Methodological and conceptual extension of initial fieldwork practice questions.

Observation at a weekly market has established which sellers hold the sheltered pitches by the wall, but not how those pitches are allocated. Which extension is best justified by that particular gap?

• A questionnaire handed to shoppers asking which stall they visit most often and why they prefer it.
• Interviews with the sellers themselves about how a pitch is obtained and how it is kept.
• A further four weeks of observation from the same vantage point, recorded in the same way as before.
• A search of published trade statistics for the region, in order to place the market in its wider economy.

Answer: Interviews with the sellers themselves about how a pitch is obtained and how it is kept.

A method is chosen for the kind of data it yields, not for variety: the unanswered question is about sellers' own accounts of an allocation process, which only they can supply. Repeating the observation would produce more of the data that has already failed to answer it.

A student's first observation was made at one branch of a bakery. For the second round she is deciding between working at a second branch and staying where she is to interview the people she has already observed. Which consideration should settle it?

• Which of the two would yield the larger quantity of material in the time she has.
• Whether the staff at the second branch would be easier to approach than those at the first.
• Whether her question is about how the arrangement works here or whether it holds elsewhere.
• Whether a second site would make the findings representative of the bakeries in the town as a whole.

Answer: Whether her question is about how the arrangement works here or whether it holds elsewhere.

A second site earns its cost only when the question is comparative; if the question is why this branch works as it does, the people already observed are the source that can answer it. Two thinly documented sites are worth less than one case documented well enough for someone else to compare against.

A student's first round at a school canteen established that pupils sit in the same groups every day and that two tables are avoided. For her second round she proposes a questionnaire to be distributed to two hundred pupils asking how they choose where to sit. The main objection to this plan is that

• questionnaires may not be used at all in fieldwork for this course
• two hundred responses would take longer to analyse than the time available allows
• the question asks about a practice people follow without deciding, so short written answers will not reach it
• the pupils would be unlikely to return the forms in sufficient numbers for the results to be reported, and those who did return them would not be a representative group

Answer: the question asks about a practice people follow without deciding, so short written answers will not reach it

Where a pattern is habitual, asking people to explain it directly yields tidy after-the-fact reasons. A few unhurried conversations with pupils who sit at different tables, set against what she has watched, will produce far more than a large number of one-line answers.

Observation at a village hall has shown that some groups get the main room on Saturday mornings and others are given the annexe. The student now wants to find out how that allocation is decided. The most economical extension is to

• observe for a further four Saturdays to see whether the pattern of allocation continues
• interview the person who keeps the booking diary and read the diary itself alongside the interview
• ask every group that uses the hall to complete a form about their satisfaction with the room they receive
• compare the hall with a similar hall in a neighbouring village during the same period

Answer: interview the person who keeps the booking diary and read the diary itself alongside the interview

The question is about a procedure, and procedures have a keeper and usually a document. Watching longer would confirm the pattern without explaining it, and the two sources together allow the stated rule and the recorded practice to be set against each other.

For her second round at a community centre a student plans to interview the four people who were friendliest to her during the observation. The difficulty this creates, and its remedy, are best described as

• the four may give long answers, so she should limit each interview to a fixed number of minutes
• she may already know what they will say, so she should ask them questions she has not prepared
• friendly participants are unreliable witnesses, so she should interview people who were hostile instead, since they have less reason to present the place favourably
• the four are likely to share a position in the setting, so she should also seek out people placed differently

Answer: the four are likely to share a position in the setting, so she should also seek out people placed differently

Those who approach a visitor are rarely a cross-section: they tend to be the confident, the long-standing and those with something to say about the place. The remedy is not to exclude them but to add people the first round showed to be marginal, and to say in the write-up how each interviewee was reached.

A student framed her first round at a housing estate around the concept of community. Her material keeps showing that some residents are excluded from the meetings and events she is recording. Extending the project conceptually would mean

• keeping the concept and gathering more examples of community activity to strengthen the account, since the concept had been settled before the fieldwork began
• abandoning concepts and reporting only what she directly observed on each visit
• bringing in a concept that fits what the data show, such as boundary or belonging, and reworking the question
• changing to a different estate where the community is more clearly in evidence

Answer: bringing in a concept that fits what the data show, such as boundary or belonging, and reworking the question

A conceptual extension is a response to the material: the data have raised a question the original concept cannot hold, and the honest move is to change the concept rather than the site. Note that the observed exclusion is a finding, not an obstacle to be designed around.