Hypothesis testing questions define the error types and ask what changes power.
Errors
Type I error (alpha): rejecting a true null hypothesis; a false positive; usually set at 0.05. Type II error (beta): failing to reject a false null hypothesis; a false negative. Power equals one minus beta, conventionally 80 percent. Power increases with larger sample size, larger effect size and higher alpha.
P-values and intervals
The p-value is the probability of seeing data at least this extreme if the null hypothesis is true. A 95 percent confidence interval that does not include the null value (0 for a difference, 1 for a ratio) corresponds to p less than 0.05.
Which test
T-test compares means of two groups; ANOVA compares means of three or more; chi-square compares categorical data; paired t-test compares before and after in the same group; correlation (Pearson) measures linear relationship and regression predicts outcomes. Nonparametric alternatives are used when data are not normal.
Keep going in biostatistics, epidemiology and ethics
- Normal distribution, standard deviation and standard error
- Screening and prevention: incidence, prevalence and prevention levels
- Survival analysis: Kaplan-Meier curves and hazard ratios
- Patient safety and quality improvement: errors, RCA and handoffs