Hi Ahmed

The rejection rule is

Reject the Null hypothesis if p-vslue less than or equal 0.05 at. 95% level of significance.

This is a feature of bootstrapping as you are trying to approximate a sampling distribution for you statistic by random draws with replacement from the raw data (or some function of the raw data). If the bootstrapping produces unstable results on different runs you need to increase the number of bootstrap samples or (if that is computationally problematic) you can average over a number of bootstrap runs to get a more stable result. If the variation is not great enough to change the results then you can pretty much ignore it.

Thank you for your comprehensive explanation.

Due to the random generation of data

Thank you !

Please,

I draw a model, i have exactly in p-value = 0.050 and ( β = 0.234 and Std. error = 0.119 )

Is my hypothese is supported or not ? how can I interpret it ?

In advance thank you for your cooperation.

Hi Ahmed

The rejection rule is

Reject the Null hypothesis if p-vslue less than or equal 0.05 at. 95% level of significance.

Thanks a lot.