collecting evidence through surveys Flashcards
what is a sample survey?
systematic collection of information from large study groups, usually by means o interviews or questionnaires administered to a sample of units in the population
- most prominent research design in social science
- used to explain circumstances events and relationships
- systematic method
what are characteristics of sample survey?
- generalisable
- usually cross sectional at one pint in time (snapshot)
- Research design is post test
- researcher does not directly observe the subjects or their behaviour
what are the stages in a sample survey
- conceptualisation of research problem
- analytical or descriptive
- sampling
- data collection
- data entry
- initial data analysis
- statistical tests of hypotheses / in depth description
- presentation of results
what is the importance of sampling?
- key step in social survey research
- survey is a subset of a population
- population savers are expensive / impractical
- can aid accuracy in larger samples
- smoke statistics used to estimate population parameters
steps in how you sample?
- define target population
- identify sampling frame
- select sampling procedure
- determine sample size
- elect sampling units
- collect data from sampling units
what is a sampling unit?
an individual
what is probability sampling?
- each individual in a population has a known non zero probability of being elected
possible calculate sampling error
more common that non probability
what is sample error?
arises from unrepresentitivness of the population sample
what are the types of probability sampling?
- simple random sample
- stratified
- cluster
what is simple random sample?
- like lottery or throwing a dice
- equal probbaailty of being chosen
what are the advantages of simple random samples?
- ease of use
- unbiased, mor representative
what tare the disadvantages of simple random sampling?
- sampling error if does not represent
- normally used when researcher knows little about population
what is stratified sampling?
- parent population dividd into exclusive and exhaustive strata anda random sample is selected from each strata
- strata his based on attributes or characteristics such as income
what are the advantages of stratified sample?
- captures key characteristics
- smaller error in estimation
what are the disadvantages of stratified sampling?
- time consuming
what is cluster sampling?
divide population into separate groups calle clusters
simple random sample of cluster is selected rom population
what is non probability sampling?
umbrella term to cover all sampling that does not satisfy conditions of probability sample
fairly common
what are types of non probability sampling?
- convienece
- quota
- self selection
- snowball
what is quota sampling?
looking for specific characteristics
subsets of the population
e.g. 100 males between 20 - 30
what are the sources of error in sample survey data?
- sampling error
- coding error
- sample design error
- non reposes error
- measurement error
how do you determine sample size?
no definitive guide to optimum sample size
where a poplaultion is less than 50 use cencus
samples of more than 30 are needed for statistical tests
increase samples sizes needed in certain circumatsnaces
why is sampling important?
determining accuracy of results
when should you use a questionnaire?
- where the study is descriptive or explanatory
(underlying approach is deductive) - seeking to prove a particular hypotheses or describe / compare characteristics
(researcher knows in advance) - researcher
