Sampling Flashcards
Population vs. Sample
A population is the total group of interest
A sample is a subgroup of the group of interest
What is a target population?
the group to whom researchers wish to apply their research findings.
What is an accessible population?
the pool of potential research participants who are actually available for a given study.
*A smaller subset of the target population
Inclusion vs. Exclusion Criteria for subject identification
Inclusion criteria define in more detail the characteristics that individuals from the population of interest must possess to be eligible for study (demographic, clinical, and/or geographic).
Exclusion criteria define in more detail characteristics that will make individuals ineligible for consideration as subjects (may also be demographic, clinical, and/or geographic).
What are the two general approaches to subject selection?
1) Probabilistic Sampling
2) Non-probabilistic Sampling
What is Probabilistic Sampling?
refers to the use of a method for randomly selecting subjects for participation in a study
*Generally unbiased, however it is more expensive and time consuming
What are the 4 types of probabilistic selection methods?
1) Simple Random Sampling
2) Systematic Sampling
3) Stratified Sampling
4) Cluster Sampling
Describe Simple Random Sampling
The most basic probabilistic sampling method in which each potential subject has an equal chance of being selected
For Example: numbers drawn out of a hat
Does Simple Random Sampling increase or minimize sampling error and why?
It minimizes sampling error because each potential subject has an equal chance of being selected
When is simple random sample selection useful?
When selecting from several hundred or fewer subjects
Describe Systematic sampling
This type of subject selection method is used when the subject candidate pool is in the thousands. Subjects are selected in specified intervals such as all evens, odds, certain pattern, etc.
In systematic sampling what subject is selected at random?
Only the first one, then all remaining subjects are chosen based on their numerical distance
Describe Stratified Sampling
A more complex probabilistic selection method used when investigators have an interest in capturing homogenous subgroups within the population
For Example: selecting 10 freshman, 15 sophomores, 15 juniors, and 10 seniors
Describe cluster sampling (aka multistage sampling)
This type of probabilistic selection method first randomly selects a large subgroup (cluster), and then selects smaller units from these clusters at random For example: first randomly select 10 schools, then randomly select 1 class from each school, and then randomly select 4 students from each class
What is Non-probabilistic Sampling?
refers to methods for choosing subjects that do not involve a random selection process.
*Easier to implement and cost less
What are the 3 types of non-probabilistic selection methods?
1) Convenience Sampling
2) Snowball Sampling
3) Purposive Sampling
Describe convenience sampling
The most basic non-probabilistic method in which researchers select subjects who are readily available (such as students).
Describe snowball sampling
A non-probabilistic method in which researchers start with a few subjects and then recruit more via word of mouth from the original participants.
Describe Purposive Sampling
A non-probabilistic method in which researchers make specific choices about who will serve as subjects in their study
**This method is very biased
What is effect size?
The magnitude of the difference between two mean values (change in status)
How is effect size calculated?
is the mean change score divided
by the standard deviation of the pre-treatment score
What is power?
The probability that a statistical test will detect a relationship between two or more variables.
In other words the ability of a statistical test to find a significant difference that really does exist.
How do you calculate power?
1 - β
* β is type II error which is accepting a false null hypothesis
How large must a sample size be?
How large is a very good sample size?
Need at least 30 subjects
100 is very good
Describe a single case design
A type of quasi-experimental study in which one subject receives in an alternating fashion both the control condition (A) followed by the experimental condition (B)
What are the 3 variations of the single case design we discussed in class?
1) A-B-A
2) A-B-A-B
3) A-B-C
Describe A-B-A design
There is a baseline data period (A) followed by the experimental treatment period (B) after which they return to the baseline condition (A).
When is the A-B-A design useful?
When experimenters want to observe behavior before treatment, during treatment and after treatment.
Describe A-B-A-B Design
There is a baseline data period (A) followed by experimental treatment measurement (B) the treatment is withdrawn (A), and then the treatment is re-introduction (B).
When is the A-B-A-B design used?
If you are designing an experiment in which you want to see an improvement
Describe A-B-C Design
There is a baseline data period (A) followed by experimental treatment measurement (B) which is then followed by another experimental treatment
What is DF?
Degrees of Freedom
Statistical concept indicating the number of values within a distribution that are free to vary.
How do you calculate degrees of freedom?
n-1