For the simple Delta I mounting an analysis is presented for the difference between thermistor and air temperature. For an alternative mounting, e.g., the STS-I, using a Mylar film sheet to isolate the sensor from the support posts, the same problem is treated by an extension of this analysis. It is concluded that the Delta I mounting has the advantage of only a small variation due to solar radiation, i.e., between daytime and nighttime measurements. Its disadvantage is that the difference between thermistor and air temperature is greater, i.e., a greater correction needs to be applied. On the other hand, this correction can be calculated with less uncertainty, because (i) the analysis is simpler, (ii) the effect of uncertainty in the support post temperature is largely removed once the thermistor leads exceed a certain length, and (iii) the coefficient of heat convection is more reliably computed for spheres and cylinders than for a flat plate, which has a varying angle of incidence with the direction of flow.