Cold and performance: the effects of information load, analgesics, and the rate of cooling

In the first experiment, 8 subjects took part in two experimental sessions--one after receiving one of two kinds of analgesic and the other after receiving a placebo. For each session, subjects completed 8- and 4-choice versions of a serial choice reaction time (SCRT) task at three times: before, during, and after exposure in a cold room at -5 degrees C. Body temperatures were monitored throughout and comfort ratings were recorded. The results indicated that cold increased error rate on the SCRT task, particularly for the 8-choice version which was also associated with faster RTs. However, a speed-accuracy tradeoff explanation for the data was discounted on the grounds that, although overall latencies were faster in the cold compared with precold, these were no faster than postcold RTs. The analgesics had no effect on comfort nor did they have an interpretable effect on performance. In experiment 2, where cooling was much slower for 5 subjects, no such increase in error rate on the SCRT task was observed. Similarly slow cooling did not impair general intellectual ability as measured by Raven's Progressive Matrices.