Surveys and Sampling methods
- Surveys: Are a means of collecting primary data from large no. of people.
- Who uses the survey method: Produce quantative date- Favoured by Positivists= Townsend- "Poverty in the UK 1979" Used survery method to proudce a mass of statistical data, about the causes of poverty, Goverment- Carries out census, Market researchers, election pollesters
- Sample: Smaller representative group drawn from the survey population= Sociologists rarely have time /money to question everyone in a macro approach
- If it is a representative sample results can be generalised and applied to the whole survey population
-The representaviness of a sample can be affected by: Sample size- Too small would not be representative, The sampling frame (List of names of all included)- No individuals should be missing, The sampling method used.
- Random Sampling: Where every possible participent has an equal chance of being chosen
-Advantages: Cheap, quick & easy, everyone has an equal chance of being chosen,minimise the possibility of bias
-Disadvantages: May not be representative,If the population is very heterogeneous the results can vary considerably. May be difficult to do for large/ dispersed populations.
1. Stratified Random sampling: This involves dividing the target population into non-overlapping categories called strata. A sample is then drawn randomly from each sub population.
-Advantages:If the population is heterogeneous- gives a more realistic picture than simple random, and ensures representation of key strata.
-Disadvantages: More time consuming
2. Systematic Random sampling: Names are selected from the sampling frame at regular intervals until the size of sample is reached.
-Advantages:Less chance of mistakes Particularly useful with large populations.
-Disadvantages: Results may be biased, not everyone has a chance of being chosen
- RS CONTINUED - 3. Random Cluster sampling: Is a type of sampling in which the target population is first divided into smaller groups or clusters. Random samples are then drawn from the clusters.
-Advantages: Useful for a very large or dispersed target population, more representative
-Disadvantages: Important subgroups may be left out, more time conusming
- Non- Random sampling:
1. Quto sampling: Randomly the researchers go out looking for the right number of each sort of person required in each category
-Advantages:Quick and easy to do. Sampling frame not needed.
-Disadvantages: Unlikely to be representative of the target population.
2. Snow ball sampling: Involves collecting a sample by contacting a no. of key individuals who are asked to suggest others who might be interviewed & so on adding to sample until enough data been collected (Laurie Taylor)-Advantages: Allows access to hard to access group. - Disadvanatages: Lack of control, sample is self selecting bias
Stages of a survey
- Stages of a survey: Any research, always starts with some kind of hypothesis/ question to investigate and any survey needs to operationalize concepts
- Importance of a Pilot Survey: Before carrying out large-scale survey, it is important to carry out PL- Small scale pratice surve carried out before the final survey to check for any possible problems- Purpose? Identify/Iron out any problems, which researcher may have overlooked, avoid wasting time & money in FS, after PL is completed, results are reviewed, changes are made, main survey can be then proceed
- Problems of the social survey: Posistivits might call "Scientific accuracy"
1. Validity: Surveys need to be carefully planned,Interpretivists- Argue statistical data fail to describe acurately people's meaning & motives
2. Generalization: Concerened with representativeness- Process of extending a specific instance to a wider universal set
3. Reliability: Whatever survey finds, should be found by anyone else conducting the same survey again