KINE 2050 Lecture Notes - Lecture 17: Statistical Inference, Negative Number, Chi-Squared Distribution
7 Apr 2018
School
Department
Course
Professor
Document Summary
Chi squares can be used for inferential statistics, to make inferences and conclusions. It is possible to complete a chi square for one variable with 4 possible categories. Therefore there is no signi cant di erent between when students write their test and how they do: since you had 4 groups, its called a 4 celled chi square test. To x this problem, statisticians came up with a di erent formula called the yates. Correction: its the same as the chi square equation, but you just take the absolute of the observed - expected if its a negative number, and subtracting 0. 5 from the number you get. Typically the yates correction is used when your degrees of freedom is equal to 1, so when you have a simple two category test, or when your expected frequency for any category is less than 5. For this class, only use it if told, otherwise use normal chi square formula.
