM)

August 29, 2016 Austin Lasseter, Ph.D.

On July 31, 2016, I presented a research paper at the Survey Research Methods Section of the JSM Conference in Chicago. Mine was one of six contributed papers on the topic of Raking, Post-Stratification, and Calibration Methods in survey research.

Austin_speaking_low_res.jpgTogether with my co-author Jonathan Phelan of the American Institutes for Research (AIR), I discussed:

  • Non-coverage bias, a common problem in survey data which occurs when the sampling frame does not match the target population
  • Methods for measuring non-coverage bias, and the differing opinions about the degree of bias that should be considered excessive

Our paper demonstrated how raking adjustments to sampling weights could substantially reduce bias due to non-coverage when extrapolating the High School Longitudinal Study (HSLS) follow-up results to all high school seniors in the nation in 2013.

The study used data from three datasets of the National Center of Education Statistics (NCES): the HSLS, the National Assessment of Educational Progress (NAEP), and the Common Core of Data (CCD). Our research paper was written during my time at AIR as part of a contract with the National Center for Education Statistics (NCES).

During the same session there were several other authors presenting papers on topics related to ours. A few highlights from our session included the following:

Thanks to Summit for supporting my participation at this conference! I look forward to continuing the dialogue.


In early August 2016, I presented at the National Academy of Education (NAEd) “Big Data: Balancing Research Needs and Student Privacy” workshop. You can read my blog post about NAEd here.

 

Share This: