Lecture 7 Flashcards
What is a statement of risk?
A statement about the dangers of something that can either be believed or disbelieved.
What are important factors to be included in a risk/benefit experiment?
- Realistic time scale
- Statistically valid (clear outcome)
- Multiple tests and repeats
- A control group (while considering fairness)
- Minimal side effects
- Ethical
- Unbiased
- Representative
Define relative risk (RR)
The increase in incidence of a disease in the exposed group compared to the non-exposed group, found by calculating ratio of the probabilities of an event occurring in the exposed group to that in the non-exposed group.
What is the equation for relative risk?
RR = relative risk
p = probability
What is the equation for absolute risk increase?
ARI = absolute risk increase
p = probability
What is the equation for relative risk increase (RRI)?
RRI = relative risk increase
p = probability
What is the equation for the number needed to harm (NNH)?
NNH = number needed to harm
ARI = absolute risk increase
What are relative risk increase and number needed to harm measured in?
%
Define relative risk of success (RR)
The ratio of the probabilities of a positive outcome in the treated group (i.e. reduced incidence) to that in the non-exposed group.
What is the equation for the relative risk of success?
RR = relative risk
p = probability
What is the equation for absolute risk reduction?
ARR = absolute risk reduction
p = probability
What is the equation for relative risk reduction?
RRR = relative risk reduction
p = probability
What is the equation for the number needed to treat?
NNT = number needed to treat
ARR = absolute risk reduction
What method can be used to determine if a trial result is significant?
The null hypothesis can be tested against.
What is the p-value of significance if we want to be 95% confident that a result is not due to chance?
p = 0.05 (a.k.a 5% probability that the result is due to chance)
Define null hypothesis
The hypothesis that there is no change in the outcome of a trial for a treated group compared to a control group.
Define standard error
A measure of the statistical accuracy of an estimate.
For a graph (bar chart) of the recovery rate of a control group and a treated group, how can the significance of the treatment be determined?
The treatment is significant if error bars do not overlap.
Define standard deviation
The square root of the variance.
What is the equation for standard deviation?
s = standard deviation
N = number of measurements made
x = measurement
x̄ = mean measurement
What is the equation for the standard error of the mean (SEM)?
SEM = standard error of the mean
s = standard deviation
N = number of measurements made
What is the student’s t-test?
A useful measurement when comparing small samples of two distinct groups (test and control) to look for a significant difference.
The student’s t-test is an _________ test when the two populations don’t depend on one another.
Unpaired
What is a single-tailed test?
A test that is only concerned with the significance of one end of the data (so either greater than or less than a specific value).