Methods Flashcards
What is the test-retest method of testing reliability?
Participants take test twice - if similar scores, evidence for reliability
What is the parallel forms method of testing reliability?
Two versions of the same test - if similar scores, evidence for reliability
What is stratified sampling?
Important population subgroups defined, members of subgroups are sampled randomly
What is cluster sampling?
Everyone within a randomly selected cluster of population is researched
What is quota sampling?
Define important subgroups, but use whoever is available from these
What kind of sample is needed to ensure high internal validity?
Restricted sample - similar performance on measures to reduce extraneous variability
What kind of sample is needed to ensure high external validity?
Representative sample - differs widely in performance due to a variety of factors affecting both measure reliability and validity
Re-arrange the following from controlled to uncontrolled: Correlational, Experimental, Quasi-experimental, Descriptive
Experimental, Quasi-experimental, Correlational, Descriptive
What are the three conditions to infer causation?
Covariance between X and Y Temporal precedence (X occurs before Y) Exclusion principle (X only possible cause)
How is temporal precedence ensured?
IV is manipulated to ensure temporal precedence
How is the exclusion principle met?
IV isolated to ensure it is only possible cause
What are the four methods of error control in research?
Eliminate
Keep Constant
Balance
Randomise
What is statistical power?
Assuming the effect is genuine, how likely is it that the experiment will detect it?
How is statistical power increased?
By prominent effects, large samples, and robust methods
What must one minimise and maximise to design a powerful study?
Minimise error variance and maximise the differences in scores between conditions