Track 8 · Hypothesis Testing · Lesson 2 of 14
Null and alternative hypotheses
Write null and alternative hypotheses about a population parameter, and choose a one- or two-sided alternative before seeing the data.
25 minIntermediate7 graded exercisesMembers
A courier company introduces new routing software and wants to know whether delivery times have changed. Before a single number is computed, the team must decide exactly what claim is being tested — about which quantity, in which direction. Get this wrong and even a perfect calculation answers the wrong question.
Every test compares two competing statements about a population parameter — a fixed but unknown number that describes the whole population or process:
- the null hypothesis H0 is the sceptical, “no effect / no difference / status quo” claim. It always contains an equality, because it must pin down one specific value we can compute with;
- the alternative hypothesis H1 (also written ) is what we would conclude if the data provide enough evidence against H0. It is usually the claim the study hopes to support.
Included with a Formula Quest membership
It has 7 graded exercises with feedback, hints and worked solutions. After it you will be able to:
- Write H0 and H1 about a population parameter, with the equality in H0
- Identify the right parameter: μ, p, μ₁ − μ₂, μ_d or p₁ − p₂
- Choose a one- or two-sided alternative from the question, before seeing the data
- Explain why picking the direction after looking at the data doubles the false-positive rate