Two recent publications have reported intriguing analyses, tentatively suggesting that some aspects of IceCube data might be manifestations of quantum-gravity-modified laws of propagation for neutrinos. We here propose a strategy of data analysis which has the advantage of being applicable to several alternative possibilities for the laws of propagation of neutrinos in a quantum spacetime. In all scenarios here of interest one should find a correlation between the energy of an observed neutrino and the difference between the time of observation of that neutrino and the trigger time of a GRB. We select accordingly some GRB-neutrino candidates among IceCube events, and our data analysis finds a rather strong such correlation. This sort of studies naturally lends itself to the introduction of a false alarm probability, which for our analysis we estimate conservatively to be of 1%. We therefore argue that our findings should motivate a vigorous program of investigation following the strategy here advocated.