The analysis of the uniformity of a spherically averaged Hubble expansion in the Local Group frame of reference by Wiltshire, Smale, Mattsson and Watkins (2013) is extended. We carry out an investigation to constrain the frame of reference from which the spherically averaged Hubble expansion is the most uniform by applying arbitrary Lorentz boosts to the data. The proposition of a systematic boost offset between the Hubble expansion in the Local Group and CMB reference frames is verified within statistical uncertainties. This evidence further supports the claim that the Local Group is closer to the frame of reference in which Hubble expansion should be considered. We subsequently carry out a statistical analysis in search of a frame of minimum expansion variation and find consistent results with the systematic boost offset analysis. However, there is a considerable degeneracy to perform boosts in the plane of the galaxy, which may be a consequence of a lack of constraints from the Zone of Avoidance where data is absent. The COMPOSITE sample of 4,534 galaxies is used primarily, with the key results repeated with the recently released Cosmicflows-2 sample of 8,162 galaxies.

The treatment of Malmquist distance bias is investigated in the context of the Cosmicflows-2 and COMPOSITE samples. We find systematic differences in the inclusion of the large SFI++ subsample into these catalogues. These differences are explored and the origin of Malmquist distance bias reviewed. We find the Cosmicflows-2 data produces results which naively suggest more variation of cosmic expansion than would be expected in any cosmological model when the methods of Wiltshire et al. are applied. We trace this discrepancy to the fact that the distribution Malmquist biases have not been corrected for in the Cosmicflows-2 survey.