Name that Data
Blood type (A, B, AB, O)
Nominal
Continuous data with a normal distribution comparing two samples that are related to one another
Paired T-test
This tells us how likely one can correctly conclude there is a difference
Power
Tells us how large is the difference in study outcome between two treatments under comparison
effect size
This is usually chosen as the value for alpha
0.05
Stage of Cancer
Ordinal
Ordinal data comparing two independent samples
Wilcoxon Rank-sum/Mann-Whitney U
Increasing the sample size has this effect on power
Increases power
The larger the effect size the _________ the power
Higher
Alpha is the maximum allowable probability of making this type of error
type-1 error
Temperature in Celsius
Interval
Two large, unrelated samples of nominal data
Chi-square
This factor does not directly enter into the equation for power and may affect the observed variation in the outcome
study design
Name one way to calculate effect size if the data is binary
1. odds ratio
2. absolute difference in proportions
This type of error is considered a false negative
type-II error
Blood pressure (mmHg)
Continuous
Four samples of independent, normally distributed data compared
One-way ANOVA
This factor has a negative association to power
variation of the outcome
Unknown actual effect size? This is a medical and scientific judgment rather than a statistical decision.
minimum clinically significant effect size
This error is considered a false positive
Type-1 error
Weight categorized into underweight, normal, overweight
Categorical
Four small, unrelated samples of data
Fisher's Exact
This factor is often used interchangeably with power analysis
sample size determination/calculation
Ways to calculate effect size with continuous data
Rejecting the null hypothesis whenever it is actually true
type-I error