In day-to-day interaction with each other we move swiftly and easily from one subject to another. However, there are difficulties for the researcher, who must encourage the free flow of speech, yet keep hold of any questions that arise. For example, the following extract about a youngster's first burglary throws up a number of questions that might fruitfully be pursued:
I was wagging school, and we had about 15p. And we wanted these pies from the bakery, they were about 40p and 50p. So we didn't know what to do to get them or anything, but we saw a house and I said, 'Shall we have a go?' and he says, 'Yeh'. So I says, 'Go and knock at the door and see if anybody's in'. So he went to the door and nobody was in. So these three kids went round the back and I was keeping watch. And they smashed a window, I heard this window smash, and I was waiting about five minutes, but it seemed like an hour. So I went round and got into the house and started searching after that. Robbed £90 out of a glass. And I just started doing more after that.
During the interview the researcher must reflect upon and pursue at least another nine and possibly many more items, all of which are worthy of further discussion. These are: wagging school; the importance of the discrepancy between the cost of pies and the money the boys had; why it was that particular house they 'saw', the inconsistency of the numbers involved; the apparent passivity of his contribution; the time dimension (it seemed like an hour); what they were searching for; the distribution and use of the £90; and the subsequent burglaries.
The question of validity is more thorny. How can the researcher be sure that respondents are not distorting the truth, telling outright lies or exaggerating a point? This problem is accentuated when there is a limited amount of time allocated to fieldwork and the number of interviews is restricted. Moreover, distortions in participants' accounts have been a particular problem for this study, as the young prisoners' ability to distort the truth is one of the central theoretical propositions being explored. If it is accepted that young prisoners actively deceive, is it possible to rely upon evidence generated by interviews?
First, what qualitative interviews lose in terms of accuracy is usually balanced by the insights gained into the lifestyle and culture of the interviewee (Hammersley and Atkinson, 1983). Second, an interviewer well versed in the subject matter will not easily be deceived. The plausibility of the account can be checked during the interview, old ground can be re-turned, information constantly checked. Follow-up interviews can clarify problems. In addition to these strategies, a further subsample of inmates were re-interviewed in groups of five. The relationship between information released in group and individual interviews is most important. The groups act as a check upon the information volunteered by individuals and questions to a prisoner repeated in the presence of a group of inmates can reveal inconsistency.Gold (1966) used a similar technique when he checked the validity of subjects' accounts with their friends. The nature and quality of the contradictions are the key to this study as they tell us which aspects of behaviour can be exaggerated or lied about and the ability of inmates to relay confusing messages about their behaviour. Care was therefore taken to monitor but not to discourage distortion in participants' accounts.