EXAM 3 Flashcards
Experimental variables
Manipulate an independent variable
Must have at least two levels
Confounding variables
type of extraneous variable that covaries with your IV
Provide alternative explanation for findings
Biggest threat to internal validity in an experiment
Extraneous variables
individual differences in physical and mental states, environmental concerns
Related to DV
As extraneous changes, DV changes
Not the thing you are studying
Research designs
Only have measured– correlational design
could lead to quasi
Only have manipulated– experimental design
Can be a within subjects design (when people receive every condition)
Could also be a between subjects design– randomized and matched
Randomized
utilizes random assignment
Relies on chance
More people you have, more beneficial it is to use
Matched
guarantees that groups will be equivalent on the characteristics you use for match
Placing people into groups that score similar in extraneous variable
If you are worried about an extraneous variable– choose matched or within subjects design
Correlational research
predicting that there will be a relationships
Specify the relationship you are expecting
Quasi experimental design
you can not infer causality
Experimental research
IV is under control and DV is measured
Hypothesis is a prediction for how the IV affects the DV
Factorial designs
more than one independent variable
Each condition is a combination of levels of the IV
Internal validity
quality of your study
Can you trust the findings
Traditionally defined as how close you can get to causality
Are there alternative explanations for your findings?
External validity
can you generalize the findings you get to other people and other situations
Threatened by
Factors that make the research setting unlike the real world
Problems with the sample of your study– not representative of the population
Probability sampling
probability of selecting person in the population is known
Simple random sampling
everybody is in a pot and you grab them at random
Stratified random sampling
researchers divide subjects into subgroups called strata based on characteristics that they share (e.g., race, gender, education level)
each subgroup is randomly sampled using another probability sampling method.