35 lines
1.4 KiB
ReStructuredText
35 lines
1.4 KiB
ReStructuredText
Statistical Tests
|
|
------------------
|
|
|
|
|
|
Statistical tests are used to decide whether a hypotesis is true or not. To do so, two hypothesis must be formulated:
|
|
|
|
#. :math:`H_0` : The one that is true at first glance (nothing particular, null, boring):
|
|
|
|
- The mean of two populations are equals
|
|
- The average weight of the children in that school is not different from the one of the country
|
|
|
|
#. :math:`H_1` : The one that says: "Something strange is happening"
|
|
|
|
- The mean of two populations are differents
|
|
- The average weight of the children in that school differ from the one of the country
|
|
|
|
.. note::
|
|
|
|
:math:`H_1` can also be noted :math:`H_a` (alternative hypothesis).
|
|
|
|
A statistical test usually works with a **p-value** noted :math:`\alpha`.
|
|
It corresponds to the probability of obtaining the results that you have under the assumption that the null hypothesis is correct.
|
|
In other words, how lucky you are of obtaining these results.
|
|
|
|
Prior performing a statistical test, you must choose a minimal *p-value*.
|
|
:math:`H_0` will be rejected (meaning :math:`H_1` considered true) if the *p-value* obtained from the statistical test
|
|
is lower or equal to the one you choose initially.
|
|
The lower your initial *p-value* is, the more difficult it is to reject the null hypothesis.
|
|
|
|
.. toctree::
|
|
:maxdepth: 2
|
|
:caption: Categories
|
|
|
|
parametric/index
|
|
non-parametric/index
|