AGRC1023 Lecture Notes - Lecture 10: Type I And Type Ii Errors, F-Test, Analysis Of Variance
Document Summary
Statistical test in anova is the f test. 2-factor experiment may get more than one f-test. Not all population means are equal, but which ones are different?". Used to determine whether difference between two means is significant at prescribed alpha level. Does not take into account type i error rate. Useful for small amount of pre-planned comparison of means. Looking at data that isn"t linear correlation data must be linear! We need to know the assumptions before looking, know there are no biases. Ensure looking at all data, not just part of it, know where it"s coming from. Affected by sample size large sample can make vague correlation that may be significant, but not that meaningful just from large sample. Can be outlier for several reasons think about why you may have outliers. Impacts on data e. g. watch out for curvilinear correlation is not linear in this case o. If you change the units, it will change the correlation coefficient.
