Methods of inquiry Flashcards
(16 cards)
Whats the Difference between Stratified Sampling and Cluster Sampling
h
Discuss the Survey errors
- mis-specification population –
chase wrong population/target the wrong people - poor choice of sampling frame –
obtaining sample from methods which arent effeicnent. EG phone books which no one uses anymore - sampling error –
biased coverage and a poor choice of sampling scheme. EG choosing top 10 companies that all begin with A
4.
Discuss the Survey errors
- mis-specification population –
chase wrong population/target the wrong people - poor choice of sampling frame –
obtaining sample from methods which arent effeicnent. EG phone books which no one uses anymore - sampling error –
biased coverage and a poor choice of sampling scheme. EG choosing top 10 companies that all begin with A - NRE –
a lot of people dont respond
5.
Discuss the Survey errors
- mis-specification population –
chase wrong population/target the wrong people - poor choice of sampling frame –
obtaining sample from methods which arent effeicnent. EG phone books which no one uses anymore - sampling error –
biased coverage and a poor choice of sampling scheme. EG choosing top 10 companies that all begin with A - NRE –
a lot of people dont respond - MRE –
EG: how much you earn p/m – not many people know answer to this - PE –
innaporiate measures
average scews results - IE–
misinterpreting the results
What stages do you have to go through to conduct a Sample Survey
- Define the Objectives, Accuracy needed, Costs involved, etc.
- define the population
- set up sampling frame
- Decide on Sampling Method, Estimate Sample Size and Implement (Pilot Study? Why do a Pilot Study?)
- Calculate measures and assess their accuracies
- Review results and procedures (Quality checks!)
Discuss the process of induction
infer a rule based on evidence
process of updating rules if there is more eveidence
Discuss the process of abduction
plausable hyposthesis - EG companies run out of cash. ONE REASON FOR THIS IS RUNNING OUT OF CASH (PLAUSABLE HYPOTHESIS)
What is the difference between observations and experiments?
observation - DONT control allocation of subjects for specific group
experiment - control as many aspects in situation as possible before observing the variables under a study
What is population in this context and give an example of this
A defined group within which you want make your survey
People working in the service sector
(What do we mean by ‘working’, ‘service sector’?)
Pet cats in Wales
What is Sampling units in this context and give an example of this
Elementary items of interest within the population
Individual pet cats
Household groups of cats
What is sampling frame in this context and give an example of this
an exhaustive unduplicated list of the population
All households in Cardiff County listed from the electoral roll
What are some examples of probability sampling?
units have non-zero chance of being selected
Simple Random Sampling, Systematic Sampling, Cluster Sampling, Stratified Sampling, etc
What is purposive sampling?
probabilities not known or even if they are non-zero
Convenience Sampling, Snowball Sampling, Quota Sampling, etc
What gives stratified random sampling an edge over simple random sampling?
This allocation of sampling effort to ‘where it is needed’ is what gives stratified sampling
We use prior information about the population to make the sampling more efficient!
Characteristics of Stratified Random Sampling
Strata’ (singular ‘stratum’) are mutually exclusive and collectively exhaustive subpopulations with defined characteristics
If we have heterogeneity between strata and homogeneity within them, it is more efficient to sample each stratum separately and combine those results
Characteristics of cluster Sampling
Heterogeneity within Clusters –
Homogeneity between Clusters
