Statistics Flashcards
What is the standard deviation?
Measure of how much variation there is from the average set of data
Define relative risk reduction
Percentage decrease in risk comparing intervention to control
i.e.
Mediterranean diet decreased risk of heart attack from 2% to 1 % - RRR 50%
Define absolute risk reduction
Risk in control - Risk in intervention
i.e. Mediterranean diet decreased risk of heart attack from 2% to 1 % - ARR 1%
Risk ratio calculation
Risk in treated group / risk in control
Define p value
Probability that results occurred by chance
Define confidence interval
Range of value that we can be 95% certain the true value lies within
What is paired data?
Two or more measurements taken from the same subject/group
What test would you used for paired parametric data?
Paired T test
What test would you use for paired non-parametric data?
Wilcoxon matched paired-test
What test would you use for unpaired parametric data?
Unpaired T test
What test would you use for unpaired non-parametric data?
Mann-Whitney U test
NNT calculation
1/ARR
Define incidence
Number of new cases per population at given time
Define prevalence
Total number of cases per population at a given time
What is the relationship between the mean and SD in normal distrution?
68.3% lie within 1 SD of the mean
95.4% lie within 2 SD of the mean
99.7% lie within 3 SD of the mean
Define sensitivity
Number of true positives detected by test
Define specificity
Number of true negatives detected by test
Define positive predictive value
Proportion of those who have a positive test who actually have the disease
Define negative predictive value
Proportion of those who test negative who do not have the disease
Define a type 1 error
Test rejects a true null hypothesis i.e. false positive
What is the null hypothesis
There is no different between the intervention and control
What is a type II error?
Test fails to reject a false null hypothesis
How can you reduce the risk of a type II error?
Increase the power of the test (sample size)
Sensitivity calculation
True positive / (True positive + False negative)
Specificity calculation
True negative / (True negative + False positive)
Which tests are prevalence dependent?
Positive and negative predictive values
Which tests are non prevalence dependent?
Likelihood ratios
