|
Activity Number:
|
466
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 7, 2003 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Survey Research Methods
|
| Abstract - #301059 |
|
Title:
|
Using Network Sampling in a Phone Survey to Locate Nontelephone Households
|
|
Author(s):
|
Nadra Garas*+ and Johnny Blair
|
|
Companies:
|
American University and Abt Associates, Inc.
|
|
Address:
|
2929 Connecticut Ave. NW, Washington, DC, 20008,
|
|
Keywords:
|
network sampling ; frame coverage ; multiplicity estimation ; telephone survey ; survey methodology ; nontelephone households
|
|
Abstract:
|
A limitation of telephone surveys is the exclusion of nontelephone households and resulting potential bias. Statistical adjustments can address this problem, but the only way to include nontelephone households in the survey is to locate them using a different frame and expensive screening procedures. In this research, we explored locating nontelephone households with network sampling. The counting rule asked telephone survey respondents whether anyone used their phone because they did not have phone service. The probabilities of inclusion using this approach are unequal. But if the contacted non-telephone households can report how many different households they used for telephone service in the reference period, appropriate weights can be constructed. In two national and one statewide survey (1997-2001), about 10% of telephone households reported such use of their phones and were disproportionately from lower income households. Attempts to contact the identified households were less successful. Given costs of including coverage of nonphone households, this methodology seems worth further examination. We report the results of these studies and plans for further research.
|
