The random jumps in the resistivity around 250 K suggests that there is very bad contact resistance or a possible alternate current path. The drop in resistance doesn’t look like a typical phase transition (it is very broad in temperature). My guess would be that the voltage leads are not very well connected to the region of the sample that the current is flowing through. This could actually give a drop in the measured voltage (and therefore “resistance”) even though the actual resistance is increasing as the sample is cooled. It would be interesting if they also included two-point resistance across the current leads.