Menu
Home
Log in / Register
 
Home arrow Environment arrow Research Methods in Anthropology: Qualitative and Quantitative Approaches
Source

VALIDITY—AGAIN

There are at least five reasons for insisting on participant observation in the conduct of scientific research about cultural groups.

1. Participant observation opens thing up and makes it possible to collect all kinds of data. Participant observation fieldworkers have witnessed births, interviewed violent men in maximum-security prisons, stood in fields noting the behavior of farmers, trekked with hunters through the Amazon forest in search of game, and pored over records of marriages, births, and deaths in village churches and mosques around the world.

It is impossible to imagine a complete stranger walking into a birthing room and being welcomed to watch and record the event or being allowed to examine any community’s vital records at whim. It is impossible, in fact, to imagine a stranger doing any of the things I just mentioned or the thousands of other intrusive acts of data collection that fieldworkers engage in all the time. What makes it all possible is participant observation.

2. Participant observation reduces the problem of reactivity—of people changing their behavior when they know that they are being studied. As you become less and less of a curiosity, people take less and less interest in your comings and goings. They go about their business and let you do such bizarre things as conduct interviews, administer questionnaires, and even walk around with a stopwatch, clipboard, and camera.

Phillipe Bourgois (1995) spent 4 years living in El Barrio (the local name for Spanish Harlem) in New York City. It took him a while, but eventually he was able to keep his tape recorder running for interviews about dealing crack cocaine and even when groups of men bragged about their involvement in gang rapes.

Margaret Graham (2003) weighed every gram of every food prepared for 75 people eating over 600 meals in 15 households in the Peruvian Andes. This was completely alien to her informants, but after 5 months of intimate participant observation, those 15 families allowed her to visit them several times, with an assistant and a food scale.

In other words: Presence builds trust. Trust lowers reactivity. Lower reactivity means higher validity of data. Nothing is guaranteed in fieldwork, though. Graham’s informants gave her permission to come weigh their food, but the act of doing so turned out to be more alienating than either she or her informants had anticipated. By local rules of hospitality, people had to invite Graham to eat with them during the three visits she made to their homes—but Graham couldn’t accept any food, lest doing so bias her study of the nutritional intake of her informants. Graham discussed the awkward situation openly with her informants, and made spot checks of some families a few days after each weighing episode to make sure that people were eating the same kinds and portions of food as Graham had witnessed (Graham 2003:154).

And when Margaret LeCompte told children at a school that she was writing a book about them, they started acting out in ‘‘ways they felt would make good copy’’ by mimicking characters on popular TV programs (LeCompte et al. 1993). [1] [2] [3]

BOX 12.4

THE MEANING OF DATA

In 1957, N. K. Sarkar and S. J. Tambiah published a classic study, based on questionnaire data, about economic and social disintegration in a Sri Lankan village. They concluded that about two-thirds of the villagers were landless. The British anthropologist, Edmund Leach, did not accept that finding (Leach 1967). He had done participant observation fieldwork in the area and knew that the villagers practiced patrilocal residence after marriage. By local custom, a young man might receive use of some of his father's land even though legal ownership might not pass to the son until the father's death.

In assessing land ownership, Sarkar and Tambiah asked whether a ''household'' had any land, and if so, how much. They defined an independent household as a unit that cooked rice in its own pot. Unfortunately, all married women in the village had their own rice pots. So Sarkar and Tambiah wound up estimating the number of independent households as very high and the number of those households that owned land as very low. Based on these data, they concluded that there was gross inequality in land ownership and that this characterized a ''disintegrating village'' (the title of their book).

Don't conclude from Leach's critique that questionnaires are ''bad,'' while participant observation is ''good.'' I can't say often enough that participant observation makes it possible to collect quantitative survey data or qualitative interview data from some sample of a population. Qualitative and quantitative data inform each other and produce insight and understanding in a way that cannot be duplicated by either approach alone. Whatever data collection methods you choose, participant observation maximizes your chances for making valid statements.

severity of punishment and independent variables other than severity of crime. Then you could test those hypotheses on a sample of courts.

Think this is unrealistic? Try going down to your local traffic court and see whether defendants’ dress or manner of speech predict variations in fines for the same infraction. The point is, getting a general understanding of how any social institution or organization works—the local justice system, a hospital, a ship, or an entire community—is best achieved through participant observation.

  • [1] Participant observation helps you ask sensible questions, in the native language. Haveyou ever gotten a questionnaire in the mail and said to yourself: ‘‘What a dumb set ofquestions’’? If a social scientist who is a member of your own culture can make upwhat you consider to be ‘‘dumb’’ questions, imagine the risk you take in making up aquestionnaire in a culture very different from your own! Remember, it’s just as important to ask sensible questions in a face-to-face interview as it is on a survey instrument.
  • [2] Participant observation gives you an intuitive understanding of what’s going on in aculture and allows you to speak with confidence about the meaning of data. Participantobservation lets you make strong statements about cultural facts that you’ve collected.It extends both the internal and the external validity of what you learn from interviewing and watching people. In short, participant observation helps you understand themeaning of your observations (box 12.4).
  • [3] Many research problems simply cannot be addressed adequately by anything exceptparticipant observation. If you want to understand how a local court works, you can’tvery well disguise yourself and sit in the courtroom unnoticed. The judge would soonspot you as a stranger, and after a few days you would have to explain yourself. It isbetter to explain yourself at the beginning and get permission to act as a participantobserver. In this case, your participation consists of acting like any other local personwho might sit in on the court’s proceedings. After a few days, or weeks, you wouldhave a pretty good idea of how the court worked: what kinds of crimes are adjudicated,what kinds of penalties are meted out, and so forth. You might develop some specifichypotheses from your qualitative notes—hypotheses regarding covariations between
 
Source
Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >
 
Subjects
Accounting
Business & Finance
Communication
Computer Science
Economics
Education
Engineering
Environment
Geography
Health
History
Language & Literature
Law
Management
Marketing
Mathematics
Political science
Philosophy
Psychology
Religion
Sociology
Travel