PS296 Lecture Notes - Lecture 5: Null Hypothesis, Type I And Type Ii Errors, Statistical Significance
Document Summary
Reduces chance of a type i error. Reduce the chance of a type ii error. Might not have enough power to reject the null. Ie. give your participants a large enough iv that it makes a difference. If we decrease the type i error rate, we automatically increase the type ii error rate. If we increase the type i error rate, we automatically decrease the type ii error rate. Sometimes with a very large sample size, small differences between means are statistically significant. Must differentiate between statistical significance and meaningfulness. When you reject the null, you"re saying that your results support the research hypothesis or provide evidence in favor of the hypothesis. This does not meant that the results prove the research hypothesis, nor that the research hypothesis is true. When we deal with stats, we are always making decisions based on probability. Results of research studies are based on probabilities.