PSY 250 Lecture Notes - Lecture 9: Scientific Method, Randomness, F-Test
Document Summary
Inferential statistics: error, null hypothesis, t test, f test. External validity: exact replication, conceptual replication, meta-analysis. Inferential statistics: we can tell main effects and interactions, these suggest that ivs affect the dv. One main effect per factor or variable. Fail: the results that we obtained could be wrong, sampling. Finding truly representative samples is extremely difficult. Scientists make mistakes in measuring: validity. Scientists make mistakes in experimental design: scientists combine all forms of mistakes. Error: a measure of any variability in all of the variables in an experiment that are not controlled, randomized, or manipulated, cannot know absolute truths. What we know is relative to other things. Scientists cannot and will not prove absolute truth: the dv is affected by so many things other than the iv. Effects of error: the effects of manipulations must be stronger than the effects of error. Null hypothesis: scientists always disconfirm, never confirm. Therefore, we use error to our advantage.