Academic journal article Psychonomic Bulletin & Review

How Inherently Noisy Is Human Sensory Processing?

Academic journal article Psychonomic Bulletin & Review

How Inherently Noisy Is Human Sensory Processing?

Article excerpt

Like any physical information processing device, the human brain is inherently noisy: If a participant is presented with the same sensory stimulus multiple times and is asked to press one of two buttons in response to some property of the stimulus, the response may vary even though the stimulus did not. This response variability can be used to estimate the amount of so-called internal noise-that is, noise that is not present in the stimulus (such as random dynamic dots on the screen) but in the participant's brain. How large is this internally generated noise? We obtained >400 independent estimates on 40 participants for a range of protocols (yes/no, two-, four-, and eight-alternative forced choice), modalities (auditory and visual), attentional state, adaptation state, stimulus types (static, moving, stereoscopic), and other parameters (timing, size, contrast). Our final estimate at ~1.3 (units of external noise standard deviation) is generally somewhat larger than that previously inferred from smaller and less varied data sets. We discuss the impact of high levels of internal noise on a number of experimental and computational efforts aimed at understanding and characterizing human sensory processing.

Imagine running a typical two-alternative forced choice (AFC) experiment in which a human participant is presented with two sensory stimuli on every trial, one containing a signal with some added noise, the other one containing noise alone. The observer is required to choose the stimulus containing the signal, and the task is repeated for 100 trials. The two noise samples presented on individual trials are different on every one of these 100 trials, and the intensity of the signal is similar to the intensity of the noise. We end up with a sequence of 100 binary responses. We then run a second experiment in which we show the same exact 100 trials to this participant. This means that on each trial, the noise samples are identical to those used in the previous experiment, the signal to be detected is added to the same stimulus, and everything that is shown to the observer is identical to what was shown during the previous experiment. We end up with a second sequence of 100 binary responses. We then ask the question: Out of 100 trials, how many times did the participant give the same response during Experiment 1 and during Experiment 2?

It is perhaps surprising that in a typical experiment, the same response happens only on roughly three out of four trials (Burgess & Colborne, 1988; Green, 1964). Considering that it is expected to happen on one out of two trials simply by chance (i.e., even if the participant pressed buttons randomly), three out of four may strike one as a rather poor degree of agreement between the participant and himself /herself. The stimulus is exactly the same: From the point of view of the sensory information that is delivered to the participant and the task that he/she is required to perform, the question is exactly the same. Yet, the participant cannot agree with himself /herself on more than three out of four trials. We must conclude that the specific choice generated by a human participant on a given trial is not a deterministic function of what is happening on the monitor but also depends, to a large extent, on a loud source of variability that is not under direct experimental control: internal noise (Barlow, 1956; Pelli, 1990). The importance of this source of variability was first emphasized by Green. Over the decades that followed, some important studies (e.g., Burgess & Colborne, 1988) have added relevant knowledge, but there has been no attempt to provide a more comprehensive view of this phenomenon. Our motivation here is to sketch such a view, given current knowledge. In this sense, the present article can be intended as an update to Green's original contribution.

We are not concerned here with the exact source of this variability. The participant may respond differently on a repeated trial because he or she sneezed the second time, or blinked, or inadvertently pressed a button other than the one that he or she meant to press. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.