Research Design

Research Design

1. Social surveys give quangative data

   • 1) Social surveys collect information about large target populations, using questionnaires or interviews.
     • 2) Social surveys tend to be used by postiits as a primary source of quantative data. This data can be anaylysed to discover overall patterns and trends.
       • 3) They're reliable, sothey're used by governemnt agencises and research companies.

2. Before you can start social survey you need a sample

   • 1)It's too expsnsive and time consuming for sociologists to survey the whole targetpopulation. They selct a sample.
     • 2) If the chartersistics of the sample reflect the charteristics of the rtaget population, with similar propertions of people in terms of age, class ehtnicity and gender, then the sample can be said to be representative of that traget population is known as its representativenss
       • 3) If the sample is sufficently large and represetitive, then it should be possible t make generalisations from it about the wider traget population. The extent which you can accurately do this is the sample's generalisibility.

3. Representaitve Sampling

   • Representaitce sampling involves picking names out of a 'sampling frame'. A sampling frame is a complete list of population being sampled, which needs to be accurate, complete and without any duplicate enteires.
     • 1) In simple random sampling, names are taken completley at random, e.g radnomly slecelted froma list by a person or a computer, so each memeber of the population has an equal chance of being selected.
       • 2) Systematic sampling involves choosing a radnom starting point in the sampling frame and selcting evey nth value, e.g. evey fifth name. There may be bias if thee's a underlying patterb un the sampling frame.
         • 3) Multi stage sampling means selecting a sample from within anothers sample. If often used to select samples for opinion pools to meassure voting intention. First , a selction of consititunecies is choosen to represent the whole country, then postcodes within that consitutncy are selected, then houses from those postcodes.
           • 4) In startfified random sampling the population is put into segments called 'starta' based on things like age, gender or income, e.g age 18-24, 25-34, 35-44. Names are selected at random from within each segment.
             • 5) Quota sampling is a bit like startififed radnom smapling, but it's not radnom. The selction is made by the interviewwer, whoi'll have a quota to meet, e.g, "interview 20 women between 25 and 34". Interviews tend to pick people who look 'nice', which intriduces bias. It's quick and useful, though.

4. Non representative sampling

   • Some target populatuons may be diffuclt to access, e.g criminals, the very young or very old, or the charcteristics of the population may be unknwon. In these cases you can use non-representative sampling methods.
     • 1) Snowball samping - finding the intiital contacts and getting them to give you more names for your research
       • 2) Purposive sampling - is when researchers select non-representative samples, iften in order to falsify a hypothesi. E.g femist sociologists trying to disprove the ide that gender roles are determined by biological difference delibertaely looked for samples where women's roles weren't diffeent from men's roles, or weren't traditoanlly 'feminine'.
         • 3) Oppertiunity smapling - is used when resarchers need to select a non-representative sample quickly and easily. Researchers can use captive audainces, these are gorups of people who are gathered tigther for another reaso, like a gorup of school children or office workers. Reserachers also go to publuc areas and slecte people who are neraby.