Field work

Field work is research conducted in a natural environment, rather than in a laboratory or other artificial setting.

For example, you might go and live with a community for six months, or two years. You would spend your time participating in the community, and also observing and interviewing people. Each evening you would likely write huge amounts of notes (field notes).

The archetypical example of field work in the social sciences is an anthropologist living with an indigenous community some where in the world, and documenting their culture, language, and way of life.

However, field work can be done in any community. For example, a family home in Macquarie Park, or an office in the CBD of Sydney.

In depth interviews

In depth interviews involve sitting down one to one with a respondent, generally for an hour or more. Sometimes over several days. A detailed interview is conducted, generally from a set of prepared questions. However, there is considerable flexibility in the structure of the interview, and the questions are more a checklist for the interviewer to make sure they covered important topics. The interview has the feeling of a conversation, rather than an interogation or questioning.

Focus groups

Focus groups assemble a group of people to discuss a product, issue, or topic as a group. They are often used in marketing, political campaigns, and finalisation of film and TV programs.

Open ended survey questions

One of the simplest and cheapest ways to get qualitative data is simply to include open ended survey questions in your survey. For example, if you were surveying migrant domestic workers about exploitative working conditions, you might ask this pair of quantitative and qualitative questions.

1. Has your employer ever stopped you from leaving the house?

Yes No Not sure

2. If you said Yes, or Not sure, please provide details. _______________________________________________________________

Ethnographic sandwich

The ethnographic sandwich is a way of thinking about ethnographic research (field research) in a mixed method (i.e. mix of qualitative and quantitative data) study.

The key idea is that ethnography can be like the two peices of bread on a sandwich - coming before or after (or both) the quantitative research (such as a survey).

Each peice of the sandwich has very different purposes.

Thematic analysis

Most analysis of qualitative data focuses on identifying key themes that are repeated across the data (such as interview transcripts). If themes recur across interviews, then generally the researcher will draw conclusions and inferences from these.

For example, we can imagine a hypothetical study involving interviewing parents from a marginalized community. Researchers asked parents what were the major problems which were preventing their students from getting a quality education. Themes identified included:

1. lack of food, as some students went to school without breakfast or packed lunch;
2. attendance issues, as some students did not go to school regularly;
3. falling behind and not being able to catch up, as once students were behind the learning stage of the majority of students they tended to not be able to follow class lessons;
4. constraints on parents, such as shift work, long working hours, lack of English or formal education, which prevented parents of disadvantaged youth from helping their children with their studies and learning;
5. **overcrowded home environment8*, which made it difficult for students to complete homework.

Saturation

When qualitative researchers from a positivist methodological orientation (i.e. those who see social science as simply applying the methods of the natural sciences to the social world) analyze data they generally aim to reach a point of ‘saturation’.

Saturation means that in the later interviews or data collection, no new themes emerge. The implication of saturation is that there is no point in further data collection.

For example, in the hypothetical case above (education in a disadvantaged community), the researchers might argue that saturation had been reached after 25 interviews because in the last 3 interviews, the interviewees only raised issues which had already been identified in the previous 22 interviews.

Triangulation

Triangulation is an analytical method used in many different settings, including qualitative research.

It involves using multiple methods, or multiple data sources, to test the validity of an idea, concept, theory, or claim.

The concept of triangulation comes from navigation and mapping, where one often determines the position of an object by measuring it from at least two different positions.

For example, in the ‘Hairifying Truth’ study in Week 1, the students argued that sex with men tended to drive pubic hair grooming in women by pointing to five separate findings:

• sex with men was statistically significantly correlated with all types of gromming except bikini waxing;
• sex with women was not significantly correlate with grooming;
• qualitative interviews with lesbians indicated that there were no strong preferences or pressures to groom or not;
• qualitative interviews with gay men indicated that there were strong normative pressures to groom;
• qualitative data from hetrosexual men showed strong preferences against pubic hair in women.

In this case, the hypothesis appears to have strong support because it is supported by multiple independent data sources.

Ethnography

Ethnography is a detailed (sometimes called scientific) description of peoples and their cultures. It generally involves describing the culture from the view of the insiders in that culture.

An ethnography of a community is generally the result or end product of qualitative field work.

Participant-observation

Participant observation involves simply spending significant time immersed in a community as a participant and observer. It is a type of field work.

Naturalism

Naturalism is the philosophical belief that social scientists do the best research when they study societies in a natural setting, rather than an artificially created setting, such as a survey or a laboratory.

For example, when some qualitative researchers express a belief that you can only truly get to know a community by spending significant time emersed in it, they are expressing a commitment to (or belief in) naturalism.

Field site

Field site a place in which field research is done. Field sites are normally places where there is regular interaction between people who share a culture.

Examples of field sites include:

• a soup kitchen where injured migrant workers are given breakfast and dinner
• a Papua New Guinea village
• a block of rental flats
• a coffee shop
• a boarding house for children with behavioural disorders or adjustment issues.

Strategies for success

Good for attitudes and experiences

In particular, focus groups allow you to get a very wide range of opinions on a single topic is a short period of time.

Allows debate/conflict

You also get to see how different attitudes and beliefs interact in a conversation. So if you had a focus group with people of different backgrounds and politics, you could see how they debated and discussed a particular controversial issue.

Empowers marginalised groups

Focus groups can provide a space to empower marginalised groups. I have worked with researchers who have used focus groups as spaces to empower domestic workers, and sex workers. The experience of hearing other members of your marginalised community can be empowering. It can help participants see their problems as systemic, not individual. They can also see that they are not alone. They can even talk about solutions together.

Bad for detailed individual cases

One of the main problems with focus groups is that you have limited time, so they are bad for getting detailed information about individual cases. They are also not very useful when you want to cover a wide range of topics, as you quickly run out of time.

Avoid leading questions

Leading questions assume a particular answer or assumption. This biases responses towards particular answers.

For example, “Is your employer a bad man?” is a biased question. Better would be “Would you say our employer is a bad man or good man or what?”

Avoid yes/no questions – give options (“Did x happen first or did y happen first?”)

Particularly with marginalised communities, it is important to avoid simple yes no questions. The problem is that the default answer is always “yes”. People try to be agreeable.

One of the best ways I have found to avoid this is to instead give a series of options.

I had a migrant workers who I interviewed give a very detailed case history, and everytime I asked if something happened he would say “yes”. In the end I found the only way I could get a straight answer out of him was to say “Did you go to the public hospital or private hospital first?”

You should be flexible, not them. Follow their train of thought

You should be the one who is flexible, not them. You will have your script, your list of questions. The temptation will be to run through them like a list. This is bad interview technique.

What you should be doing is opening up the conversation with the first question or two, and then try to let the participant talk freely, and slowly ask questions to make sure you get the details you want for your study.

The main way to do this well is to know your questions very well. This way you can make sure you can raise questions in whatever order best fits the participants train of thought.

Linear stories are easier for participants – start at the beginning

Remember that linear stories are easier for people to tell. They have a natural flow. And even when people get distracted or wander off, the timeline provides a place you can come back to an orient the conversation.

A famous study by Professor Martina Morris used qualitative interviews to obtain the entire sexual histories of hundreds of people in Thailand, Uganda, and the United States. And apparently the way she got very honest answers was by starting by asking about their first romantic or sexual experience, and then simply going through them in sequence.

Expect digressions

Simply expect digressions. People will get off topic. Some people will find it impossible to stay on topic. You need to use your judgement and humanity to decide how much and how often to steer the conversation back on topic.

Double check timelines

People’s memories are terrible. Double check timelines (if they really matter). I have heard people get major events completely out of sequences - incorrect years and orders of events.

Be careful of assumptions

This is a hard one to monitor, but just make sure you check some of your assumptions. Many migrant workers, for example, will tell you they have good boss, and give big smile and be genuinely happy. But then if you ask about some common problems, in simple ways, you find problems. “What time do you start work? Finish? How many days per week? Ever work overnight? Does boss deduct saving money?” I had a worker I interviewed who insisted he was happy at work and treated well. He had not had a day off in months, and regularly worked in a construction site for 24 hours straight so as to help his boss meet deadlines. His employer also kept a proportion of his pay as savings, which the worker would only receive when he left Singapore. This is actually an indicitator of forced labor (because it is a penalty held over the worker that makes them unable to say no to the employer). But his first answer was that he was happy, he had good boss, etc.

Useful questions:

These are the simple questions I find most useful for in depth interviews.

Purposive

When we conduct qualitative research, we don’t just interview anyone. We try - like all reasearchers - to get an unbiased sample. We want to talk about a group of people, and use this to draw conclusions about a much large group or community they represent.

In qualitative research, we generally can’t do true random sampling.

Instead we tend to use what is called ‘purposive’ sampling, which simply means that you use your judgement to try to get a sample that contains signficant levels of variation and difference.

By choosing a sample that has significant variation and difference, it becomes harder for research to be accused of bias in sampling.

Events: Routine, special, unanticipated

One way qualitative researchers talk about appropriate sampling is to talk about sampling across different types of events.

Routine events are those which happen regularly. They are the normal parts of life. Sunday church. Doing the laundry. A dinner party with friends.

Special events are those which are more rare. A wedding. A funeral. An economic crisis.

Unanticipated events are those which we did not expect when designing our study. These can often be quite insightful, as they break our preconsceived ideas about a community or the world.

Descriptive

Descriptive questions: these are questions that are very open ended and general. They might ask the participation to describe their typical day, or a particular event, example, or experience.

For example: “How would you describe the type of fashion (clothing) that you wear to university? How about your friends? And other people at this university?”

Structural

Structural questions: these are questions which use conceptual categories - normally from the participant or the field site - and which asks interviewees to provide more detail.

For example “Some people say that you can classify students based on their approach to fashion. What do you think are the major types of students, based on their fashion?” or “You say there are a group of female students who dress in ‘business’ attire. Can you tell me more about the clothes they wear? And who are these women? Why do they wear busines clothes?”

Contrast

Contrast questions: These are questions that ask the interviewee to compare and contrast cultural/conceptual categories. “What are the main differences between smart casual and business fashion at this university?”

Types of field notes

Tips for note taking