What is nonparametric data?
Data that does not fit the normal distribution. It is often discrete (nominal or ordinal)
How should nonparametric data be analyzed?
Chi square or Mann-Whitney U
What is type I error?
Incorrectly rejects the null, ie interpreted as a difference when there is none
What is type II error?
Incorrectly accepts the null, ie interpreted as equal when in fact a difference exists
What causes type II error?
sample size too small
What is variance?
The spread of data around the mean
When should a t test be used?
When data is quantitatve and you are comparing means.
When should Students t test vs a paired t test be used?
student’s t test: independent groups
paired t test: dependent groups–ie, before and after intervention
When should ANOVA be used?
When comparing means from more than 2 groups.
Formula for relative risk?
incidence in exposed/incidence in unexposed
Formula for power?
Power = probability of making the correct conclusion = 1 - probability of making type II error
Prevalence is
of people with disease in a population
Incidence is
of new cases over period of time
Formula for sensitivity?
true positives/(true pos + false negatives)
Formula for specificity?
true negatives/(true neg + false positives)
In normally distributed data, what proportion is within 1, 2, and 3 standard deviations of the mean?
1- 67%
2- 95%
3- 99.7%
How to calculate SD?
square root of variance
