Hypothesis Testing

Run twenty tests on random data and one comes back significant. That is not a flaw in the method, it is what the method promises. This course teaches you to write the claim first, test it once, and read the answer for what it says.

₹499/- ₹0

Offer ends in 5 days

1,000 seats total 262 left

738 already enrolled

Six things you will be able to do

Go past the p value threshold. Learn what a test actually asks, and what a significant result does not establish.

01 State the claim firstA null and an alternative, written before the data is opened.
02 Read a p value correctlyWhat it measures, and the far more common thing people think it measures.
03 Choose the right testOne sample, two sample or paired, and what each one assumes about your data.
04 Weigh the two errorsFalse positives against false negatives, and which one your context can afford.
05 Report the effect sizeHow large the difference is, alongside whether it is distinguishable from noise.
06 Count every test you ranMultiple comparisons, and why the twentieth result needs a higher bar.

What changes after this course

You stop reading a small p value as proof and start asking how many tests preceded it, and how large the effect actually is.

A question written as a testable claim before any data is opened

α

A p value read for what it measures, and not for what it is assumed to prove

An effect size and a confidence interval, so significance has a magnitude

What our learners say

LN
Learner name Role or college

One line about what changed for them.

LN
Learner name Role or college

One line about what changed for them.

LN
Learner name Role or college

One line about what changed for them.