16 Nov 2023

Posted on:

25 Jun 2021

0

# Reject or not reject the H0?

Hello,

I don´t know why, but I am a bit confused now. In the lesson "Fitting Higher-Lag AR Model" from the Time Series data section it seemed to be said:
If the p-value is greater than 0.05 we can reject the H0.

But as I know, it is otherwise:
p-value < alpha  --> Reject the H0
p-Value > alpha --> Can not reject the H0

Many thanks in advance. I don´t know why I am struggling.

Posted on:

16 Nov 2023

0

Hi Sebastian, Generally in Statistical Hypothesis Testing ,

1) Setting appropriate Null Hypothesis and Alternative Hypothesis.(note: "=" sign only occurs in null Hypothesis)

2) We need to fix the value for alpha(P(Commiting type 1 error)). Most preferable is 5% level (eg; I like to measure a particular distance from a point A to point B, If I measure it for 100 times that i would make error in measuring that for 5 times out of 100 times).

3)Choose the appropriate test statistic for testing.

4)From the test statistic we used to get p values based on its respective statistical tables( F-test table, Chi sq test table and etc..)

5Now, we need a conclusion for our fixed hypothesis- Accept or Reject the Hypothesis statements.

Using the P value (probability value), a) if obtained p value <= alpha value, we need to reject the Null Hypothesis based on the samples.

b) if p value > alpha value,  we failed accept the Null Hypothesis based on the samples.

Note: P value- is  how likely that the  data results occured under Null Hyphothesis is true. Eg; If my pvalue is 0.08---> My Null Hypothesis  is true based on the samples i have used to calculated and obtained p value supports  8%  to my fixed Null Hyphothesis(0.08 > 0.05).

I hope, you understood. Thank You!

Posted on:

16 Nov 2023

0

And also some statistical test may vary their Hypotheis statements. Be careful while fixing your Hypothesis statements.