Comparing two methods of informing operators about what might happen next

Operator performance is dependent upon the level of certainty of what will happen next. Knowledge about this certainty can be given by long-term experience of the occurrence of events or by secondary information given immediately before the event. We define these methods as, respectively, memory-dependent and display-dependent. In this experiment nine groups of twelve subjects performed four-choice finger-pressing tasks in response to visual signals. The tasks were based upon the unbalanced frequencies and response priming paradigms of experimental psychology corresponding to the two methods of providing information thus defined. The tasks varied in mean certainty as well as in source of knowledge about the certainty. Mean reaction times varied, to a statistically significant degree, with level of certainty and, for events of low certainty, with the method of presenting information about certainty. Error rates varied inversely with the same factors but not all differences in this respect were statistically significant. We conclude that the display-dependent method of presenting information about certainty produces slower but more accurate responses for events of low certainty. We discuss the implications of these results for occupational tasks.