Research Design

Research Design

1. preference and knowledge - sociologists often specialise in a field they find interesting; steve bruce specialises in religion, sociologist that want to make a change often try to research things that could help develop solutions to the problem.

2. Funding and Co-operation - there are many sources of funding: charity (the jowseph rowntree foundation), industry, the government (lots of quantitative studies). the funder is often called the gatekeeper as they have the final say. government agencies often fund stuff related to policy, industrial providers fund things which will benefit them. the researcher needs to get the co-operation of the groups they're studying cos if not the topic may not be viable.

3. career - many sociologists want to improve their employability, original or popular topics with good clear results improve an academics standing (particularly if published by a big journal). a quick way for a sociologist to progress their career is to falsify or add something to another sociologist's work.

   • means that one particular group could be researched alot; office workers are routinely researched for stratification, some say they are working class but other say they're middle class.

4. Reviewing - reviewing existing data and literature of a topic the sociologist wishes to study is crucial; if they analyse existing material it can give useful info on methodology and whether it worked well, what sample sizes to use etc...

5. Research questions - these questions give focus to the sociologist, it is a question their research aims to answer; a good question should focus on one part of the topic and be clear and easy to answer.

   • should be value free and without bias, shouldn't suggest certain social changes. shouldn't be a leading question.

6. Hypotheses - a statement that makes a prediction, research will aim to either prove or disprove this; can also state a relationship between 2 things.

7. operationalisation - some concepts are hard to measure i.e. democracy, so they need to be operationalised; this can be done using an indicator e.g. electoral participation. sociologists need to be able to justify how they operationalised their concepts (this is subjective and usually criticised).

8. Triangulation - when sociologists try to combine different methods or data to get the best results.

   • gives a more detailed picture, more valid.
   • allows different sets of data to be compared against eachother.
   • combines strengths and weaknesses of different types of data.
   • can be expensive and time consuming so is not always possible.

9. Pilot Study - lets you test accuracy of questions and any technical problems in research design; makes studies more reliable and valid, can test time and help to train interviewers. show project is feasible and help secure research funding.

10. Social Surveys - collect info about large populations using questionnaires and interviews; tend to be used by positivists and can be analysed to show trends and patterns; reliable so used by govt and research companies.

11. Sample - can't research the whole population so they select a sample; if the characteristics of the sample reflect that of the population in the correct proportions then this can be said to be representative, if the sample is sufficiently large and representative then it can be said to be generalisable.

    • Representative Sampling - this is done by picking names out of a sampling frame (complete list of the population being sampled, needs to be accurate , complete and without duplicates).
      • Random - randomly selected, equal chance of being chosen.
      • Systematic - selecting every nth person, may be biased if pattern in sampling frame.
      • Multi-stage - taking a sample from within another sample (usually done to measure voting intention) e.g. pick a town and then a postcode within that town.
      • Stratified Random - put population into strata (based on things like age, income, gender) then randomly sample.
      • Quota - selection made by interviewer and will have to meet a quota of whatever group they are sampling e.g 20 beautiful girls. (interviewer tends to pick people who look nice which introduces bias, quick and easy).
    • Non-representative sampling - some populations may be difficult to access e.g. criminals, in this case you use a non-representative sampling method.
      • Snowball sampling - finding initial contacts and they give you names
      • Purposive sampling - when a sample is chosen to falsify a hypothesis; feminists may choose a sample to disprove gender roles so choose manly women.
      • Opportunity sampling - when research needs to be done quickly this can be done; stopping people on the streets or a captive audience e.g. office workers.