PSYC 201 Lecture Notes - Lecture 18: Type I And Type Ii Errors, Confounding, Internal Validity
Document Summary
Psyc 201 lecture 18: counterbalancing and factorial designs part i. Logic of experimental theory is ideally seeing the difference between the conditions. Allow a lot of random conditions, even induce it, for the experiment to run randomly. Extraneous variables: increase extent to which conditions likely to differ by chance alone and so make it harder to find small differences. You can eliminate ev"s by holding them constant: e. g. only testing people during specific times and dates, reducing the amount of variability. Holding a lot of variables constant creates limited generalizability creating external validity. Systematic (confounding) effects: something that is correlated from some sources within. Confounds reduce internal validity (ability to make inferences of what causes what) Subject selection effect: occurs when there is a systematic way of who goes into a certain group; incorrect inferences can occur. Ideal: complete random chance of who goes into a group, can use probability to estimate how applicable it is to real life.