Casuality test interpretation help

#1
Dear all,

I'm analysing time series using VECM model and Wald/Granger casuality test. As my reference article I'm using : https://www.researchgate.net/public...growth_Evidence_from_European_Union_countries

I am getting confused at Granger casuality test. In the article, writers ACCEPTED NULL hypotehsis at sig.level <0.05? Is that even true in theory?
My reasearch is also based on EU-28 countries with similar timespan, but my results get almost opposite, since I REJECT null hypothesis at sig. level <0.05 (as it is explained in many guides)

Please advise on this manner.

Thanks a lot.