Contingency Analysis Flashcards
describe the response and explanatory variables in contingency analysis
both are CATEGORICAL variables
what does a contingency analysis test
tests and estimates associations between categorical variables
what does a contingency analysis test between two categorical variables
whether they are independent of each other
does one variable in a continency test tell us about the other
NO - are independent
what is a 2 x 2 table
a table where there are two categorical variables with TWO levels EACH
what is the probability of success variable
p
what is the probability of failure
1 - p
what is the association between the probability of success and failure
they are mutually exclusive
how to calculate odds of success for a sample data set
what is an odds ratio (OR)
a ratio used to quantity the difference between the odds in different groups
how to calculate the odds ratio of a sample data set
what is the OR ratio when the odds of both groups are the same
1
what is the OR when event in first group is HIGHER than the second group
O1 is greater than O2
what is the OR when event in first group is LOWER than the second group
O1 < )2
what is a shortcut for calculating the OR ratio for the sample
how to calculate the standard error of the sample odds ratio
what is the confidence interval for the ln(OR)
what are two tests done under contingency analysis
- chai squared contingency test
- fisher’s exact test
what does the chai squared contingency test do
tests the goodness of fit to the data of the null model of INDEPENDENCE of the variables
what type of table does a chai squared contingency test usually apply to
2 x 2 x 3 tables
what type of table does a fisher’s exact test usually apply to
2 x2 tables
how to find expected frequencies for chai squared contingency test
how to calculate test stat for chai squared contingency test
degrees of freedom for chai squared contingency test
r = total row
c = total column
what does the chai squared contingency test test
the null hypothesis of no association between variables (independent) between 2 OR MORE categorical variables
does the fisher’s exact test give exact P value
YES
