PSYC 251 Lecture Notes - Lecture 3: Carl Gustav Hempel, Null Hypothesis
Document Summary
Hypothesis: people prefer my new spaghetti sauce. Null hypothesis: there is no difference in preference for my new sauce. Give a taste test to a sample of 1000 people. Or 65% people say my sauce is better. Variability will result in random deviations from pure chance (50%) How far away from 50% do we need to be before we can conclude the answer. In small samples, deviations are more likely to occur by chance. Our statistics are going to have to get more conservative as we have smaller samples. Bigger values mean effects much larger than you would expect by chance alone (alpha=. 05) Smaller values mean effects that are small, and so you cannot rule out that they are just due to chance (1-alpha = . 95) A hypothesis is a theory about the world. Because we can never prove a hypothesis is true, we setup a null hypothesis and disprove it.