Presentation on theme: "The MICS3 Process. By now we have decided to do a MICS3 We have answered these questions: Why is the survey being done? How will the survey results be."— Presentation transcript:
The MICS3 Process
By now we have decided to do a MICS3 We have answered these questions: Why is the survey being done? How will the survey results be used? To whom will the survey results be addressed?
Objectives zAn overview of the entire survey cycle zTo review the key decisions for survey organisers zCritical steps and sequences zHow decisions in one area affect other survey areas
Main activities 1)Definition of survey instruments 2)Pretest the survey instruments 3)Sampling (presentation) 4)Main training (presentation) 5)Fieldwork (presentation) 6)Data processing (presentation, workshop) 7)Data analysis and report preparation (presentation and workshop) 8)Dissemination of results (presentation and workshop) 9)Archiving and further analysis
Decisions to make before starting zIdentify survey coordinator and key senior personnel zDecide on level of aggregation and sample size zCalculate the number of interviewers required and days needed in the field zEstablish a timetable zCalculate how much the survey will cost Review all items above to match budget!
Level of aggregation for estimates zNational Level 6,000 households, 300 clusters relative error of 12% or less Subnational Estimates Regions Urban and rural Provinces or states Other population groups of interest
Factors influencing cost zUnit and total costs vary considerably across countries zDepend on a variety of factors, including sample size labour costs currency fluctuations etc. zMain component: fieldwork expenditures
Factors influencing fieldwork costs: example zSample size: 6,000 households zEach interview takes 60 minutes plus 10 minutes to move from house to house (70 minutes total) depending on questionnaire length! zWorking day is 6 hours (or 360 minutes) excluding lunch break and transport to and from cluster zThus 360 / 70 = ~5 households per interviewer per day (also allows for call-backs) z6,000 / 5 = 1,200 interviewer days required
Variables in field work zThree variables to adjust to determine best combination of interviewers and time frame Number of interviewers Number of days of field work Size/number of teams zOptions: 30 days of fieldwork x 40 interviewers 35 days of fieldwork x 35 interviewers 40 days of fieldwork x 30 interviewers
Timetable zWhen to start? Major reports, other surveys, weather zGood Planning What do we need to learn from MICS? Uses of the data
Tasks Identify resource person and survey coordinator, plan survey Adapt and pretest questionnaires Carry out sampling Complete logistic arrangements Select and train interviewers Conduct pilot study and collect data Enter and clean data Complete data processing Prepare reports (preliminary and final) Months Timetable for a survey of 6,000 households Assumptions: 6,000 households; interviewer completes 5 households per day; 1,200 interviewer-days; 30 interviewers; 40 days in the field; plus 7 days for one day off per week; plus transport days = 54 days
Survey budget items - 1 Note: This assumes a sample size of 6,000 households and excludes listing/mapping exercise.
Survey budget items - 2 Note: This assumes a sample size of 6,000 households and excludes listing/mapping exercise.
Survey budget items - 3 Note: This assumes a sample size of 6,000 households and excludes listing/mapping exercise.
How decisions affect other areas of process zMICS3 Standard Methodology Goals ==> Indicators ==> Sampling and Core Questionnaire ==> Field Work Procedures ==> Core Data Processing ==> Core Data Analysis and Reports zThe Accumulative Effect zTiming zCosts