Biostats Flashcards
(49 cards)
study type that can show risk factor association without establishing causality?
cross sectional study
odds ratio is used in what type of study?
case control
triple blind refers to
the additional blinding of the researchers analyzing the data
which phase of clinical trial assess efficacy, optimal dosing, and adverse effects?
phase II
what is equivalent to 1 - false negative rate?
sensitivity
what is another name for sensitivity?
true positive rate
explain how PPV and NPV are affected by pretest probability?
high pretest probability leads to
- high PPV
- low NPV
pretest probability is equivalent to
prevelance
equation for odd ratio?
ad/bc
equation for relative risk?
a/(a+b) / c/(c+d)
equation for the attributable risk?
incidence in exposed - incidence in non exposed = a/(a+b) - c/(c+d)
equation for the relative risk reduction (RRR) ?
RRR = 1 - RR
equation for absolute risk reduction (ARR) ?
c/(c+d) - a/(a+b)
equation for number needed to treat (NNT)?
NNT = 1/ARR
equation for number needed to harm (NNH)?
NNH = 1/AR
3 terms that are equivalent to “precision?”
consistency, reproducibility, reliability
2 terms that are equivalent to “accuracy?”
trueness, validity
how does increasing precision affect standard deviation and statistical power (1-B)?
- dec standard deviation
2. inc statistical power (1-B)
what is Berkson bias?
example of selection bias, study population selected from hospital is less healthy than general propulation
what is Pygmaliion effect?
researcher’s belief in the efficacy of a treatment changes the outcome of that treatment
in positive skew distribution, explain the relationship of mean, median, mode?
mean > median > mode
what statistical hypothesis do you use when there is some association between the dz and the risk factor in the population?
alternative (H1)
what is a type I error?
alpha, incorrectly rejecting the null hypothesis (when null hypothesis is true)
type I error is also known as
false-positive error
