Big data sampling and spatial analysis: “which of the two ladles, of fig-wood or gold, is appropriate to the soup and the pot?”

Following from Krivoruchko and Bivand (2009), we consider some general points related to challenges to the usefulness of big data in spatial statistical applications when data collection is compromised or one or more model assumptions are violated. We look further at the desirability of comparison of new methods intended to handle large spatial and spatio-temporal datasets.