Unit 3 - Collecting Data Flashcards
Observational Study
A researcher records what is already happening. Subjects choose what they do.
Experiment
The researcher randomly assigns subjects to treatments. Subjects do not get to choose.
Blind Experiment
The subjects do not know which treatment they are receiving
Double Blind Experiment
The subjects AND the evaluators do not know who received which treatment.
Placebo
A fake treatment (sugar pill, saline shot)
Placebo Effect
Patients show improvement because they think they received treatment even though they only got a fake treatment
Factors
What you are testing in the experiment. Same as the explanatory variable (x)
Response Variable
What you measure at the end of the experiment. The y variable.
Levels
The amounts or types of factors.
Examples:
Factor: Sleep; Levels: 5 hrs, 7 hrs, 9 hrs
Factor: Drug; Levels: Drug A, Drug B
Block
A group of subjects for an experiment that are alike. Blocks must be split up to different treatments.
Matched Pairs
A block of size 2. 2 subjects who are alike and are randomly assigned to different treatments.
A subject can be a pair for himself/herself.
A strata
A group of people who are alike based on a characteristic. Used for sampling.
Stratified Sampling
Each strata gets its own hat. Randomly select people from every strata to survey.
A cluster
A mixed group of people that should be representative of the population
Cluster Sampling
Put the names of the clusters in the hat. Randomly select one or more clusters and survey some/all the people in those clusters only.
Systematic Sampling
Begin with the nth person. Sample every ___th person.
Simple Random Sample
Every person in the population goes in the hat. Randomly select from the hat. Every person has an equal chance of being chosen. Every group of people has an equal chance of being chosen.
Convenience Sample
Survey the people who are around you; who are convenient and easy to survey. A bad sampling method.
Response Bias
People give answers that don’t represent their true feelings
- wording of the question is confusing
- question is about something embarassing
- intimidated by person asking the question
- wording of the question makes people feel like they should answer a certain way
Voluntary Response Bias
People self select to answer the survey. Usually only people with strong opinions respond.
Non-Response Bias
The researcher randomly selects people to participate in the survey, but those people do not respond
Undercoverage
The survey is done in a way that leaves out a certain segment of the population
Three Types of Experimental Design
- Completely Randomized Design
- Randomized Block Design
- Matched Pairs
Three Principles of Experimental Design
Control
Randomization
Replication
What is control?
Treat all subjects the same except for the treatment to limit the effects of lurking variables.
What is randomization?
Randomly assign subjects to the treatments. This helps reduce the effects of any lurking variables that cannot be controlled.
What is replication?
Assigning multiple subjects to each treatment.
Experimental Units
Non-human subjects.
- animals
- plants
- objects
- bacteria