Methods of Collecting Primary Data

When data must be collected first-hand, the investigator chooses one of several standard methods, balancing accuracy against cost, coverage and the nature of the informants.

1. Direct personal investigation (direct interview). The investigator personally contacts and questions each informant face to face. It gives highly accurate, original and reliable data and allows doubts to be cleared on the spot, but it is costly, slow and suited only to a small field; it also risks the investigator's personal bias. Best where the field is small and high accuracy is essential.
2. Indirect oral investigation. Instead of the informants themselves, third parties who possess the information (witnesses / persons connected with the matter) are questioned. Used when the informants are unwilling, unavailable, or the matter is sensitive (e.g. enquiries into drinking or gambling habits). Care is needed because the third parties' answers may be biased or based on hearsay; several witnesses are usually questioned to cross-check.
3. Information through local correspondents (agents). Local agents or correspondents in each area collect and send in the data in their own way. It is cheap and gives wide, continuous coverage (used by newspapers and government departments for regular reports), but it is less accurate, since it depends on each correspondent's judgement and honesty. Suited where regularity and wide coverage matter more than pinpoint accuracy.
4. Mailed questionnaire method. A questionnaire (a printed list of questions) is mailed to the informants, who fill it in themselves and return it. It is economical and covers a wide, scattered area, but it works only with literate informants, suffers from non-response (many are never returned) and cannot clear doubts. The wording of the questionnaire is critical. …