chapter. 6 Flashcards
why President Alf Landon poll was wrong
Sampling frame
• Sampled telephone subscribers and automobile owners which are a wealthy sample
2 types of sampling methods
Random sample (probability theory)
o 1948
o Is the primary method for selecting large samples
o Many research situations make SRS impossible
Nonprobability sampling
- : technique in which samples are selected in some fashion not suggested by probability theory
Sample vs Sampling
sample
- a group of people to represent the population selected to be observed
sampling
- the creating of the group of people
elements
- set of observations
ex: people, families, organizations
4 reasons for sampling
- adequate representation
- Cost, time, and accessibility
- Manageability
- Accuracy of information
EPSEM
Equal probability of selection method): samples that conforms to the principles of probability sampling are labeled as EPSEM.
probability sampling types
Simple random sampling (SRS
Systemic Sampling:
Stratified sampling:
Multi stage cluster sampling
Systemic Sampling:
Systemic Sampling: A type of sampling in which every Kth element in a list is selected for inclusion in the sample.
sampling interval vs sampling ratio
Sampling interval
Standard distance between elements selected in the sample
sampling interval=(population size)/(Sample size)= 10,000/ 1,000
Sampling ratio
Proportion of elements in the population that are selected
sampling ratio=(sample size)/(population size)= 1,000/ 10,0000
Stratified sampling:
Stratified sampling: a type of probability sampling that involves grouping subjects into homogeneous strata before sampling.
Stratified sampling procedure is used to ensure representation of different subgroups in the population,
It is used in conjunction with simple random sampling and systematic sampling.
Multi stage cluster sampling:
Multi stage cluster sampling: A sampling procedure in which natural subgroups (clusters) are sampled initially, with the members of each selected group being sub-sampled afterward. We use this sampling process when there is no single list or when the list is too long to use SRS or SM. MSCS involves a cycle of listing and sampling.
Non-probability sampling
Purposive judgemental sampling
Snowball sampling:
Quota sampling:
Reliance on Available
Subjects (Convenience) Sampling
Purposive/ judgemental sampling
a sampling process where the researcher uses his won knowledge or judgment about which unit will be useful while selecting a sample. It is a useful method when the purpose is to obtain a general description of the target population
Snowball sampling:
Snowball sampling: a sampling procedure where each people interviewed are asked to suggest additional people for interviewing. (when it is difficult to identify group members).
Quota sampling:
Quota sampling: where units are selected into the sample on the basis of pre-specified characteristics. It is used to ensure that the sample contains the same distribution of characteristics that exists in the population being studied.
Reliance on Available Subjects (Convenience) Sampling
a sampling strategy where a researchers select conveniently available subjects to observe
errors with samples
Random error
Mistakes that are equally likely to occur
• Ex: crime study ask people if they were a victim of crime they may say yes even if not
• In overall random errors cancel out in their effects
Bias • Form of systematic error • Where pattern of mitake are more likely than others • No not cancel themselves out • Can sirioudly distort results o All are prone to error
PPS probability proportionate to size sampling
multi cluster sample in which clusters are selected not with equal probabilities but with probabilities proportionate to their size
- Ex: city block with 200 households has twice the chance of selection as one with 100 households
Multistage designs and sampling errors
- Very efficient but less accurate sample
- Sampling errors
o Sample shows population only within a range of sampling error
o Sample of elements within the cluster represent the elements in that cluster only
o Ex: selecting a wealthy block with wealthy people in the houses - Should max # of clusters selected while decreasing the number elements within each cluster