We use ECOPIA Hall Effect measurement system (Model no. HMS-5300 + AMP55T magnet) in our lab. I am trying to determine the carrier concentration and mobility of PbS thin films. I have made electrodes at the four corners using Gold. I have measured I-V curves of the samples, the shape of the I-V curve is linear (an ohmic contact is formed between PbS and Gold). I did hall measurement of a sample at same condition for three times, it gave me positive hall coefficient for two times and negative hall coefficient for one time. The value is not in the same order and the polarity changes randomly. Even successive measurement runs on the same sample without changing anything at all, is showing completely different results. Why does it show positive and negative Hall Coefficient randomly in different for the same sample at the same temperature? What could possibly be causing this behavior and how can it be overcome?