L6 - Finding and fixing flaws in research Flashcards
What is treatment variance?
Desired variance due to experimental manipulation
What are extraneous variables?
unrelated variables that may affect the DV
What is error variance?
random error that affects all conditions equally
What is confound variance?
systematic error that affects all comditions differently
What is a perfect experiment?
total variance = treatment variance
What is a real experiment?
Total variance = treatment variance + confound variance + error variance
What is statistical validity?
correct use of statistical procedures
What’s internal validity?
Whether IV manipulation causes changes in the DV
What’s construct validity?
Measuring what’s intended to be measured
what’s external validity?
generalisability of results
How do you ensure internal validity (within subjects design/ matched pairs)?
ensure similar treatment for all groups
use random assignment and control conditions
How do you ensure internal validity (between subjects design/ IDG)?
Use counterbalancing and control of ppt variables
randomly allocate ppts to minimise order effects
How do you ensure construct validity?
Link constructs clearly with manipulation/ measure
Use objective measures
How do you reduce expectancy effects?
Use standardised procedures
Implement blind/double-blind procedures
Control for experimenter and participant expectations
How do you ensure external validity?
Use diverse samples and replicate studies across different settings.
Conduct meta-analyses to synthesise results.
