The Questionnaire and the Schedule

In one line

A questionnaire is filled in by the respondent; a schedule is filled in by the investigator who puts the same questions to them.

In the wording a student can write in an exam: a questionnaire is a list of questions, printed or written, sent or handed to respondents who read and answer them themselves and return them; a schedule is a similar list of questions which is carried by an investigator who puts the questions to the respondent orally and records the answers in the respondent's presence. The distinction is not in the questions but in who administers the instrument, and every other difference between the two follows from that.

The questionnaire

Its forms of question

Closed or restricted questions, offering fixed alternatives: yes or no, a list to select from, a scale. They are quick to answer, easy to code and count, and they force the respondent into categories the designer chose.

Open or unrestricted questions, inviting the respondent's own words. They reach the unanticipated and are laborious to analyse and often left blank.

Mixed or cafeteria questions, offering alternatives with an "other, please specify" line, which is the usual practical compromise.

Contingency questions, answered only by those who gave a particular answer earlier.

Scale questions, measuring intensity rather than presence.

The rules for constructing one

These are examinable and they are also practical.

1. Keep it short, because length is the chief cause of non-response.
2. Simple language, in the respondent's own vocabulary, avoiding technical and legal terms.
3. One idea per question. Avoid the double-barrelled question, such as "are the police prompt and courteous", which cannot be answered by anyone who thinks they are one and not the other.
4. No leading questions, which suggest their own answer.
5. No ambiguity. Words such as often, usually and regularly mean different things to different people, so ask for a number.
6. Avoid questions requiring calculation or long memory.
7. Order: easy and interesting questions first, sensitive questions late, and personal particulars usually at the end.
8. Logical sequence, grouping related questions.
9. Provide for every possibility, including "do not know" and "not applicable", since forcing an answer manufactures data.
10. Pre-test it, which is the step that finds the defects the other nine did not prevent.
11. State the purpose, the sponsor and the assurance of confidentiality at the head.

Its merits

Cheap; can reach a very large number and a wide geographical area; requires no field staff; free of interviewer bias; gives the respondent time to consider and to consult records; and anonymity, which is its most important advantage on sensitive subjects, since a respondent will write what they would not say to a stranger's face.