SUPPORTING STATEMENT
B. COLLECTION OF INFORMATION EMPLOYING STATISTICAL METHODS
1. Universe and Respondent Selection
We will conduct the November 2013 Email Address Collection Test Supplement in conjunction with the CPS, for which the universe is 118 million households. From this universe, we select a sample of approximately 72,000 households each month, of which approximately 60,000 households are eligible for interview.
We actually interview about 55,000 households each month. To minimize burden, we are only conducting this sample in half the rotation groups – approximately 27,500 households. Based on previous response rates for similar size supplements, we expect the supplement response rate to be approximately 90 percent.
2. Procedures for Collecting Information
The November 2013 Email Address Collection Test Supplement is a supplemental survey associated with the November CPS. Attachment E gives an overview of the CPS sample design and weighting methodology. The statistical properties of these supplemental items will fall within those associated with the CPS.
For testing the best way to collect email addresses, we have two question methodologies being tested. One is a one-step approach while the other is a two-step approach. See Attachment A for wording. The input includes a random number to make sure either Question A or Question B is randomly selected among the respondents. This creates two subsamples that will be able to be weighted using the standard weighting and will allow for not only comparison at the whole universe level but at lower demographic levels with each subsample being representative of our respondent population. We will be comparing the two to see which method provides the most viable email addresses as well as lower refusal rates and false positives. This method should allow for us to measure a detectable difference of .4% at a 80% power analysis. For the smaller demographic populations (Hispanic Origin, Race, Sex), the detectable difference should be .4% for the larger demographic groups such as White Only and the largest should be around a 2.4% detectable difference for smaller race groups such as Other Race.
3. Methods to Maximize Response
We maintain response rates and data accuracy for the CPS at high levels through computer edits, interviewer instruction and training, and close monitoring of these data. (Refer to paragraph 5 of Attachment E for a discussion of the CPS nonresponse.)
4. Testing of Procedures
We do not require testing of procedures and methods given that we used the same procedures and methods as basic CPS. We did perform cognitive testing of the questions to make sure the questions were properly worded and not sensitive. Attachment B is a copy of that report.
5. Statistical Aspects and Data Collection
The following individuals may be contacted on the statistical data collection and analysis operations:
Statistical Design/Statistical Analysis
Yang Cheng
Demographic Statistical Methods Division
U.S. Census Bureau
(301) 763-3287
Data Collection/Survey Design
Lisa Clement
Office of Associate Directorate of Demographic Programs
U.S. Census Bureau
(301) 763-5482
Attachments
A. Supplemental Questionnaire
B. Report of Cognitive Testing of the 2013 Email Address Collection Test Supplement
C. Confidentiality Brochure
D. CPS Advance Letter
E. Overview of CPS Sample Design and Methodology
| File Type | application/msword |
| Author | woods009 |
| Last Modified By | kml |
| File Modified | 2013-11-01 |
| File Created | 2013-07-22 |