This research investigated the effect of four thermal environmental conditions on the performance of automatic and controlled processes. The environmental conditions included temperature and relative humidity combinations of (1) 25°C with 60% r.h., (2) 33°C with 38% r.h., (3) 33°C with 66% r.h., and (4) 33°C with 92% r.h . These combinations corresponded to vapor pressures of 15 mm Hg, 15 mm Hg, 25 mm Hg, and 35 mm Hg respectively.

To analyze the data from both the automatic and the controlled processing task, data were transformed to an equivalent scale using proportion scores. F-values well below 1 indicated that variability dominated the experiment. Type of processing was the only significant factor in the experiment.

Upon analyzing each task separately, it was discovered that the major source of variability was the controlled processing task. The automatic processing task had no significant main effects or interactions. The three levels of vapor pressure were almost found to be significantly different (Pr> F=O.07). All analyses of the controlled processing task were dominated by variability.

A larger sample size would be needed to find statistically significant differences in observed means and standard deviations. Power analyses indicate hundreds, and in some cases even thousands, more subjects would need to be run for the controlled processing task or proportion score analyses. Although fewer subjects are needed for the automatic processing task, the power of the experiment was very low.

Future researchers are advised to improve or replace the controlled processing task and to use more subjects. In addition, the variability of the experiment should be reduced by (1) choosing a more homogeneous group of subjects, (2) providing an incentive to the subjects to provide a constant level of effort, and (3) using more extreme environmental conditions.