Experimental Methods 7.3 Flashcards
What to experiments test?
Experiments test causal relationships
•Manipulate an independent variable (IV) and measure a dependent
variable (DV)
•Highly controlled situations and equivalent groups
Variable
Characteristics that vary
- Variability based on the participants response
Between group designs
- Ppts are exposed to only one level of the IV
* Helps reduce demand characteristics
Subject variable between group design •Nonequivalent groups
•Limits causal inference
•Best practice is to try and match ppts on relevant variables
Within group designs
- Participants are exposed to every level of the IV
- Ppts compared to themselves
- Stronger statistical power
- Can increase demand characteristics and “good participant” confounds
- Equivalent groups
- Counterbalance the order of the conditions
Factorial designs
- Factorial designs: more than 1 factor (IV or subject variable)
- Fully within, fully between, mixed factorial designs
- Useful when predicting interactions
True experiments
- Test causal relationships between an IV and a DV
- Internally valid
- It is the IV and not something else that led to the differences in the DV
Quasi-experiments
•Causal conclusions about the effect of an IV or subject variable on a
DV cannot be made with a lot of confidence
•Incomplete control over the subject variable (or IV)
Longitudinal designs
•Within group measurements are repeatedly taken over an extended
period of time•Special case where a subject variable can differ within a ppt
Challenges of longitudinal studies
- Take many years
* Subject attrition
Cross sectional designs
ex: •Measure people of different ages at the same time
Cohort effcects
A cohort effect is a research result that occurs because of the characteristics of the cohort being studied. A cohort is any group that shares common historical or social experiences, like their year of birth.
