|
Activity Number:
|
479
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2007 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Survey Research Methods
|
| Abstract - #308418 |
|
Title:
|
Nonresponse in the Survey of Industrial Research and Development
|
|
Author(s):
|
Perez Adriana*+
|
|
Companies:
|
The University of Texas at Houston Health Science Center
|
|
Address:
|
80 Fort Brown SPH RAHC N102B, Brownsville, TX, 78520,
|
|
Keywords:
|
complex-survey ; establishment survey ; outcome ; assessment ; survey bias and precision
|
|
Abstract:
|
The Survey of Industrial Research and Development (SIRD) is important because provides statistics on research and development (R&D) by companies in the United States, which is used to evaluate the status of science and technology diffusion in the United States relative to other nations. Missing data is present in many of the SIRD key variables, varying from 1.1% to 68.2% (year 2003). Several imputation techniques (auxiliary ratio and auxiliary trend) are implemented to overcome missingness. Currently, estimates are reported for SIRD as if imputation values were observed values. A maximum likelihood estimate (MLE) of the data accounting for the uncertainty due to missing values is proposed. Empirical and simulations studies were developed to compare the results of current and proposed estimators through precision and accuracy measures (bias and mean square error).
|
