MultiTopic Household Surveys: Fundamentals

1
Multi-Topic Household Surveys Fundamentals

• Diane Steele
• DECRG
• Session II Designing and Implementing
  Multi-Topic Household Surveys
• January 23, 2008

2
What makes a good survey?

• Relevance
• Answer the policy question
• Quality
• Methodology
• Accuracy and reliability
• Adherence to international statistical guidelines
  
• Timeliness
• Punctuality
• Production time
• Comparability over time
• Sustainability
• Public dissemination
• Documentation

3
What makes a good survey?

• Questionnaires
• Sample design
• Field work
• Data management

4
Brief History of the LSMS

• 2006 onward
• Methodological Research
• Concepts Consumption, Financial Services
• Technology CAPI, GPS

5
Whats in a name the typical LSMS Survey

• LSMS survey is not standard survey, but each
  survey shares important common characteristics
• There are other multi-topic household surveys
• Integrated Survey
• Priority Survey
• Core Welfare Indicator Questionnaire (CWIQ)
• Household Income and Expenditure Survey/
  Household Budget Survey

6
Whats in a name the typical LSMS Survey

• An integrated survey Multi-topic and multi-level
• 1) Household questionnaire
• collected at the household level HH roster,
  consumption, housing, agriculture, household
  enterprises, misc. income, savings, credit
• information collected directly from individuals
  rather than proxy respondents Education, health,
  employment, migration, anthropometry, fertility

7
Whats in a name the typical LSMS Survey

• An integrated survey Multi-topic and multi-level
  (cont)
• 2) Community questionnaire
• 3) Price questionnaire (food/non-food at
  community markets)
• 4) Facility questionnaires (schools, health
  facilities)

8
Relationship Between LSMS Purposes and Survey
Instruments
9
Whats in a name the typical LSMS Survey

• Sample Nationally representative, but relatively
  small samples.
• Survey instruments Quantitative questionnaires
  (few open-ended questions)
• Data Entry concurrent DE (done quickly following
  interview in many cases where feasibly in the
  local field office or in the field itself)
• Documentation Basic Information document as well
  as other appropriate documentation

10
Whats in a name the typical LSMS Survey

• Rigorous quality control
• Direct informants
• Precoded, close-ended questions
• Explicit skip patterns
• Extensive piloting of draft questionnaires
• Extensive training
• Well-defined field procedures (manuals)
• High level of supervision
• Concurrent data entry
• Small sample size
• Revisits to households to correct errors
• Emphasis on capacity building and public access

11
No two LSMS surveys are exactly the same

• Jamaica core and rotating module survey design
  (annual)
• Kazakhstan LSMS merged into the HBS
• Albania GPS (household, community, facilities)
• Nicaragua (1998) special follow-up for
  households affected by Hurricane Mitch
• Guatemala (2000) qualitative village data
  collection for sub-sample of villages to
  complement the LSMS data

12
No two LSMS surveys are exactly the same

• Special purpose topics in the questionnaire
• Tanzania (1993) contingent valuation questions
  (willingness to pay)
• Guatemala (2000) social capital module
• Bosnia (2001) mental health module
• Kagera region, Tanz. (1991-94, 2004) extensive
  module on adult deaths
• Special purpose samples
• Northeast China (1995 97) focus on
  agricultural activities in rural households
• Northeast and Southeast Brazil (1996-97)
• Kagera region, Tanz focus on HIV/AIDS

13
Survey Alternatives