Comment by sokoloff

One way would be by “[reviewing] the underlying research and [deciding] which one is more correct.”