Читать книгу Sociology - Anthony Giddens - Страница 91

Three types of questionnaire are used in surveys. Some contain a standardized, or fixed-choice, set of questions, to which only a given range of responses is possible – for instance, ‘Yes/No/Don’t know’ or ‘Very likely/Likely/Unlikely/Very unlikely’. Standardized questionnaires have the advantage that responses are easy to count and compare, since only a small number of categories is involved. On the other hand, because they do not allow for subtleties of opinion or verbal expression, the information they yield is likely to be restricted in scope and can sometimes be misleading.

Other questionnaires are open-ended, giving respondents more opportunity to express their views in their own words rather than limiting them to fixed-choice responses. Open-ended questionnaires typically provide more detailed information than standardized ones. On the other hand, the lack of standardization means that responses are likely to be more difficult to compare statistically, and this limits attempts to draw general conclusions from the study.

Questionnaire items are normally listed so that a team of interviewers can ask the questions and record responses in the same predetermined order, and all the items must be readily understandable to interviewers and interviewees alike. In the large national surveys undertaken regularly by government agencies and private research organizations, interviews are carried out more or less simultaneously across the whole country. Those who conduct the interviews and those who analyse the results could not do their work effectively if they constantly had to check with each other about ambiguities in the questions or answers.

Questionnaires should also take into consideration the characteristics of respondents. Will they see the point the researcher has in mind in asking a particular question? Have they enough information to provide an answer? Will they be able to answer at all? The terms and concepts used in a questionnaire might be unfamiliar to the respondents. For instance, the question ‘What is your marital status?’ might baffle some people, and it would be more appropriate to ask, ‘Are you single, married, separated, or divorced?’ Most surveys are preceded by pilot studies in which just a few people complete a questionnaire in order to pick up such ambiguities and iron out problems that may not be anticipated by the investigator before the main survey is carried out.