Observation in the field and the observation report practice questions.

A student's record of a barber's shop consists entirely of what was said: greetings, jokes, an argument about a football result. Her supervisor suggests she add a sketch of the room before the next visit. The most useful thing the sketch would contribute is

• evidence of where people place themselves, so that talk can be related to position in the room
• a decorative element that makes the finished report more pleasant to read
• proof that the student was present in the shop on the days she claims to have visited
• an accurate architectural record of the premises that could be checked against the building plans

Answer: evidence of where people place themselves, so that talk can be related to position in the room

Who sits in the chair nearest the door, who stands, who is served out of turn and who never crosses to the back of the shop are all findings, and none of them appears in a transcript. A sketch converts the room into data rather than background.

A student's notes on a school office contain the line: 'The queuing system was chaotic and unfair.' Which rewriting turns this into a record that could support such a conclusion later?

• 'The queuing arrangements were poorly designed and caused frustration among those waiting.'
• 'Nobody in the office seemed to care about the order in which people were being seen that morning.'
• 'Two students who arrived after 9.05 were seen before three who had been waiting since 8.50; the secretary called names from a list on her desk.'
• 'The system was disorganised, as anyone watching for a few minutes on that morning would have been able to see for themselves without any difficulty.'

Answer: 'Two students who arrived after 9.05 were seen before three who had been waiting since 8.50; the secretary called names from a list on her desk.'

The entry giving times, order of service and the secretary's list records what happened, and a reader could reach a different conclusion from it. The others state a judgement, which cannot be checked against anything and leaves nothing for the analysis to work on.

During a second visit to a community centre a regular leans over and asks a student what she keeps writing in her notebook. For the record she is building, the most useful response is to

• close the notebook and continue the observation from memory alone for the remainder of the visit
• explain briefly what she is doing, and note the exchange and its effect on the room as part of the record
• say that the notes concern something unrelated, so that the setting is disturbed as little as possible, which is what the setting can reasonably ask of a visitor
• read out to him everything that has been written about him and about the other people present so far

Answer: explain briefly what she is doing, and note the exchange and its effect on the room as part of the record

The question is itself data: it marks the moment at which the student's presence became a topic in the room, and what happens afterwards is evidence about how she is classified there. Concealment would make the notes unusable, and abandoning them loses the record.

Two students observe the same weekly market on the same morning. One returns with a record of transactions: what was sold, at what price, with how much haggling. The other returns with a record of greetings: who greeted whom first, who was ignored, who stopped to talk. Neither record contains an error. The methodological point this illustrates is that

• one of the two records must be closer to what actually happened at the market that morning
• observation is always guided by a question, so what is recorded is selected before anything is written
• students working at the same site should always be assigned different days to avoid this outcome
• the two records can simply be added together to produce a complete account of the morning's events, since neither observer missed anything that the other one recorded

Answer: observation is always guided by a question, so what is recorded is selected before anything is written

There is no record of everything, and the selection is made by the question, whether or not the observer has formulated it. The consequence for a report is that the question has to be stated, so that a reader knows what the record was built to catch and what it necessarily let through.

A student concludes her observation report by writing that four visits to one launderette prove that people in the town use such places as social meeting points. The precise objection to this sentence is that

• four visits to a single site cannot establish a claim about the town, though they can describe this site closely
• observation can never support any conclusion, since only interviews yield evidence about what people do
• the claim would be acceptable if the student had visited eight times rather than four
• the report should contain no conclusions at all, since conclusions belong to the second round of fieldwork, once the material has been gathered and set aside for later use

Answer: four visits to a single site cannot establish a claim about the town, though they can describe this site closely

The error is in the scope of the claim, not in the amount of work. A careful account of one site is a genuine finding, and the useful move is to state what it would take to know whether the pattern holds elsewhere — which is a second-round design, not more of the same.

A student has noticed that arguments break out at the counter of a village post office when the queue is long, and she wants her next visits to establish how often this happens and who steps in. Which recording plan fits that question?

• Sweep the room every ten minutes and note how many people are present and what each of them is doing
• Follow one member of the counter staff for the whole visit and record everything she does and everything said to her
• Record each dispute as it occurs, with the time, who was involved and how it ended
• Write a continuous account of the visit from arrival to departure, noting whatever seems worth reporting at the time

Answer: Record each dispute as it occurs, with the time, who was involved and how it ended

Event sampling records every occurrence of a defined event, which is what a question about frequency and intervention requires. A scan sample answers 'what is going on at this moment' and would miss anything happening between sweeps; a focal follow answers questions about one person; a running account leaves nothing that can be counted.