There is no such minimum number in practice, I believe. However, the interviews or focus group discussions will not add any value to the data, once the theoretical saturation is reached, i.e. responses from informants are no longer different to add variation to the data.

Can you elaborate more on your question, please? Do you mean the total number of groups that you will involve in the study or the number of people in each group? There is no specific rule of sample size in qualitative research. The total number of groups that you need to include in your study will have to be determined by your study objectives and the amount of data you need to cover the objectives. Since your unit of analysis is group, the number of informants in each group should be determined by the extent to which the voices of the informants represent the group. The formal position of an informant within the group in this case should also be taken into account.

Great answers! As an ethnographer, I agree with what was said.

In Ethnography you do not have such rules as a required number.  It is wise, though, that you do show triangulation,  meaning, that you gather data from different angles and sources before you reach any conclusions.  As our colleagues mentioned above, you stop gathering data when you feel that there is a saturation and that you know enough about the phenomena.  Another thing: look at the quality of informants, because if you increase numbers of pax, but they are all  marginal.... then you will surely miss something.  It is also important, if possible, to identify the leaders and get to be with them as well

I do not think there are any numerical regulations apart from the theoretical saturation requirement. In order to achieve saturation as quickly as possible, I think it is important the selected informants to be in the "well-informed" category.

I agree that the number of informants in ethnography is context dependent and should not be bound by hard and fast rules.  However, consistent application of participant observation reports are indispensable to providing reliable context.  I have also found that informant interviews and recorded interactions are greatly augmented by bringing multiple informants into joint conversations and activities.  Social conventions, reflections, and improvised discussions have been extremely important in providing access to multiple perspective and unforeseeable insights.  

Although scholars view ethnography as a subjective method, it is highly recommended as a qualitative and holistic  approach that takes into account multiple and differing aspects of human behavior.  From comparative (ethnological) perspectives, it is good to have a large group (or groups) as individuals belonging to it/them might provide you with different worldviews enriching the outcomes of your study.

 The rule is the saturation point in the answers.

I was going to answer, but everything I had to say was covered in a much more informative manner by the above pros o/

Thank you everyone for the answers given.

Yes, I will elaborate.

My question is "The number of people in each group".

I will only have one group in my study.

Many thanks to all.

Amnah

Amnah -

Sorry to keep the waters muddy but the answer depends greatly on the kinds of questions to be posed. If you are asking about rice planting techniques or seasons of the year, you can get by with just a handful. However, if you are dealing with issues that are more variable the number goes up. My original study of values used scores of villagers and more than a hundred others.

The number to be targeted in a given group will be determine on the resourcefulness you may have from a particular group. but at least one each should be the minimum target. Then you base more on the resourceful group. My take

Amnah-

Sorry to muddy things a bit, but the number of people needed will depend heavily on the nature of the questions being asked. I they concern mundane, non-abstract cultural elements such as the names of the seasons, how many paths in a village, etc. than the number can be quite small once you get agreement among your respondents. If the issues involved are complex and abstract, then many more will be needed. In my original values study I interviewed more than 20 villagers and sought corroborating information from more than a hundred others.

Yes, agreed-- it dependsIABISMO DE PASIÓNrsgsgswwe

It depends on the topic, competence of respondents,age, etc.At least you will need minimum three respondents to form a group.

Amnah..it is totally depend on your scope of study. Maybe you just can select the main ethnicity to make it simplify. All the best.

Yes, thank you very much to all.

Yes, I do believe as ethnographers, there are no strict rules to the number of informants in a particular group. It is up to the ethnographer to justify and decide. Also say the number of informants in a group has a limited time to be together, is another element to think about. Again here, the ethnographer decides.

My best regards to all and I welcome other point of views.

Thank you.

An important point is for the ethnographer to make clear what number he/she is using as a sample or how many informants the ethnographer is using.

Yes, Amnah the ethnographer decides. But, good ethnography also requires knowing and speaking the colloquial/native language. Translators, in case if ethnographer is not a native speaker, migt help, but tI'm afraid to say that the analysis of the data will not be highly reliable.

Thank you Julie, for that point, it is important always to mention the number of informants.

Yes, El-Sayed, I remember reading on translators and how translators could even impose their own interpretations if the researcher is not careful or aware. However, in my study I do not need a translator but point taken, to ensure trustworthiness "thick" description or audit trail could accomplish this.

In ethnographic research, there is no hardbound rules about number or units. However, a thick description of the events, situations and cases is very important. The cases chosen should answer to the research problem. The sample taken should justify the research.

Hi Amnah, you are right as far as you do not need translator. Indigenous ethnographers in such a case have a good cognitive privilege.

Thank you for your point of view, Sukanta and El-Sayed,

I have an inescapable question which I will start in researchgate, and I would be grateful if you or others could comment.

It's concerning ethnography's literature review.

Thank you.

Amnah

I agree the all answers, that in ethnogrphy research there is no rule about amount or size. All depend on research problem. The researcher as instrument in qualitative research, and data collected must be valid, through triangulation and other method.

Thank you Santi.

You mentioned "triangulation and other method". Could you help me with "other method"? It would be interesting to know.

Best regards and thank you

Amnah

Consider the emic/etic aspects of the word "group" -- that is, what YOU might consider a "group" of (fill in the term you use to identify those you study), might not be what (those you study) consider to be a "group" -- that is, they might contest the inclusion or exclusion of some members, or might contest the way you configure the group (i.e., by location, kinship, etc.)

You'll get your best answer if you ask several of your most helpful informants what THEY would consider the optimal number & configuration of a representative group.

Ethnographic work emphasizes the emic perspective (or interviewees' point of view), though the etic (the researcher's view or analysis) is also considered. The inquiry of asking the informants to suggest number of representative interviewees might be questionable.

In folklore research, the stance of the investigator is "shoulder to shoulder" with the informants, meaning we consider their wisdom and knowledge to be (at the least) equal to ours. It's considered best practice among folklorists to value the informants' explicit instruction as part of forming notions about constituency. The co-author of a classic work on ethnography, ("People Studying People") Michael Owen Jones, who directed my dissertation, used to pose to students in his Fieldwork seminar, "How can you find out what a narrator means by [a phrase with a meaning known to members of the group but opaque to outsiders]?" The answer? "Ask 'em!" He advocated being open and inclusive. I follow in his footsteps; many others do, as well.

Thank you for your views Michel. I take it you are looking into Belief system i.e. religion, folk philosophy, and ritual. Please correct if I misunderstood.

Amnah

Once again, informants are a great source of information about their communities, believes, religions, rituals and oral history (and so like). When an ethnographer ask them (informants or interviewees), they are supposed to provide answers about their cultures/societies. They also may introduce the ethnographer to other people (informants/interviewees), but they do not get involved in the methodology or what is the good number of the people being studies because such a question must be the focus of the ethnographer/anthropologist who assess the number of informants based on that nature of the research/project being addressed. I hope it is clear.

In much ethnography of small scale societies, one of the greatest problem is small sample sizes. Studies in industrialized countries or in cities can get responses from thousands of folks (but not necessarily fine-scale data). When working with smaller communities, I would suggest getting as many people as possible in your sample. If you do any statistical analyses, sample sizes are critical to being able to posit meaningful comparisons, even for simple descriptive statistics. I heartily endorse El Sayed El-Aswad's point about working in the native language. Far too many of my behavioral ecology colleagues are doing short-term fieldwork using translators, and relying heavily on questionnaire data. In my behavioral ecology work, observational data were an improvement over the quality of interview data about topics that needed quantified information for statistical analyses of simple frequencies of basic behaviors such as how often folks went out foraging, how far did they travel, how many food resources did they encounter and how many of these did they take, what was the weight of returned foods, what kinds and mow much raw materials were collected on trips, etc. Some critics of scientific ethnography think these are boring creature adaptations and that what people think is much more interesting. Maybe that is true, but to get to the mind, I first need to understand adaptations to any particular environment. I strongly support behavior observation methods, but I subscribe to quantitative data collection as an important way to determine what people actually do, not just what they say they do or their culture considers appropriate. I'm interested in subsistence quantification, land-use, demography, etc. so I strongly value quantified observations. I certainly spend as much time talking with folks as possible. I'm not studying the people I work with, I am trying to learn from them about how they live and what they do. I only have an anthropologist's perspective, they are lifetime hunter-gatherers (I work with Pume hunter-gatherers in the savannas of Venezuela) or small-scale agriculturalist (I also work with Maya farmers in the Yucatan peninsula of Mexico), and it is my ethnographic job to learn from them. I have greatly benefited from constantly improving my indigenous language skills, and no longer believe in anthropological ideas such as taboos, superstition, myth, and probably not even religion, as most anthropologists conceive of it. The ability to work in an indigenous South American language has shown me the profound scientific knowledge these people have about the natural world around them. The only things they can't accurately ascribe causation to are things that require different tools to look at the world (i.e., microscopes and ideas about germ theory in the case of disease) or that are hard to predict; among foragers in Venezuelan one important example of that was lightning. Their explanations for things might sound silly in translation (hence ascribed by the anthropologist to superstition or myth), but they often are literary phrases or very literal, and cannot be appreciated without good language competence. However, direct answers to questions are not always forthcoming. During my 30 months of fieldwork in Venezuela, many of my questions initially seemed ridiculous to the folks I was working with. Although I was an adult, I was asking a child's kind of question, and often the answer was just what would be said to a child: "Because! That is what we do!" The Pume appreciated that I operated only in their language, but it seemed crazy to them to explain to an adult something every child should know. They knew there were many basic things I did not know, but like many adults, they assume children will absorb the reasons and understand them eventually. Sometimes it would take 18 - 24 months to get someone to voluntarily explain something to me so that I could begin more sophisticated learning about a topic.

Thank you El-Sayed and Rusty. I strongly agree your comments on sample size. Again emphasizing your stand in writing is additionally important, e.g. positivism or postpositivism. Yes, ask any multilingual person, to understand a community in-depth, you HAVE to know the language.