|
Activity Number:
|
437
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Wednesday, August 6, 2008 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Survey Research Methods
|
| Abstract - #300756 |
|
Title:
|
Assessing Mode Bias in a Mixed Mode Reproductive Health Survey
|
|
Author(s):
|
Holly B. Shulman*+ and Nedra Whitehead
|
|
Companies:
|
Centers for Disease Control and Prevention and Research Triangle Institute
|
|
Address:
|
205 College Ave, Swarthmore, PA, 19081,
|
|
Keywords:
|
mode effects ; mode bias ; surveillance ; mixed mode
|
|
Abstract:
|
PRAMS is a mixed mode (mail/telephone), state population-based surveillance system of women who recently delivered a live-born infant. Because of the sensitive nature of some of the topics addressed, we examined the extent to which the mode of completion biased survey estimates. We selected a nested matched case-control sample (n=8617) from all respondents, with telephone respondents treated as cases and mail respondents as controls. Matching variables were related to the likelihood of response and to mode of response. We examined 20 sensitive questions. Several socially undesirable events were reported more frequently by mail respondents. The magnitude of the bias ranged from 0.03 to 1.04 percentage points for the 6 events with the strongest mode effects. Our results indicate that mode effects do produce bias in some prevalence estimates from PRAMS, but the magnitude is small.
|
