Surveys and Sampling Methods Flashcards
(43 cards)
Population
The entire group of individuals that we want information about
Sample
The part of the population that we actually examine in order to gather information.
Sampling
Involves studying a part of the population to gain information about the whole.
Census
Attempts to contact every individual in a certain population.
Bias
Systematically favoring certain outcomes (want to avoid)
Sampling Methods: Voluntary Response
Flawed method because people with strong opinions are most likely to respond.
Sampling Methods: Convenience
Flawed by choosing individuals who are easiest to reach.
Sampling Methods: Simple Random Sample (SRS)
Each member is equally likely to be selected and each possible group (sample) is equally likely. One population/ One sample
Systematic Sample
Begin at a random point; select every nth member.
Sampling Methods: Stratified Sample
Divide the population into similar groups (strata) and then pick a random sample within each strata.
Sampling Method: Cluster Sample
Divide population into groups. Randomly pick a group and then use each individual in that group
Observational Study
Good: Ethical, easier to conduct (people are already doing the behavior)
Bad: Can’t conclude anything
Experiment
Good: Can use control to conclude causality between the explanatory and response
Bad: Can be unethical, harder and time consuming to set it up
Placebo Effect
When an untreated subject incorrectly believes that they are receiving real treatment and reports an improvement
Control Group
Group of patients who receive a fake treatment; enables control over outside variables on the outcome
Principles of Experimental Design: Replication
Use enough subjects/trials to reduce chance variation. Natural variability occurs.
Principles of Experimental Design: Randomization
Use of chance to assign experimental units to treatments
Statistically Significant
The finding is unlikely under the assumption of what should happen. Rarely happen by chance.
Principles of Experimental Design
Control, Replication, and Randomization
Block Design
Another form of control. The same idea for an experiment as stratified is to selecting a sample.
Match Paris Design
Subjects are matched in pairs to compare responses to two treatments. Pairs could be 2 subjects with similar characteristics that would influence response or 1 subject with both treatments done, with their own control.
Single Blind Experiment
Subjects are unaware of what treatment they receive
Double Blind Experiment
Neither the Subjects nor those who measure the response know which treatment the subject received. Contributed to controlling the Placebo Effect; ensures no unconscious bias on the doctors part
Confounding Response
When variables can’t be distinguished from each other.