Data Collection Flashcards
What is a population?
- A collection of the whole of something
- All elements sharing some set of characteristics;
- If we have a population we can get parameters - true values
- Census
- A study of all elements of a population
What is a sample?
- A set of individuals drawn from a population
- If we have a sample we can get statistics - values that estimate the parameter
- Sample characteristics = statistics
- Needs to be representative of full population
What kinds of error are there?
- Sampling error
* Non-sampling errors
What is sampling error?
- The difference that exists between a population and a sample as a result of the sample selection.
- Will be reduced by taking a larger sample
Why are non-sampling errors a problem and what kinds are there?
- Larger samples won’t reduce non sampling errors so more serious
- Coverage error
- Nonresponse error
- Measurement Error
What is coverage error and how is it reduced?
- Results from selection bias
- No matter what, some people will never be in sample
- Reduced by having an adequate sampling frame (up to date list of all items from which you will select the sample)
What is non-response error?
- Some people will never choose to do surveys
- Upper and lower economic classes have differing survey response frequency
- Can’t assume that people who do respond to surveys are the same as those who don’t
What is measurement error?
- Ambiguous wording of questions
- Hawthorne effect
- Respondent feels obliged to please the interviewer
- Respondent error
What are the steps to taking a sample?
- Define the sample frame
- Choose nonprobabliity or probability sampling
- Chose a sampling plan
What is a sample frame?
- A list of all the items that make up the population
- Choice of frame important
- Sample is then chosen from the sample frame
What is the difference between probability and non-probabilty sampling?
- Non probability sampling
- Choose the items to be included without knowing their
probabilities of selection - Cannot apply stasitical inferences
- Cheap and dirty
- Choose the items to be included without knowing their
- Probability sampling
- Select items based on known probabilities
- Statistical inferences allowed
What are the non-probability sampling plans?
- Convenience
- Judgemental
- Quota
- Snowball
What are the probability sampling plans?
- Simple random
- Systematic
- Stratified
- Cluster
What is cluster sampling?
- Divides the population into clusters, a number of clusters is selected to represent the population and all units within selected clusters are included in the sample
- Cheap, easy, reduced field work
- Imprecise, difficult to compute and interpret results
- Properties, households, homes
What is stratified sampling?
- The population is divided into groups called strata from which a simple random sample is drawn
- Objective is to form strata such that the population values of interest within each stratum are as much alike as possible
- Includes all important sub-populations, precision
- Difficult to select relevant stratification variables, not feasible to stratify on many variables
What is systematic sampling?
- The population is arranged and there is an interval between each selection (production line)
- Can increase representativeness, sampling frame not necessary
- Can decrease representativeness
What is simple random sampling?
- Each item in a population has an equal chance of inclusion
- Simple & easy for small populations
- Difficult to construct sampling frame, expensive, lower precision
What is snowball sampling?
- Existing study subjects recruit additional ones
- Can estimate rare characteristics
- Time consuming
What is quota sampling?
- Population is segmented into subgroups, then judgement is used to select units from each segment based on a specific proportion
- Samples can be controlled for certain characteristics
- Selection bias, no assurance of representativeness
What is judgemental sampling?
- Researcher chooses samples based on their view of who would be helpful for the study
- Low cost & time
- Does not allow generalisation, subjective
What is convenience sampling?
- Members of the population are selected by the ease of access to them
- Least cost & time
- Selection bias, sample not representative