Temperature sensitivity: One subjective continuum or two?

Abstract

The full range of nonpainful temperatures (approximately odeg C to 52 deg C) was scaled by magnitude estimation with the standard at physiological zero (32 deg C). Apparent magnitude bore an exponential relationship to temperature over this range, although the same Os gave power functions when judging warmth or cold alone. A transformation is presented, which, when applied to the appropriate sections of the exponential relationship for the full range, produced functions indistinguishable from the power law for warm and cold ranges. The conclusion Stevens and Stevens (1960) drew from the results of scaling warmth and cold alone, namely that a ϕo threshold correction should be applied to the physical scale of all continua to derive power law functions from magnitude estimations, is shown to be an invalid generalization from temperature scaling.