These surveys aim to find out the opinions of a wide range of people. But they have one drawback - everyone is asked their opinion at one time, that s, of course, when the survey takes place. The problem is that people change their minds and so what they think today they may possibly not think tomorrow. So much surveys are fixed reflection of one moment in time.
Longitudinal surveys
These are surveys conducted on the same people over quite a long period of time, sometimes over as long as 20 years. They provide us whith clear, moving image of the changes in attitudes over time.
Unfortunately longitudinal surveys are difficult to do, mainly because people drop out of the survey as they get bored answering the questions, or they might move and the researches lose track of them. If too many "drop out", this may make the survey unreliable.
Panel surveys
An alternative version of the longitudinal survey is the panel survey. A panel survey usually lasts for a shorter time than a longitudinal survey and asks questions of panel members on a more frequent basis. Typically, a person may be a member of a panel for a couple of years and be asked questions every month. In a longitudinal survey, they will probably be asked questions only once a year.
Before any survey is actually conducted, the researchers will want to conduct a "dry-run" just to make sure they have got everything right. This dry-run is known as aPILOT SERVICE. There may be problems with the way the questions have been written so that people cannot understand them, or perhaps there are too many questions so that people get bored. They may also discover that the people they have chosen for the survey may not be appropiate for their purpose.
There are about 64 million people in Britian. Finding out everybody's views would be terribly expensive and complicated to do. Only the government has the resources to carry out a survey of every household in Britain (known as the Census), and even this is so difficult and expensive that the government only does it once even ten years.
So, in order to find opinions and get information, sociologists have to study a smaller number of people.
This is not necessarily a bad thing, as long a the small group chosen (the sample) is a true cross-section of people, so that the opinions of this sample represent the opinions of everyone in the population under study. Clearly, if the sample is not a true cross-section, the whole study will be inaccurate.
Getiting an acurate sample of the population you wish to study is therefore one of the most important things that a researcher must do. This is what we need to discuss next. But before we do so, there is one problem we have to mention, that is the sampling frame.
