We cannot tell by looking at individual families that there is such a severe problem. The relative proportion of boys and girls in any individual family could always be compensated by a different proportion in other families we have not looked at. It is only by using methods like a census or large scale survey that the overall ratio for the community as a whole can be calculated and the problem can be identified. Can you think of other social issues that can only be studied by surveys or censuses?
a picture based on a collectivity rather than on single individuals taken separately. Many social problems and issues become visible only at this aggregative level — they cannot be identified at the more micro levels of investigation. However, like all research methods, survey also has its disadvantages. Although it offers the possibility of wide coverage, this is at the cost of depth of coverage.
It is usually not possible to get in-depth information from respondents as part of a large survey. Because of the large number of respondents, the time spent on each must be limited. Moreover, since the survey questionnaire is being taken around to respondents by a relatively large number of investigators, it becomes difficult to ensure that complicated questions or those requiring detailed prompting will be asked of all respondents in exactly the same way. Differences in the way questions are asked or answers recorded could introduce errors into the survey.
That is why the questionnaire for a survey (sometimes called a ‘survey instrument’) has to be designed very carefully — since it will be handled by persons other than the researchers themselves, there is little chance of corrections or modifications in the course of its use. Given that there is no long-term relationship between investigator and respondent and hence no familiarity or trust, questions that can be asked in a survey have to be of the kind that can be asked and answered between strangers. Questions of a personal or sensitive kind cannot be asked, or if asked are likely to be answered ‘safely’ rather