“…Two categories of determinations based on hadronic measurements enter the World Average: Those based on PDFs [16,17,18,19], which are essentially obtained by optimizing Eq 2 as a function of α S , and the t t production, currently including only the CMS measurement at 7 TeV [20], which is instead based on minimizing over a χ 2 function that considers explicitly the t t data only (we call this the partial χ 2 method ). We shall discuss the relation between these categories, and also try to elucidate the noticeable fact that determinations of α S based on an hadronic dataset, such as Ref [20] as well as more recent ones like Ref [21], give significatively different results from the determinations based on the PDFs that they use as input to compute the predictions in Eq. 1.…”