Social Surveys - Questionnaires and Interviews (Part 1)

There are two ways of conducting a social survey:
Written Questionnaires which respondents are asked to complete and return by post or e-mail.
Interviews which are carried out either face to face or over the phone.

Several different decisions have to be made before carrying out the social survey:
1. Types of Question
Closed-Ended or Structured Questions - These are questions which have a limited number of set answers which, generally speaking, cannot be deviated from.

Answers from these questions are normally fed into a computer and then turned into quantitative data for quick and easy analysis. These types of questionnaires are easily repeated and therefore are reliable. They are usually done on a large scale making them representative of the target population.

Open-Ended questions - The person answering is free to give whatever answer they wish in their own words.

Because of the possible differences between each answer, there will be a wealth of qualitative data and these types of questionnaires tend to lead to greater levels of validity.

2. Formulating an Aim or Hypothesis
Aim: An aim is a statement that identifies what the sociologist intends to study and hopes to achieve by collecting data on a particular topic.
Hypothesis: A hypothesis is more specific than an aim. It is a statement of fact that the sociologist tries to prove either true or false. A hypothesis is usually written in the form of a statement i.e. "differences in educational achievement are the result of differences in family income". Creating a hypothesis can give direction and purpose to a piece of research.

3. Operationalising Concepts
This is the idea that you must have a working or "operational" definition of the concepts you are researching. Take the hypothesis above:

"differences in educational achievement are the result of differences in family income".

Sociologist must make clear what they mean by educational achievement i.e. Does this mean a certain number of GCSE passes, and by family income i.e. is this measured by money coming in from wages or do you include other forms of wealth?

4. Pilot Studies
Before carrying out a survey a sociologist will do a pilot study. This is a small-scale study to:

a) Check the feasibility of the research.
b) Improve the design of the research.

A pilot study is much more likely to be used with quantitative data i.e. structured questionnaires and interviews.

• Some advantages of pilot studies are that it avoids wasting time and money on an inadequately designed project. 
• Questions can be checked to make sure they make sense to the respondents. 
• It may help you to establish a rapport with the people you are studying Gavron - The Captive Wife. 
• It can be used to develop the research skills of those taking part in the interviewing. 
• A pilot study may determine whether the research goes ahead. Insurmountable practical problems may arise. 
• A pilot study may be used to convince a funding organisation of the usefulness of a project. 

5. Sampling
A sampling frame is a list of all the population we are interested in studying. The bigger and more accurate your sampling frame, the more representative your sample will b.

A sample is the group of people chosen from the sampling frame to take part in the research.

Random sampling is where the sample is selected purely by chance e.g. names may be drawn from a hat. The sample may not be representative of the sampling frame for example, there may be an imbalance of gender.

Quasi-random sampling is where every 10th or 100th name in a list is selected. As above however, the sample may not be truly representative as not everyone has an equal chance of being picked.

Stratified random sampling is when the sample is divided into groups according to relevant variables such as gender, ethnicity, ages etc. As an appropriate number are taken from each group this is the most representative sampling technique.

Quota sampling is where the researcher goes out to look for a set number or a quota of each group that they wish to sample for example 10 males and 10 females. This relies on the researcher to make judgements which may be easy when picking someone by gender but not so easy by other variables such as social class.

Snowball sampling is when a key individual is contacted who is asked to suggest others who may be contacted. This is a useful technique with groups who it may be difficult to get to take part such as criminals. This is not a representative technique but may be chosen by interpretivist sociologists who are more concerned with validity in their research than making generalisations.

Opportunity sampling involves choosing individuals that are easiest to access for example passers-by in the street or a class of pupils etc. This is very unlikely to be representative of the target research population.