Using ethnographic methods you put yourself in a strange position. In a foreign country with a new language, you already feel like the outsider coming in. Using ethnographic methods you must let go of your own presumptions and assumptions about a group of people in order to effectively learn anything about them. You cannot look at them the same way you look at people in your own culture. It is important to put aside personal feelings and not to judge, otherwise conflict might arise.
People are intuitive and can sense others feelings. If you disapprove of a native practice, people might close down to you. If you really want to get something specific you have already decided out of your research, you will unwittingly push people in that certain direction. People, by nature, like to please other people, especially someone who they might have a sense of reverence or respect. If someone realizes that you want them to say a certain thing, or talk about a certain subject, even when they don’t believe it, they might do so simply to please you. This is a bad thing.
That is the difficulty when performing an ethnographic study. While you may have a certain hypothesis, you cannot let that be known to your study group, and sense your hypothesis was developed on the terms of your own culture, you must be prepared to throw it out. Those who follow the basic principles of grounded theory believe that you should approach your study with NO PRECONCEPTIONS or hypotheses.
Remaining objective is essential. In this case, objectivity means that you must look at the culture in its own terms and not push your own reality or judgments upon it. How can you judge one culture based on the ethics of another? Cultural relativity is necessary when approaching new research in order to conduct an effective ethnographic study.
Interviewing is an effective method to learn from people what they believe, how they think, and how that affects their life. There are two different forms of interviews: fully-structured and semi-structured. Interviews should be crafted to answer a research question. What is it that you want to learn? This is important because if an interview is aimless, it might not produce anything you can use for your research. If you find that the questions you are using are not giving you the type of responses you want (like they are not giving you details about how the drought affects their life, just implying that it does), you should revise your set of questions until you get one that is effective. Your first few interviews might not produce the results you get in the end, but that is a necessary learning stage.
Fully-structured interviews are interviews where the questions are set in stone. You ask the questions to the interviewee, and you do not vary from those questions. Semi-structured interviews also begin with a set of questions, but you have the flexibility to add additional questions based on the interviewees responses. This form is much more flexible and allows for more specific questions to each interviewee. However, with fully-structured interviews you can more easily compare the responses from one person to the next, and do not have any separate questions that were asked of one person but not the next.
For sample interview questions, click here.
Participant observation is beneficial to immerse oneself into a community to gain a deep knowledge about the intricacies and inner workings that could not be obtained from literature or a type of method where information is learned second-hand. It yields insights into peoples lives and customs that they would not be able to tell you if you just asked. The researcher lives within a certain context, maintains relationships with people, participates in community activities, and takes extensive and elaborate notes on the experience. It enables the researcher to see the culture without imposing their own social reality on that culture. Years or months of research lead to months of analysis of the journals and field notes to convey the research findings within a theoretical context.
Participant Observation is also useful in order to gauge the difference between what people do and what people say they do. For more information on participant observation, click here.
Surveys are usually used as a quantitative method. Surveys should be crafted to get opinions and beliefs without leading the person being surveyed in a specific direction or to a certain conclusion. They should be non-inflammatory and they must not be too long. It is difficult to get all the information you want to learn out of a medium-length survey, but the longer it is, the less likely people will complete it. You must avoid putting your own emotion into a survey, because the likelihood is that you might anger someone, or push them in a certain direction because they believe that is what you want to hear. If you ask which hazard is more dangerous to a community (a. earthquake, b. volcano, c. landslide, d. geothermal activity, e. disease) and then proceed with twenty questions about volcanoes, you are leading them in one direction- and from that place they might over-exaggerate their personal beliefs or quit filling out the survey.
The survey is developed to get demographic information on the person being surveyed as it should be done anonymously. The survey questions may include true/false, ranking questions, or multiple choice answers so that the answers can be assigned to a number to be quantified. Using demographic information, you can see if there is a difference in responses depending on race, gender, age, economic status, religion, political affiliation, or any other classification that might be present. Many different types of statistics programs exist to best calculate the results. Open-ended questions can be used, but a system has to be developed to quantify the answers, or the answers can be used more qualitatively.
Who do you distribute the survey to? How do you distribute? Surveys should be distributed at random if you are looking to estimate some population parameters. Then, a scientifically drawn unbiased survey is the only option. A random draw is by using a means of chance, like choosing every 8th person in the phone book until you have enough recipients, in order to attempt to get a representative population. (Again, if that is your goal.)
It is acceptable to distribute surveys to a more targeted audience, for instance to understand a local healing ritual, if you are looking for cultural data. Cultural data requires experts. Then you ask people who can “offer expert explanations and who represent the intracultural variation that we find in all societies” (Bernard 2002). This is called nonprobability sampling.
Survey can be distributed a number of ways, however in the developing world literacy rates need to be taken into account. If the population your are studying is mostly illiterate, perhaps going door to door with your survey will be a more effective means for a higher return rate. You should always take into account that odds are you will only receive a fraction of your surveys back (30% would be a good return rate), and this should affect the number of surveys you should distribute.
How many people should you distribute to? You should always take into account that odds are you will only receive a fraction of your surveys back (30% would be a good return rate), and this should affect the number of surveys you distribute. Your sample size is very important, look further into detail with what you want to do and the recommended sampling size before starting the process. If you try a more focus group approach results cannot be generalized. Because focus group participants do not represent a big enough sample size, the information gathered cannot be used to make statements about any larger population. Thus the researcher is restricted in what s/he can say about the significance of the findings. Sample size depends on the following:
· The heterogeneity of the population
· Number of subgroups in the analysis
· The size of the subgroup
· How precise you want your sample statistics to be
For single proportions, like how many people in a population approve or disapprove of something, you need about 100 respondents to be 95% confident. But if you have more specific questions for more specific groups, you will need more respondents.
For some tips on writing surveys, click here.
For a sample survey, click here.
While the interview and participant observation are widely used and valued qualitative research methods, problems can arise.
It is extremely important to record all of any interviews you may conduct as memory is not a sufficient source for citation. Before a project is begun, you must decide for yourself whether you will perform full or partial transcriptions. The more you transcribe, the more protection provided for your research and analysis. However, the facts remain that full transcriptions are very time consuming, and the majority of your interview will not be useful for your research. Therefore, partial transcriptions are good for transcribing only the necessary or most valuable parts of your interview. You may not know what is the most valuable information you receive when you hear it, which is why it is beneficial to take light notes during the interview and extensive notes of what you remember after the interview. Do not focus heavily on your note-taking so that you miss what your interviewee is saying.
Participant Observation Tools
A research journal for field notes is a very practical way to keep track of your observations. After an extended period, flipping through your journal, you might notice patterns that you had not realized were prevalent as you witnessed them in person. Also, as memory fails it is very important to have detailed notes of what you observed; otherwise your observations are meaningless.
Journals are also valuable for realizing your own biases or prejudices. It is here that you might begin to question some of your interpretations, and here where you might realize possibilities that had not yet occurred to you. When conducting social fieldwork, you should always question yourself to make sure you are not getting in the way of your own research. It can also be valuable if you are having trouble connecting to your interviewees to see what it is you are doing that offends or distances them.
Always, always, always test your equipment before travel, and everyday before going out into the field. Fieldwork is a learning experience, however, it would certainly be more valuable if the learning experience was about your research topic rather than temperature at which tapes melt.
Qualitative vs. Quantitative research? A resource for the best approach, click here.
Books on Methods*:
Bernard, H. (2002) Research Methods in Anthropology: Qualitative and Quantitative Approaches, Third Edition. New York: Altimira Press.
Orcher, L. (2005) Conducting Research: Social and Behavioral Science Methods. Glendale, CA: Pyrczak Publishing.
Stallings, R. (ed). (2002) Methods of Disaster Research. California: Xlibris Corporation.*
*For information on qualitative methods, the following chapter is particularly valuable: Phillips, B. (2002) “Qualitative Methods and Disaster Research.”
Yow, V. (1994) Recording Oral History: A Practical Guide for Social Scientists. Thousand Oaks, CA: Sage Publications.
*Ethnographic methods are huge subjects and generally have books written about them rather than articles. The following pages contain articles and case studies for a literature review of some social science theories and topics on hazards. These materials use both qualitative and quantitative methods.