Flashcards in Session 5 Deck (67):
What is a 'population'?
the large group of ALL possible scores that would be obtained IF the behavior of every individual of interest in a particular situation could be measured
What is a 'sample'?
A relatively small subset of a population that is selected to represent or stand in for the population; a subset of the complete group of scores found in any particular situation
The __________________ is everybody you are interested in learning about.
What is a census?
gives an accurate demographic of a population
A _____________ is subset of population and stand for and represents the population of interest.
What is the link between sample and population?
What is inference?
researchers infer that the characteristics of the sample probably are the characteristics of the population
Inferences must have ______________ ____________, setting limits on the possibility that we are wrong. 1% or 5%
What are the two questions researchers ask when evaluating a sample?
1. Is the size adequate?
2. Is the sample biased?
What makes a sample unbiased?
random samples in which every member of the population of interest have an equal and independent chance of being included int he sample
What is another name for random sample?
simple random sample
What is the biggest task of sampling?
defining the population as specifically as possible
What is a bias free sample called?
Simple random sample
Can a bias free sample contain error?
yes, but it is still bias free
What is potentially biased but easier way to approach sampling?
Describe two ways of volunteerism recruitment?
1. researchers issue a call for volunteers
2. researchers identify a random sample from a population, but only certain people choose to participate
Why is volunteerism a source of bias?
because those who volunteer may be fundamentally different from non-volunteers
What are the two types of errors that random samples are subject to?
1. Sampling Error
2. Systematic Error
What is sampling error?
the difference, due to random choice, between a sample statistic and the population parameter it represents
What is Systematic error?
the difference, due to non chance errors, between a sample statistic and the population parameter it represents
Is a sample EVER perfect?
1. there will never be a perfect representation of population.
2. there will always be sampling error
Sampling error is created by_______ ___________.
Sampling Error is reduced by ___________________.
adequate sample size
Sampling error is evaluated with ________________.
Systematic Error is created by ______________.
bias in sampling
Systematic error is reduced by _______________.
using sampling methods that are less subject to bias
Systematic error is evaluated with ________.
review of sampling methods used for possible sources of bias
Why is systematic error worse than sampling error?
random chance errors can be cancelled out by the introduction of another subject however systematic is throughout and full of bias
What is systematic sampling?
every nth participant is selected from a list of the members of the population
What is the source of bias in systematic sampling?
list of the population may be ordered in such a way that every nth participant is fundamentally different from every non-nth participant
What is precision in sampling?
the magnitude of sample errors
Precision is increased by _______.
reducing sampling errors through
List two ways to reduce sampling errors.
1. increase sample size
2. stratified random sampling
What is the most efficient sampling method?
SRS: stratified random sampling
What is stratified random sampling?
the most efficient method of sampling
List the 2 steps of stratified random sampling.
1. identification of important subgroups (strata) in the population
2. proportionate random selection of participants from each subgroup
What is the source of bias in stratified random sampling?
stratification not based on a variable that is relevant to the issue being studied
In stratified random sampling, the sample preserves the same ___________ as the entire population.
Stratified random sampling uses a number of ____________ __________.
What approach improves precision in stratified random sampling?
using multiple strata (multiple variables for stratification) in selecting a given sample.
Use of multiple strata improves precision as long as (2) ________, __________.
1. variables chosen are relevant
2. variables chosen are indecent (unrelated) to each other
What is independence?
one thing not affecting the other
What is the primary purpose of Stratification?
1. to ensure that different subgroups are represented in the correct PROPORTIONS.
2. NOT to make comparison across subgroups
3. But to obtain a single sample that is representative in terms of the stratification variable
What is cluster Sampling?
certain groups are randomly selected and ALL participants in each group are observed
What is the source of bias in Cluster Sampling?
1. clusters not drawn at random
2. each cluster tends to be more homogeneous in a variety of ways than the population as a whole.
What is Purposive Sampling?
select individuals whom researcher believes will be good source of information
What is the source of bias in purposive sampling?
Qualitative research primarily uses what type of sampling?
In Purposive sampling the ________________ has selected this population and therefore introduced their own bias.
What is Snowball Sampling?
researcher contacts potential participants who have been identified by previously tested participants
What is demographics?
the background characteristics of the participants in research
List 3 things that demographics are used for
1. provide a picture of the types of individuals who constitute the sample
2. Analyzing and interpreting the results of the study
3. allow readers to make informed judgments regarding the extent to which the results may apply to their own setting
What is Mortality Bias?
participants who drop out of the experiment at some point may be different fro those who remain or the population in general
What 3 things should be reported in mortality bias?
1. if mortality occurred
2. number and percentage who dropped out from each group
3. demographics of those who dropped out and those who completed the experiment
Precision is increased by what two factors?
1. increasing sample size
2. stratified random sampling
Precision is obtained by what 2 factors?
1. striving for an unbiased sample
2. seeking a reasonably large number of participants
List 3 reasons why TOO large a sample is not good.
1. principle of diminishing returns
2. with larger samples, the possibility of researcher error increases
3 with larger samples the possibility of supporting the research hypothesis when it is not true increases
List 5 Major factors that influence researchers' decision on how large sample size should be.
1. pilot study
2. procedures expensive/time-consuming
3. documenting incidence of something rare
4. variability of population of interest
5. looking for small differences
What are 3 details about pilot studies?
1. conducted to obtain preliminary info on how new treatments and instruments work
2. on the basis of the results, treatments and instruments are modified
3. smaller (10-100)
If procedures are expensive and/or time-consuming what size sample would the researcher use?
If documenting an incidence of something rare what size sample would be used?
If the population is homogeneous, what sample size is used?
If the population is heterogeneous, what sample size is used?
If the researcher is looking for small differences, what sample size is used?
Why does a heterogeneous group require larger sample size?
in order to capture all the variety within the population
What two factors determine whether to use a small or large sample size?
1. it depends on the size of the population of interest
2. it depends on traditional sample sizes for this type of research