Flashcards in Factor Analysis Deck (40)
How can we allow for sufficient variability of items and participants?
- inversely keyed items and homogenous items
- minimisation of serial effects
- examine invariable responses closely
- item difficulty or clarity
- ceiling/floor effects
What is the problem with constant measurements?
- they cannot really be statistically analysed.
- data analysis cannot always account for fix design errors
What is a frame error?
- errors in the sampling (Q analysis)- we sampled ppl that don't belong
What is domain error?
- errors in the domain we sampled
What are some errors in samples?
- nonresponse/participant errors
What are some errors with observed/true scores?
- measurement error
What is generalizability theory?
- it adds systematic error in the observed scores and attempts to map it and eliminate it (ie. control it)
What is item response theory?
- mathematically maps the difficulty of measurement items and maps them against participants' ability on a construct
- can be used to make predictions
What are the uses of profiling?
- criminal personality profiling: eliminating suspects, identify unknown offenders, used with unusal crimes, adaptive interrogation techniques