Confidentiality of the response from survey participant
Language translation problems for particular kinds of questions.
Telephone surveys or web based or household surveys in different countries
Special issues in countries that are low income, agricultural, non-industrial
Example(s): See http://www.websm.org, examples from ED 683 readings and, more important, the lectures.
5. Develop and undertake a systematic review of randomized trials in which the follow-up measurement of outcomes can be construed as two parallel surveys. Report on the design of the review conduct, and results using a structured format.
Example(s): School attendance, rates in Mexico (Progresa/Opportunidad), Honduras, Kenya, Colombia, or others, in RCTs that test whether a particular intervention works relative to a control condition.
6. Design a project that focuses on administrative records (which are like a census or sample). Examine whether and how well the evidence on quality of such records would match standards of quality in surveys. Or, analyze the publicly accessible records. Report on the design, conduct, and results using a uniform structure. Example(s): GIS, web accessible data from villages, cities, or states/provinces, and countries, universities.
7. Design and conduct a study (a systematic review, or pilot field research, etc.) on languages and translations as an issue in surveys of different language groups. Report on the design, conduct, and results, using a structured format. Example(s): See readings, and recall lectures from ED 683.
8. Design and conduct a study (systematic review or pilot field research, etc.) on how to pose sensitive questions and obtain candid responses. Report on the design, conduct, and results, using structured format.
This topic might include focusing narrowly on:
How to pose the questions?
Statistical methods for assuring confidentiality of response
National law on confidentiality of response
Risks to respondents
The topic could cover any country. Cross country issues are important.
9. Develop a literature review on attrition and cooperation rates in a particular survey area, or in a particular domain of randomized trials in which measures of each arm of the trial can be construed as a parallel survey.
10. Develop an essay on the appearance ideas about the ideas of statistical surveys in prose, poetry, and drama.
