Statistical distributions and tests
The statistical tools of the programme: mean, median and standard deviation, the normal distribution, confidence intervals, the null hypothesis and the p-value, alpha and beta risks and power, Student's t and chi-squared tests, correlation, and the interpretation traps.
- 36 cards
- 3 topics
- Free
The outline, topic by topic
Each topic is weighted the way the exam weighs it.
- Concepts
- Scenarios
- Understand
A few cards from the deck
Mean and median: what is it?
Show the answer
Mean: sum of values divided by their number. Median: the value splitting the sample into two halves.
The median resists extreme values.
You compare the mean blood pressure of two groups of patients. Which test?
- Student's t-test
- A chi-squared test
- A correlation coefficient
- McNemar's test
Show the answer
Student's t-test
A quantitative variable compared between two independent groups: Student's t-test.
Standard deviation: what is it?
Show the answer
A measure of how values spread around the mean.
Not to be confused with the standard error, which measures the mean's precision.
How can a study's power be increased?
- By lowering the alpha risk from 5 % to 1 %
- By switching from a one-sided to a two-sided test
- By choosing a more variable outcome measure
- By increasing the number of participants
Show the answer
By increasing the number of participants
More participants shrink the standard error: a real difference of the same size becomes easier to detect.
Normal distribution: what is it?
Show the answer
A symmetrical bell-shaped distribution in which about 68 % of values lie within one standard deviation of the mean and 95 % within two.
Many biological measures roughly follow it.
You compare the percentage cured under two treatments. Which test?
- Student's t-test
- A chi-squared test
- A correlation coefficient
- An analysis of variance
Show the answer
A chi-squared test
Cured or not is a categorical variable: proportions are compared.