How do you do survey in urban areas?

From SurveyWiki
Revision as of 19:41, 12 October 2011 by Admin (talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigationJump to search
Questionmark.png

Does anyone have ideas on how to get groups of people together to interview in urban areas of 50,000-100,000 people? One idea we've had is to go through churches, but we're not sure if only interviewing Christians will skew our results regarding language attitudes. Is there any method that you've used (or just thought about) in the past that would work well?

Markets

One way done in the past in Mexico was to use the market as a source of people. Markets can be key. Much could be learned about LWCs, but also reported inter-comprehension and proficiency from market-goers who originate from smaller towns (that may be less accessible).

Church Gatherings

In some areas several churches will get together regularly. In some areas of Papua New Guinea, for example, many churches meet together at one of the circuit's church locations on the last Sunday of the month. You might be able to arrange a similar multi-church get together. One idea is to send invitations out to people in the area or put up signs so everyone knows it is open to all. You could even offer food/entertainment like a Bar-B-Q. Two ways to get data from people at the event: Mingle and try to learn people’s stories or put up a table with a poster telling what you are doing and see if anyone wants to sit down and chat. Although you should get a wider cross-section then just church people, working with or through established churches can help defray costs.

Government Offices

Towns of 50,000-100,000 are large enough to have government administrative offices where records are kept and where a few experienced workers might be employed. Those who are experienced (with more knowledge) might be thrilled and honored to tell you what they know and also to provide information (maps, census or population figures, etc.) They may also introduce you to key individuals who know the area and people. This occurred regularly during surveys in Cameroon, and I often took such persons with us for the full duration of our research in a given language area. We paid them a per diem and of course provided for their meals.

Sampling Ideas

To avoid possible skewing from sampling only Christians, sample multiple types of religious places of worship if appropriate. Sampling large urban areas can potentially be huge surveys. It may be that by evaluating your purposes and goals, you may find a way to narrow your focus, and in that process, to simplify some of the sampling. You may also end up doing the survey in phases, and evaluating at each stage if you have answered your research questions to a sufficient level for the next step of application or program planning, or if your team will need to proceed to the next research phase.

Gathering Places

Are there culturally-appropriate gathering places where people might "just hang out" after work? I'm thinking of things like pubs or tea houses, but also anywhere that people have to wait for something. Sometimes people who are waiting around can be helpful, and they don't always have anything better to do. You would need to figure out ahead of time what information you need from people in order to find them again for follow up work. One way we approached sampling of urban areas in central Africa was by neighborhood and district (this was how a survey using markets was accomplished by university researchers as well). You could choose a kind of natural gathering spot (market, well, bus stop, "hang out" place), and pick X number in each district and neighborhood.

Social Networks

You might also do some pre-survey work by investigating existing social networks. That may provide a natural "in" for researchers, and also provide some self-selecting kinds of groupings of people. Knowing about social networks can also provide important social indicators for language use and contacts between sectors of the population that in turn can help predict areas where you expect to find variation, various levels of bi- or multilingualism, etc.

University Students

Would it be feasible to get university students together for a group interview? You'd probably want to get enough biographical data from each to ensure they grew up in the urban area you're focussing on. Also, the students might have some ideas of how/where you could get other groups of people together to interview. For example, some might have relatives that work in businesses that would be open to your doing a group interview with the employees. If you can first demonstrate to the students that the interview is non-threatening and that you truly desire to learn more about their area, they might be more eager to help connect you with other possible groups to interview.


Go back to the Survey FAQs page.