|
Activity Number:
|
167
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 3, 2009 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Survey Research Methods
|
| Abstract - #303998 |
|
Title:
|
An Empirical Study of Nonresponse Adjustment Methods for the Survey of Doctorate Recipients
|
|
Author(s):
|
Fan Zhang*+ and Stephen Cohen and Donsig Jang and Amang Sukasih and Sonya Vartivarian
|
|
Companies:
|
National Science Foundation and National Science Foundation and Mathematica Policy Research, Inc. and Mathematica Policy Research, Inc. and Mathematica Policy Research, Inc.
|
|
Address:
|
4201 Wilson Blvd., Suite 965, Arlington, VA, 22230,
|
|
Keywords:
|
Survey of Doctorate Recipients ; nonresponse adjustment ; covariate weighting class ; propensity weighting class ; propensity score
|
|
Abstract:
|
In the Survey of Doctorate Recipients at National Science Foundation, adjustments were made to the design weights to compensate missing units. In search of sensible adjustment methods, we conducted a simulation study to evaluate five different adjustment factors: the inverse of unweighted response rate in weighting cell formed by cross-tabulation of significant main effects; the inverse of weighted response rate in weighting cell formed by deciles of predicted propensity values estimated by logistic regression; the inverse of unweighted response rate in weighting cell formed by deciles of predicted propensity values estimated by logistic regression; the inverse of individual predicted propensity values from an unweighted logistic regression model; and the inverse of individual predicted propensity values from a weighted logistic regression model.
|
