Psychophysics

Psychophysics is the scientific study of relationships between physical stimuli and perceptual phenomena. For example, in the case of vision, one can quantify the influence of the physical intensity of a spot of light on its detectability, or the influence of its wavelength on its perceived hue. Examples could as well be selected from the study of AUDITION or other sensory systems.

In a typical psychophysical experiment, subjects are tested in an experimental environment intended to maximize the control of stimulus variations over variations in the subject's responses. Stimuli are carefully controlled, often varying along only a single physical dimension (e.g., intensity). The subject's responses are highly constrained (e.g., "Yes, I see the stimulus," or "No, I don't see it"). Small numbers of subjects are tested with extensive within-subject designs. Individual differences are small in the normal population. Experiments are routinely replicated and replicable across laboratories.

Theoretical treatments of the data consist of computational or physiological models, which attempt to provide an account of the transformation between stimulus and perception within the sensory neural system. The classic modeling approach is SIGNAL DETECTION THEORY. Many other experimental uses and theoretical treatments are illustrated in the related entries listed below.

A detection threshold is the smallest amount of physical energy required for the subject to detect a stimulus. The example of the intensity required to detect a spot of light will be used throughout the following paragraphs.

There are a variety of formalized methods for measuring detection thresholds (Gescheider 1997). In the method of adjustment, the subject is given control of the intensity of the stimulus and asked to vary it until her perception reaches some criterion value (e.g., the light is just barely visible). In the method of constant stimuli, a set of intensities of the light is preselected and presented many times in a random series. The result is a psychometric function, describing the percent of a particular response (e.g., "Yes") as a function of light intensity. In staircase or iterative methods, the stimulus in each trial is chosen based on the accumulated data from previous trials, using a rule designed to optimize the efficiency of threshold estimation.

When a series of trials is used, there are two major options concerning the subject's task and responses. In Yes/No techniques, the subject reports whether she did or did not see the stimulus in each trial. In forced-choice techniques, the stimulus is presented in one of two spatial locations or temporal intervals, and the subject's task is to judge in which location or interval the stimulus occurred.

Briefly, the method of adjustment has the advantage of maximal efficiency if the effects of stimulus history and subject bias are small. The method of constant stimuli reduces the influence of stimulus history. Forced-choice techniques have the advantage of minimizing the influence of subject bias, and forced-choice iterative techniques often provide an optimal balance of efficiency and bias minimization.

A discrimination threshold is the smallest physical difference between two stimuli required for the subject to discriminate between them. Measurement techniques are analogous to those used for detection thresholds.

In supra-threshold experiments, the subject views readily visible stimuli and is asked to report the quantity or quality of her own SENSATIONS. For example, in a scaling experiment, a subject could be shown lights of different intensities. The task would be to describe the perceived brightness of each stimulus by assigning a number to it. The result is a description of how brightness grows with intensity. In a color-naming experiment, a subject could be shown lights of different wavelengths. The task would be to describe the perceived hues using a constrained set of color names (e.g., red, yellow, green, and blue), and partitioning the perceived hue among them (e.g., "10% Green, 90% Blue"). The result is a description of the variations in hue across the spectrum.

Psychophysics is also marked by a variety of experimental paradigms with established theoretical interpretations (Graham 1989; Wandell 1995). For example, in the summation-at-threshold paradigm, thresholds are measured for two component stimuli and for a compound made by superimposing the two. The goal is to quantify the energy required for detection of the compound stimulus with respect to the energy required for detection of each of the components. Outcomes vary widely, from facilitation to linear summation to independent detection to interference, and lead to different theoretical conclusions concerning the degree and form of interaction of the mechanisms detecting the two components.

Similarly, selective adaptation paradigms examine the extent to which adaptation to one stimulus affects the threshold for another, and are used to argue for independent or nonindependent processing of the two stimuli. Identification/detection paradigms examine whether or not two different stimuli can be identified (discriminated) at detection threshold, and are used to determine the extent to which the mechanisms that detect the stimuli also code the properties required for identification.

Much psychophysical research is guided and united by sophisticated mathematical modeling. For example, spatiotemporal aspects of vision have been treated in extensive theories centered around the concept of multiple, more or less independent processing mechanisms (e.g., Graham 1989; see also SPATIAL PERCEPTION), and COLOR VISION has been similarly unified by models of the encoding and recoding of wavelength and intensity information (e.g., Wandell 1995).

Models of psychophysical data are often heavily influenced by the known anatomy and physiology of sensory systems, and advances in each field importantly influence the experiments done in the other. For example, the psychophysical trichromacy of color vision provided the first evidence for the presence of three and only three channels underlying color vision, and provided the impetus for a century of anatomical, physical, physiological, and genetic as well as psychophysical attempts to identify the three cone types. As a converse example, anatomically and physiologically based models of parallel processing of color and motion have importantly influenced the psychophysical investigation of losses of motion perception for purely chromatic (isoluminant) stimuli (see MOTION, PERCEPTION OF). Treatments are also available of the need for special bridge laws, or linking propositions, in arguments that attempt to explain perceptual events on the basis of physiological events (Teller 1984).

In sum, psychophysics underlies the accumulation of knowledge in many parts of perception. It has many tools to offer to cognitive scientists. Its empirical successes illustrate the value of tight experimental control of stimuli and responses. It provides experimental paradigms that can be generalized successfully to higher level perceptual and cognitive problems (e.g., Palmer 1995). The extensive modeling in the field provides successes that might be worth emulating, and perhaps blind alleys that might be worth avoiding. Finally, and most importantly, in combination with direct studies of the neural substrates of sensory processing, psychophysics provides an important example of interdisciplinary research efforts that illuminate the relationship between mind and brain.

References

Gescheider, G. A. (1997). Psychophysics: The Fundamentals. 3rd ed. Hillsdale, NJ: Erlbaum.

Graham, N. V. S. (1989). Visual Pattern Analysers. New York: Oxford.

Palmer, J. (1995). Attention in visual search: Distinguishing four causes of a set-size effect. Current Directions in Psychological Science 4:118-123.

Teller, D. Y. (1984). Linking propositions. Vision Research 24:1233-1246.

Wandell, B. A. (1995). Foundations of Vision Science. Sunderland, MA: Sinauer.

Psychophysics

See also

References