The experimental(ish*) results in the search for variation in the fine structure constant are … (excitement building) … (butterflies in stomach) … (forehead getting shiny) … inconclusive! Oh … right.
In brief, Molaro et al. have run their many multiplet analysis code to estimate Δα/α for the five line-of-sight absorbers towards quasar, HE2217-2818, which lies on a sightline for which Webb et al.’s dipole-model predicts a Δα/α of 5.4±1.7 ppm. And they come up with an aggregate observational result of <Δα/α> = 1.3±2.4±1.0 ppm (the two error terms being random and systematic, respectively). With the null hypothesis being strictly Δα/α=0 the authors have chosen to conclude that their dataset is neither consistent nor inconsistent with either competing model. Fair enough, I suppose, in the context of their on-going study … why jump to unnecessary conclusions that you might just have to retract when the rest of your data arrives?!
* [I say “ish” because a lot of dark-arts data reduction seems to inevitably go on in all these QSO absorption line studies]
Update: I have placed an R script for computing the Bayes factor of each of the Webb et al. team’s dipole models compared to the null, given the data from all four post-Webb et al. 2011 observational studies, on an ASAIP forum thread here. The result is decisive!