Question regarding hypothesis testing -- In practice

I read this from the course < Master Statistics with Python Skill Path>:

  • In practice, the alternative hypothesis and significance threshold should be chosen prior to data collection.
    Does anyone know why is the order important?
    Thanks in advance!

The idea is to have an “objective” condition to use to evaluate your hypothesis.
So you choose the significance level before doing testing a hypothesis.

If you’d collect the data first,
there may be a temptation to choose the significance level
to make the hypothesis test turn out to be the result that you prefer;
which would not be an objective result.

2 Likes

Thank you very much for the clear explanation!