MATH 140 Lecture Notes - Lecture 23: Null Hypothesis
Document Summary
Categorical variables: qualitative variables, not quantitate/numerical: ex: gender, street name. Two different types of relationships between variables: Independent variables: there is no connection between the variables whatsoever. Ex: ice cream sales and crime rates: dependent/associated variables: there is a direct relationship between the variables. You use a frequency table to find out what the relationship between the variables are. Grand total: use the expected count: the value of each cell if the variables were independent, formula for expected count: =(cid:4666)(cid:1867)(cid:1875) (cid:1872)(cid:1867)(cid:1872)(cid:1864)(cid:4667)(cid:1876)(cid:4666)(cid:1855)(cid:1867)(cid:1864)(cid:1873)(cid:1865)(cid:1866) (cid:1872)(cid:1867)(cid:1872)(cid:1864)(cid:4667) (cid:4666)(cid:1866)(cid:1856) (cid:1872)(cid:1867)(cid:1872)(cid:1864)(cid:4667) (cid:1876)2= (cid:4666) (cid:4667)2 x2: (chi-square. The chi-square test: used to find it the two variables are independent using both the observed count and the expected count, uses every cell. O: (observed count- actual value in the cell) How to compute 2 with your ti-83/84 calculator: Adjust the number of column and rows- shown in calculator as (column, row) and add in only the observed values.