Lecture 19 - Bringing it all together Flashcards
What can you not do to a theory?
Prove it - you either support it or disprove it. Falsification!
Data can either support a theory, disprove it, or…?
Distinguish between two different theories.
What is ‘exploratory research’?
Research where a theory is not available in advance.
When is the most fruitful research often conducted?
When there are two competing theories - supporters of each try to disprove the other’s theory.
What is the ideal environment in which to run an experiment?
When two theories can be compared by what they predict for the results - the data are hopefully interesting whichever way they go.
What could cause negative data?
- inaccurate measurements - insufficient power - difference could be minute and could therefore require more pts to decrease variance - confounding/noise variables - there might not be a difference.
Why can’t you look for significant similarities in data?
There’s no test for it, mainly because of the problem of how similar things would have to be - the measurement accuracy would be too critical.
How is a statistical hypothesis formed?
Grand theory –> research hypothesis –> statistical hypothesis.
What are 4 things you must do when describing data?
- show it in as much detail as is reasonable - always show variability - if it’s confusing, simplify it - include any data worth discussing (not in appendices!)
What three things do you need to know before carrying out a statistical test?
- Which test to run 2. What the test’s requirements are 3. How to perform the test in SPSS