ESRA logo

Tuesday 16th July       Wednesday 17th July       Thursday 18th July       Friday 19th July      

Download the conference book

Download the program





Tuesday 16th July 2013, 16:00 - 17:30, Room: No. 18

Surveys in Developing Countries

Chair Ms Deniz Karci Korfali (Koc University)

Paper Details

1. Notes from the EUMAGINE Survey in Turkey Challenges and Heterogenous experiences in four research areas

Ms Deniz Karcä Korfalä (Koc University)
Ms Aysen Ustubici (Koc University)

This paper focuses on the heterogeneity of experiences faced during the EUMAGINE survey conducted in four different research areas in Turkey with urban and rural parts. We argue that the difficulties of survey methodology may not be homogenous within the developing contexts and even between the research areas located in the same region. This paper sheds light on the contrasting survey experiences in terms of access to households in urban vs. rural areas, in highly developed vs. underdeveloped regions and in regions with different migration histories.


2. Interviewer Effects in Survey Questions: Evidence from Timor-Leste

Ms Kristen Himelein (World Bank)

During the development of household surveys, questions are drafted, pilot tested, and re-drafted repeatedly to precisely target an indicator of interest and to eliminate any trace of bias in the wording. This process can be particularly challenging for questionnaires that examine subjective issues. Despite the best efforts of survey designers, the questionnaires must ultimately be administered by field enumerators, introducing the possibility of interviewer effects. There are different mechanisms through which the interviewer may influence the respondent, such as through a reaction by the respondent to some physical trait of the interviewer, or by reaction to verbal or non-verbal cues given by the interviewer. This can be particularly exacerbated in the rigid social hierarchy found in many developing countries. This paper uses data from a Law and Justice survey conducted in Timor Leste in 2009 (n=1213) to examine various hypotheses regarding interviewer effects. As the data was collected as a sub-sample of the recent national poverty survey, it includes objective as well as subjective questions. The dataset also includes metadata on the interviewer's age and gender, as well as the interviewer's opinion on the individual subjective questions collected prior to the start of the survey. Making use of the randomized assignment of interviewer-respondent pairings, I examine various objective and subjective questions for the presence of interviewer effects, and attempt to separate the role of observable characteristics and the individual opinions of interviewer in introducing potential bias into resulting indicators.


3. A Look at Cultural Intelligence in Developing Survey Wisdom

Mr Türknur Brand (Central Bank of Turkey)

How incorporating cultural intelligence can improve the work of survey methodologists in a developing country? In our global world, it is important to maintain comparative research across cultures. At this stage, especially in developing countries, there are some challenges to collect data acquired. How to approach different cultures is not simply a list of specific cultural “do’s and don’ts”. It acquires universal people skills which goes a long with cultural intelligence defined as ability to adapt to new cultural settings and to deal with other people sharing common cultural background and understanding. Survey methodologists notice the progress on managing cognitive issues at the culture level in developing countries /culturally diverse populations; for example data collection interviewing, improving the questionnaire wording in surveys. There are no written rules in the survey world to handle each developing country case separately. Turkey would be only one case study that gives idea about survey characteristics in that part of the world. At a later stage, after learning more about the cultural attitudes in surveys in different regions, the aim is to develop another ability to construct right measures for a comparative research/analysis across nations once it is required


4. Creating sampling frames with good coverage

Miss Anu Abraham (Department of Humanities and Social Sciences, IIT Madras)

This paper explores the challenges involved in employing large scale survey databases for studying migratory patterns and trends in India using two quinquennial databases, one national and one specific to Kerala (the state of study). Definitional changes of variables, missing variables and data points emerge to be a difficulty which hinders comparison between surveys and between different rounds of the same survey. Information asymmetry between researchers and respondents as well as under and over reporting of facts for perceived benefits is another major problem faced. The paper tries to explore this discrepancy between the data and field reality.