Module 9: Chi-squared test Flashcards
One way expected contingency tables use
-whether there are differences among counts of that categorical variable
One way expected null and alternative
-Ho: no difference in allergy type among children
-Ha: there is a difference
Two way expected contingency tables use
-whether the counts across levels of one variable change differently depending on what level of the other variable is being looked at
Two way expected null and alternative
-Ho: type of allergy and home heating source are independent
-Ha: these are not independent
How to calculate expected tables
-one way: add all up and divide by number of cells
-two way: find marginal distributions for rows and columns and then find the total and multiply by distribution of both row and column
Chi-squared use
-measures distance between observed and expected contingency tables
Chi squared null distribution shape
-bell curve shifted to left side because it can only have positive values
Chi squared degrees of freedom
-one way table: k-1
-two way table: (c-1)(r-1)
-k is for number of cells
-c is for number of columns
-r is for number of rows
Chi squared reporting
-symbol for test (x2)
-degrees of freedom
-total count of observed table
-observed chi squared value (2 decimal places)
-p value (5 decimal places)