1/f Noise and Effort
What is 1/f noise?
1/
f noise refers to a complex pattern of variation. This pattern is intriguing in part because it seems to characterize a stunning range of phenomena (e.g., earthquake magnitude, the intensity of pulsar emissions, auditory frequencies in music). These phenomena may appear random, but the presence of 1/
f noise reveals something systematic in those fluctuations. I am currently examining changes in response latency that occur from one trial to the next as participants perform tasks like our shoot/don't-shoot simulation. Consistent with work in cognitive psychology, latencies in these tasks seem to be characterized by robust 1/
f noise, ultimately suggesting that trial-to-trial variation may be due, in part, to random noise, but also (and more interestingly) that powerful, low-frequency rhythms - distinctly nonrandom fluctuations that occur slowly over the course of 200 trials — exist in the data as well (Correll, 2008). It is critical to note that evidence of 1/
f noise in latencies strongly suggests that those fluctuations are at least partly systematic in nature. And, just as systematic differences in the average response latency to Black and White faces can inform us about implicit bias, systematic differences in patterns of variability may hold important clues about psychological states. For example, there is some evidence that 1/
f patterns, which are robust on simple tasks, can be attenuated by making the task more difficult. In essence, as the task becomes more difficult, the typically powerful low-frequency fluctuations become less pronounced. My own work extends this idea to argue that 1/
f noise depends on task-related effort. In recent studies, I have either manipulated or measured the degree to which participants attempt to avoid showing racial bias. In all cases, as efforts to avoid bias increase, 1/
f noise decreases. These data suggest that trial-to-trial variability, which researchers typically treat as meaningless error, may actually provide valuable clues about a participant's orientation to the task at hand.
