470 likes | 597 Views
SIPP Basics. National panel survey ? Since 1984 with sample size between about 11,000 to 45,000 interviewed householdsThe duration of each panel varies from 2? yrs to 4 yrsThe SIPP sample is a multistage-stratified sample of the U.S. civilian noninstitutionalized populationThe survey uses a
E N D
1. A Comparison of Survey Reports Obtained via Standard Questionnaire and Event History Calendar* Priming Evaluation *
Jeff Moore, Jason Fields, Gary Benedetto,
Martha Stinson, Anna Chan, and Jerry Maples
U.S. Census Bureau
XXV International Methodology Symposium: Longitudinal Surveys: from Design to Analysis
Symposium international sur les questions de méthodologie
Statistics Canada
October 27-30 2009
Gatineau, Qc, Canada
2. SIPP Basics National panel survey – Since 1984 with sample size between about 11,000 to 45,000 interviewed households
The duration of each panel varies from 2˝ yrs to 4 yrs
The SIPP sample is a multistage-stratified sample of
the U.S. civilian noninstitutionalized population
The survey uses a 4-month recall period – 3 interviews / year
The sample is divided into 4 rotation groups for monthly interviewing
Interviews are conducted by personal visit and by decentralized telephone
3. The Save the SIPP Campaign
New York Times Editorials
Discovering What Happens Next
Save our SIPP
The Continuing Saga of SIPP
NY Representative, Carolyn Maloney stated:
“I want to thank all of the policy groups and economists who worked so hard to help reverse the Administration’s original SIPP decision. Because of their dedication and hard work, the Administration came to understand how important SIPP is to creating and implementing good public policy.”
4. SIPP Re-Engineering Implement Improvements to SIPP
- reduce costs
- reduce R burden
- improve processing system
- modernize instrument
- expand/enhance use of admin records
Key Design Change:
- annual interview, 12-month reference pd., event history calendar methods
5. EHC Interviewing Human Memory
- structured/organized
- links and associations
EHC Exploits Memory Structure
- links between the occurrence and timing of events
EHC Encourages Active Assistance to Rs
- flexible approach to help elicit an
autobiographical “story”
6. Evaluations of EHC Methods Many EHC vs. “Q-List” Comparisons
- various methods
- in general: positive data quality results
BUT, Important Research Gaps
- data quality for need-based programs?
- comparison to 4-month reference period?
7. Field Test Goals & Design Basic Goal:
Can an EHC interview collect data of comparable
quality to standard SIPP?
- month-level data
- one 12-month ref pd interview vs. three
4-month ref pd interviews
- especially for need-based programs
Basic Design:
EHC re-interview of SIPP sample HHs
9. Stakeholder Concern About Test Design
10. Design Details (1) Main Sample:
SIPP 2004 Panel Cases from Waves 10-11-12
- reported on CY-2007 via SIPP “Primed”
Supplemental Sample:
SIPP Wave 8 Sample Cut Cases
- dropped from SIPP in 2006; “Unprimed”
EHC Re-Interview in 2008, about CY-2007
11. Design Details (2) EHC Questionnaire
- paper-and-pencil
- 12-month, CY-2007 reference period
- selected SIPP topics (“domains”)
- start with landmark events
- within domains, anchor on “now”
- month-level (at least) detail
Field Period: Mid-April thru Late June 2008
Sample of Addresses, Not People
- post-interview clerical match to SIPP
12. Control Card – 2008 Field Test Instrument
13. Page 1 of 5 – 2008 Field Test Instrument
14. Design Details (3) $40 Incentive, Non-Contingent and Non-FR-Discretionary
Same Response Rules as SIPP
- EHC interview for all adults (15+)
- self-response preferred (proxy permitted)
Field Staff: Census Bureau FRs
- most with some interview experience
- ~1/3 with SIPP experience
- 3-day training on EHC methods
15. Evaluation Plans Compare SIPP and EHC Survey Reports
- Same People
- Same Time Period
- Same Characteristics
Evaluation of “Priming” Effect
Additional Evaluations:
- Administrative Records Evaluation
- Training Evaluation
- Field Process Evaluation
16. Design Details (4) TOTAL Initial Sample Addresses
Illinois Texas Total
2007 SIPP Respondents: 487 609 1,096
SIPP Sample-Cut Cases: 427 422 849
1,945
Outcomes:
Household interviews Total: 1,627 (91%)
2007 SIPP Respondents: 935 (91%)
SIPP Sample-Cut Cases: 692 (91%)
Individual Adult (15+) EHC interviews Total: 3,318 (99%)
2007 SIPP Respondents: 1,922 (99%)
SIPP Sample-Cut Cases: 1,396 (99%)
EHC Adults matched to SIPP Total: 2,756 (83%)
2007 SIPP Respondents: 1,658 (86%)
SIPP Sample-Cut Cases: 1,098 (78%)
FINAL ANALYSIS SAMPLE:
2007 SIPP Respondents: Total 1,620
Non-movers 1,420
Movers 200
SIPP Sample-Cut Cases: Total 1,090
17. What were the overall research questions?
18. What were the results? Almost All SIPP and EHC Reports Agree
- all characteristics, all months
- in general: 97-98% likelihood that a respondent’s
SIPP and EHC reports will agree
- worst case (employment): 92-94%
Disagreements are RARE EVENTS
Results available:
http://www.census.gov/sipp/DEWS/moore-aapor2009slides-abbrev.ppt
19. Results Summary for Reinterview Evaluation 3 Patterns:
1. EHC = SIPP All Year
SSI; WIC (IL)
2. EHC < SIPP All Year
Medicare; Social Security;
WIC (TX); Food Stamps (IL)
3. EHC < SIPP, Early in the Year Only
Food Stamps (TX); employment;
school enrollment
20. The Movers The ‘Main Sample (Wave 11)’ of EHC respondents who reported for CY-2007 has an important difference from our ‘Supplemental Sample (Sample Cut).’
21. The Movers The two subsamples of EHC respondents show different characteristics, and one primary cause are the Wave 11 respondents who moved between June 2006 and mid-2007.
22. The Movers
23. Assessment of Priming (1) SIPP Wave 10-11-12 Respondents
Provide CY-2007 Data Twice
- first SIPP, then EHC
- results presented previously
Are Their EHC Reports Biased by having responded to SIPP for CY-2007 already?
- e.g., more accurate EHC response
- could bias field test interpretation
Control Group: Wave-8 Sample Cut
- last SIPP response in Jun-Sep 2006
- “Unprimed” re: CY-2007
(however still SIPP experienced)
24. “Priming” - SSI(Supplemental Security Income)
28. “Priming” - WIC(Women, Infants, and Children)
31. “Priming” - Work
34. “Priming” - Food Stamps
37. “Priming” - School Enrollment
40. Priming - Findings No evidence of priming with respect to having reported about CY-2007 in SIPP.
- For each characteristic
- Non-Movers from W11 and
Sample-Cut cases are not different
- Interaction b/w these groups and
months are not different
- Results hold for both weighted and
un-weighted models
41. Results Implications Levels of agreement from the reinterview comparison between SIPP and the EHC
- are not artificially inflated due to experience as reinterview respondents re: CY-2007
- may still be biased because both the main reinterview sample and sample-cut respondents were long-time SIPP respondents.
42. Field Test Overall Summary Successful “Proof of Concept”
Overwhelming Finding: SIPP-EHC Agreement
Valuable Lessons to Inform Next Test
- larger, broader sample
- “correct” timing of field period
- automated questionnaire
Specific Data Comparisons are Instructive
43. Comments: Jason Fields – Jason.M.Fields@Census.Gov
Jeff Moore – Jeffrey.C.Moore@Census.Gov
45. Page 2 – 2008 Field Test Instrument
46. Page 3 – 2008 Field Test Instrument
47. Page 4 – 2008 Field Test Instrument
48. Page 5 – 2008 Field Test Instrument