The different kinds of social survey methods are detailed below.
Postal Surveys
A potential participant receives a survey in their mail box which they may or may not fill out and return. Postal surveys are always structured, although they may include both qualitative and quantitative questions.
Some key points about postal surveys are:
- Participants self-select which means there is little control over the type of person which responds to the survey. This means the data may come from a narrow portion of the community. For example, the researcher may wish to obtain views from a wide cross section of ages, but find that a certain age bracket is poorly represented because they chose not to fill out the survey. The researcher can specify what sort of person they wish to respond which may help resolve this issue.
- A considerable effort is required to prepare and send out the survey because the questions must be very clear as participants cannot ask for clarification. Research into the issues being surveyed is required to ensure the questions are relevant, appropriate, are presented in participant-friendly language and cover all the potential points of interest. In addition, it is not always a simple task to obtain names and addresses of potential participants.
- The data collected may lack detail because of the closed nature of the questions or small areas to write responses. Moreover, there is often little or no opportunity for the researcher to clarify responses with the participant or ask for further detail on points of interest.
- Postal surveys do not allow for rapport to be developed between a participant and an interviewer simply because the interviewer is not present. By establishing a rapport with the participant, the interviewer tends to get a greater level of honesty and detail in the information collected. As a result a lack of rapport may affect the depth and the quality of the information.
- Postal surveys have the advantage that, because the questions and questionnaire format are standardised, data entry and comparison between answers are straightforward. If the questions are quantitative then statistical analysis is straight forward.
- A low response to postal surveys is quite common – often between 20 and 30% is normal. However, the response rate may increase if the issue is highly topical or a reward or incentive is offered for completed survey.
Telephone Surveys or Interviews
- Potential participants are contacted by phone and asked if they wish to participate in a survey. On agreement the participant may be asked qualitative or quantitative questions which are open ended or follow a set of closed questions similar to a postal survey.
- One of the clear advantages to a telephone survey is the participant and interviewer can interact. This allows a) relationship to be developed The participant to ask about the research and clarify the questions; b) points of interest to be followed up during the interview and answers clarified where necessary; c) The interviewer can select participants to meet certain criteria and thus survey a wide cross section of the community.
Disadvantages of phone interviews include home numbers not always being easy to obtain and many people in rural areas still have dial-up internet which blocks the phone line. Telephone surveys are quite common and may not be well received by potential participants, particularly at dinner time (i.e., early to mid evening) when most of the calls will have to be made. Interviews may need to call a large number of people to fulfil the number of responses required.
Data can be collected quickly and depending on the format of the survey data entry and analysis may be straight forward.
Face to Face Interviews
A face to face survey can take several forms it can be structured, semi-structured or unstructured and questions may be qualitative or quantitative. With this type of survey there is interaction between the participant and the interviewee and thus the chance to develop a rapport.
The interviewer can follow up points of interest or ask for clarification on comments made by the participant. There is more freedom to do this if the interview is less structured.
With face to face surveys respondents can be selected so that a broad cross section of the community can be represented. However, the time commitment required from the participant may be longer than for another type of survey.
The interviews may be long and questions may be asked in a different way and in a different order each time, which makes data analysis more time consuming than for postal surveys. To overcome this, special software packages may be used to help with qualitative data analysis.
The data collected from interviews tends to have more detail than postal surveys and produce a more in-depth picture of the social context but, because of the time consuming nature of collection, fewer participants are involved.
Web or e-mail based surveys
This approach is becoming more popular because data can be directly entered into a database which saves time. However, it still has many of the downsides of postal surveys with respect to a low response rate, lack of interaction between the participant and surveyor. Moreover, it is not very effective in areas with poor internet access (i.e., many rural areas with dial up) or where potential participants do not have e-mail.
Enquiries about this research should be directed to Dr Paula Blackett [email protected]