Post date: Sep 06, 2012 3:36:29 PM
(Draws liberally from Russ Bernard's Research Methods in Anthropology)
Think Jane Goodall studying chimps
Involves long term interaction and proximity with group being studied
get close to people so they feel comfortable with your presence so you can observe and record their lives
Gain rapport and learn to act so that people go about their business as usual when you are around
don't go as far as going native -- the objective of research is ultimately to understand the natives on your own terms. Even an emic study is ultimately a model, in researcher's terms, of native point of view. Once you go native you are no longer doing research -- you are just living
Usually involves removing yourself daily from the research setting to put it into perspective
Doesn't rule out administering formal surveys or other structured data collection tasks
To understand how things work
either etically or emically
To understand how it feels to be a member of a given group
Notes
Emic study tries to understand how natives are perceiving their behavior, their world. What things mean for them, using their own categories to understand the world
Etic study explicitly uses researcher's constructs to understand the field setting. An example is Harris' analysis of Hindu Sacred Cow.
A person in the midst of a group can occupy any position along a continuum from pure observer to pure participant. The ethnographic field researcher occupies a middle position, perhaps leaning toward observer (participating observer) or toward participant (observing participant).
Typically share most of the life the informants, but not all. If studying hospital patients, don't need to pretend to be patient, or worse, a nurse or doctor.
Sometimes are forced by natives to take sides
kornblum, studying gypsies, when his group was attached by a rival group. The people he was living with "had no sympathy for the distinction between participant and observer. She thrust a heavy stick in my hand and shoved me toward the door.".
Doctors and psychologists run subjects through their experiments. Sociologists ask respondents to fill out surveys. Anthropologists (and all ethnographers) learn from their informants, and managers consult experts. All four (subjects, respondents, informants, experts) are sources of information, but the attitude toward them is different.
In general, terms like subject and respondent indicate an etic approach in which the source of information is not expected to understand the meaning of the information they are giving. The physician doesn't think you know what's wrong with you, so he asks a bunch of questions about symptoms and takes measurements of your blood, saliva etc, and draws his own conclusions.
In contrast, informants and experts are treated more like co-investigators. One typically assumes that the informant actually knows how and why he does this or that.
This assumption is not necessarily warranted of course -- I'm capable of producing grammatically correct sentences on demand, but in truth I have no idea how I do it -- my own mental processes are mostly hidden from me.
For very sensitive topics with real strangers, participant observation can/should take a year or more. Studies show that ethnographies that took less time tend to make scant reference to sensitive topics (sexuality, crime, witchcraft, political feuds, etc)
For "simple" settings with little that is sensitive (e.g., studying what happens at a laundromat), a few weeks may do.
Certain data are unobtainable by other means. Some groups won't let you see anything until you become part of their lives, as in the case of studying criminal networks.
Reduces reactivity (people changing behavior because they are being watched)
Help you formulate survey questions that are sensible and appropriately phrased. How many product surveys have you encountered that have questions you can't answer because the questions don't really apply to you?
Intuitive understanding of the context that allows you interpret the findings meaningfully. Example: Sarkar and Tambia study of Sri Lankan village. Concluded that 2/3 of villagers were landless. But anthropologists who did participant observation in the area knew that the villagers had a patrilocal marriage system in which a son received use of land but did not take legal ownership until the father died. In addition, they defined household as a unit of people that cooked in same pot. But in the local system, every married woman used own pot, so the number of households appeared greater than it was.
In some settings, you cannot be anonymous observer. To understand how the police operate, you must hang out at the police station, but the police will soon want to know who you are and why you are there.
Participant observation is respectful of informants. You are not just flying in with a questionnaire and taking their data. You are investing your time in them and moderating your demands on their time.
Among a set of reasonable sites, choose the one that is easiest to get into.
Be prepared with lots of written documentation about yourself and your project. Grant proposals, resumes, examples of past work, and letters of introduction from your university, granting agency, boss, etc.
Depend on social capital. Come with a list of specific people to look up that are acquaintances of people you know. They can help in lots of ways, including legitimizing your presence
With organizations, start at the top and work down -- get to the gatekeepers first. Assure them confidentiality, and don't offer a quid quo pro that could harm your informants
Have ready answers that describe your research, what will be done with the results, etc.
Do your homework and learn about the setting before you get there. Get comfortable with the physical setting.
Learning the language. This is especially important for management researchers working in their own countries, because they think they already know the language. But typically they don't really know full business language (e.g., terms from accounting, finance, marketing, etc.) and they underestimate the unique meanings that words acquire in any given organization
Case of giving talk on networks at the Academy of Management. A questioner said this was all well and good but she found that personal relationships were a lot more important than networks. Since I thought networks consisted of personal relationships this was confusing. Eventually it came out that in her company, "networks" referred to virtual groups participating in a listserv, and she didn't realize that others use the term differently.
You will need to adopt the language of the group you are studying -- up to a point. When studying subcultures within your own culture, imitating others' way of talking can seem insincere, silly or mocking (e.g., White student trying to sound Black).
Explicitizing. Becoming consciously aware of what people are doing and saying. Otherwise you are taking in as much what you expected to see as what you actually saw.
Kronenfeld (1972) study asked diners leaving several restaurants about the what the waiters and waitresses were wearing. Respondents showed high agreement on both -- yet none of the restaurants had any female waitstaff.
In some ways, making a culture explicit is the whole purpose of ethnographic research
Maintaining posture of apprentice. Treating the informant as an expert in her own culture (although this may not be true), while making clear that you are ignorant (but learning). You will learn more if you remember this and don't become too cocky.
On the other hand, while being clear about your ignorance, you also want to behave in a way that is appropriate to the culture and does not make people nervous about giving information away or letting you into situations (e.g., boardrooms) to observe. If you are too naive people won't trust you with the information.
Storing information. The biggest challenge in ethnographic fieldwork is remembering everything you are learning. There are a number of strategies that can help:
tape record what you can.
take notes as you go along
ALWAYS write up at least a page of notes each evening before going to bed summarizing everything you can think of. This is super helpful for inserting the information in your brain so that you can ask better questions the next day.
Patience. The whole idea of participant observation is that people become comfortable with you and start revealing things they otherwise wouldn't reveal. And, because you are learning like an apprentice, you learn the answers to key questions at a time when you are ready to understand the answers. Hand in hand with this is the fact that your questions are better because you know enough to ask them properly. It also shows respect for the informants that you are willing to wait for the information you need.
Practicing Objectivity. There is a Zen to ethnographic research and part of this Zen is to hold beliefs lightly. It's not that you try to empty your mind of pre-conceived notions (that would reduce you to the intelligence of a table ), nor that you don't believe what you believe (a bad psychological state) but that you are aware that alternative views are possible.
Be extremely wary of your own feelings of personal outrage. It is a little known fact that outrage, indignation, and similar feelings are pleasurable. They confer a sense of self-worth and purposefulness. You WANT to believe things that make you feel that way, and so you can fool yourself. Also, such feelings typically have something to do with social competition -- your feelings about the exploited workers at your research site have everything to do with your ongoing mental battle with those mean Republicans that you are sure are ruining the planet.
It depends on the research question and research setting, of course. But here are some generic strategies that may be helpful in many situations
One approach is to start with the physical layout of the research site. Create a map. Who is where? This can be the basis for your first questions
Another approach is "a day in the life of ____" where you follow one person or one group around all day and get an idea of what happens in a day, what activities are engaged in.
Another approach is the oral history. Asking individuals to tell you their life histories that led to where they are today. This can also be done at the organizational level by using informants or secondary data to pull together a history of the organization
Finally, there is the environmental scan, where you get an idea of what is the environment that your research target is reacting and adapting to. I.e., what are the key contextual factors.
You may feel that for ethnographic research you should not have a research question -- that this should emerge from the research itself. This is not a good idea because it is a recipe for thrashing and/or going in every direction at once. Better to have an idea, even some very specific ideas, but allow the research question to change as you go along. This provides forward movement.
One kind of research question that is appropriate for ethnographic fieldwork is "How do people in this group accomplish X?" For example, see John Holt's ethnographic books on why children fail in school. More generally, the research is specified by the dependent variable -- some distribution of outcomes for which you are trying to understand the mechanism that produces it
Another kind of research question is centered on the independent variable. For example, you are curious about how a person's position in a social network affects what happens to them.
The most productive kind of question, in my opinion, is one that investigates the relationship between two pre-specified variables. For example, you are interested in how attractiveness influences managerial success. Or you want to know how a child's position in the classroom friendship network affects their ability to succeed in school. The objective is to understand the mechanism -- the series of causal links -- that relates the independent variable to the dependent outcome. One of the advantages of this approach is that you don't have to consider all of the causes of Y (if centered on dependent variable) nor all the consequences of X (if centered on the independent variable)