Sampling
A sample obtained by allowing subjects to decide whether or not to respond
What is a Voluntary Response Sample?
Because they are generally fearful of government intrusion, many immigrants from Latin America did not return their census questionnaire during the 1990 census.
An example of Undercoverage Bias
Deliberately imposes some treatment on individuals to measure their response.
What is an Experiment?
The most important element of any experiment and it must be incorporated either in the selection process of experimental units and/or the distribution of experimental units into treatments and control groups.
When using the Randomization element
What/Who treatment is imposed on
Experimental Units
What type of sampling sses subjects that are readily and easily available?
What is a Convenience Sampling?
A radio talk show host asks listeners to call in with their opinions of making wearing masks in public space mandatory.
An example of Voluntary Response Bias
Observes individuals and measures variables of interest but does not attempt to influence the responses
What is an Observational Study?
We want to make sure we are using a design that compares two or more treatments
When using the Comparison element
When results of an experiment are unlikely (less than 5%) to happen purely by chance and gives us convincing evidence that treatment caused the difference.
Statistically Significant
A sample that consists of individuals from the population chosen in a way that every set of individuals has an equal chance in the sample selected
Simple Random Sample (SRS)?
Our administration sends out 100 survey questions to a sample of parents in order to gage their attitudes towards returning to school in 2020. Only 23 responds.
An example of Non-Response Bias
A medical team examines the records of 5 large hospitals and compares the survival times of those cancer patients who had surgery versus those who had chemotherapy.
Observational Study
Treated identically in all respects to the group receiving the treatment except that the members of control group do not receive the treatment.
When using the Control element
Separate subjects into blocks and then randomly assign treatments within each block
Randomized Block Design
Divide the population into groups of similar individuals (strata) then select on SRS within each strata. Combine the SRS from each strata to form your full sample.
What is a Stratified Random Sampling?
People who are more thrill-seeking are likely to take part in pain research studies. This may skew the data
In gym class, the effect of exercise on blood pressure is studied by requiring that half of the students walk a mile each day while the other students run a mile each day,
Experiment
Use enough experimental units in each group so that any difference in the effects of the treatments can be distinguished from chance differences between groups.
When using the Replication element
Subjects are paired (block size of 2) and then randomly assigned to a treatment
Match Pairs Design
Divide the population into sections (clusters) then randomly choose a few of those clusters. Every member of the cluster becomes your sample.
What is a Cluster Sample?
In scientific journals, there is strong publication bias towards positive results. Successful research outcomes are published far more often than null findings.
An example of Survivorship Bias
The relationship between weights of bears and their lengths is studied by measuring bears that have been anesthetized
Observational Study
When subjects and/or experimenters don't know about treatments
Blinding
_______ ________ allows us to conclude a treatment causes changes in the response variables.
Random Assignment