Chapter 14: Biostatistics Flashcards
Gaussian distribution
normal distributoin
in normal distribution what percent of values fall within 1 SD of the mean, what about 2SD
- 68% within 1 SD
- 95% within 2 SD
what does it mean when data is skewed right
- positive skew
- more high outliers
alpha in biostats
error margin, typically set to 5%
CI = 1 - alpha
- an alpha of 5 means a confidence interval of 95
null hypothesis
there is no statistically significant difference between the groups
type I error
- false positives
- null hypothesis rejected in error
- probability of a type 1 error is the alpha
type II error
- false negative
- null hypothesis accepted when it should have been rejected
- beta, typically set to 0.1 (10%) or 0.2
- risk of type II error increases with smaller sample size
- = Power - 1
power
- the probability that a test wiill correctly reject the nell
- = 1- beta
relative risk or risk ratio implications
- if =1, no difference bteween groups
- > 1, greater risk of the outcome in treatment group
- < 1, lower risk of outcome in treatmetn group
relative risk reduction (definitio)
how much risk is reduced in the treatment group compared to control group
absolute risk reduction includes reduction in risk AND incidence rate
how to round NNT and NNH
- NNT: round up (even if 0.01), avoids overstating benefit
- NNH: round down (even of 0.99) avoids understating harm
odds ratio vs hazard ratio
- odds: used to estimate the risk of unfavorable events
- hazard: survival analysis (looks at death or disease progression), rate at which unfavorable event occurs within a short period of time
interpreted similarly
t - test
used when the endpoint has continuous data and the data is normally distributed
- one sided: data from group compared to known information on general population
- paired: when a patient is their own control
- student t-test: when there are tw independent variables
ANOVA
- analysis of variance; F-test
- used to test for statistical sifnificance when using conitinuous data with 3 or more groups
