Copyright 2010, The World Bank Group. All Rights Reserved. Part 1 Survey Design Produced in Collaboration between World Bank Institute and the Development.

Copyright 2010, The World Bank Group. All Rights Reserved.

Part 1

Survey Design

Produced in Collaboration between World Bank Institute and the Development Data Group (DECDG)


Survey design

Several decisions must be made to effectively design a survey

A. What is the goal of the survey?

B. What will be measured? (concepts)

C. Who should be measured? (units)

D. When? (frequency and timing)

E. Where? (areas)

F. How? (survey type)

Copyright: www.cse.unt.edu/



Knowing where you want to finish tells you where to start• Ask questions to define desired outcome:

1. What is the purpose of the survey?

2. What topics do we want to know about?

3. What will we do with the information (use, users)?

4. Will the survey be asked once or repeatedly (frequency)?

5. What is the resource situation?



1. Purpose of the survey• To be established at the beginning of the process in written

form• Clear and concise statement important

Written purpose useful as future reference :• when researching topics and questions used in similar

surveys• when there is uncertainty whether certain questions belong• if there is pressure to include items that don’t belong• when choosing questions to reduce the length of the survey


2. Make a list of topics to include in the survey:

• Include standard labor force survey topics• labor force status, demographics, earnings, occupation

and industry, hours worked, etc. • Written list:

• avoid topics being forgotten • holds a place for topics when ordering sections• serves as a reminder to find standard accepted

questions for topics




2. Make a list of topics to include in the survey (continued):

• Add country-specific topics of special interest or importance• ethnicity, tribe, citizenship, region of residence• refugee or displaced person status• polygamy and how polygamous households are organized• urban and rural differences, subsistence agriculture• development and/or education policies and programs• living conditions and access to infrastructure• child labor, the role of women



After listing topics of interest, ask:• How do these topics fit together?• Are any topics missing?• Are any topics included that shouldn’t be?• Have we researched existing, respected surveys to see if they

include anything we haven’t thought of and need?• Does looking at the list raise ideas of other topics that should

be included?

Unlikely every topic that makes this initial list will be included in the survey

There isn’t time or space for everything

We must be selective



3. Knowing the use of the information ensures collection of better data– Production of data that adhere to International Labour

Organization (ILO) standards has many benefits• Concepts are defined and can be used when designing the

survey • Allows for comparisons across countries• Adds legitimacy and build confidence in the data among

users• Reduces staff burden because data needed for reporting are

readily available in the required format



3. Use of collected data (continued)

Consider data analysis needs • Think through any “why” or “what about” type questions that

might arise. • Plan to collect data that will allow you to answer these

questions.

Consider policy-making needs or needs to judge impact of programs

• Important to cover all topics• Topics shouldn’t be too narrow or tailored to a particular

policy or program, especially if comparing data over time



4. Frequency of the survey

One time surveys and those that will be asked repeatedly may be very different:• One time survey can ask about current events and

programs• A repeated survey :

• must be held consistent• allows you to see changes over time• should not ask about current events


B. What will be measured (concepts)?

List all concepts to include:• Standard demographic characteristics, such as:

SexAgeMarital statusEducational attainment and school enrollmentRelationship to other household members

• And concepts that can be constructed from them, such as:Presence of children in the householdAge of youngest childshould not ask about current events



Basic labor force measures include:

• Labor force status:• Employed• Unemployed• Not in the labor force

• And measures derived from them:• Labor force participation rate• Employment/population ratio• Unemployment rate

www.lankanewsweb.com



For those who are employed, standard topics include:

• Hours:• Full- and part-time• Actual and usual

• Occupation, industry, and class of worker• Earnings• Multiple job holding:

• Characteristics of other jobs

www.blog.norway.com



Many other additional topics exist, such as:

• Discouraged workers and the marginally attached• Formal and informal employment• Job tenure• Subsistence agriculture

Eventually, the number and depth of topics chosen depends on data needs and resource constraints



After making the initial list of desired concepts:

Consider tables needed and stories you might want to tell to see if any pieces are missing.

Consider all possible outcomes. Do they suggest any needed follow up?



Have you tried to measure these concepts before?

How was it done? Did it work well? What issues, solutions, and unsolved problems were there? What lessons can be learned and applied here?

Copyright 2010, The World Bank Group. All Rights Reserved. Part 1 Survey Design Produced in Collaboration between World Bank Institute and the Development.

Documents

survey slide

survey design

survey type copyright

list of topics

world bank group

world bank institute

selective slide

written list