Stats Flashcards
Sensitivity
If a person has the condition, how likely are they to test positive
contingency table
Sensitivity = TP / TP+FN
Specificity
If a person doesn’t have a condition, how likely are they to test negative
contingency table
specificity = TN / TN+FP
Positive predictive value
If someone tests positive, how likely are they to have the condition
contingency table
PPV = TP / TP+FP
Negative predictive value
If soemone tests negative, how likely are they not to have the condition
contingency table
NPV = TN/ TN+FN
When do you use fishers test?
small sample size
When do you use Bonferroni correction
when analysing data multiple times
when do you use chi squared
not normally distributed
can put it in a contingency table
when do you use t test
normally distributed
What is a cohort test
analytical test, observational
take an exposure and look for the outcome
What is a case-control test
analytic, observational
take an outcome and then look back to find an exposure
What is a cross-sectional test
observational, analytical
looking at outcome and exposure at the same time
What is a cross-sectional test
observational, analytical
looking at outcome and exposure at the same time
