The Foundations and Methodologies of Human Psychophysics

In her Neuromatch Academy lecture, Prof. Jenny C. A. Read of Newcastle University provides a rigorous introduction to psychophysics, exploring its historical roots, its practical importance in modern technology, and the sophisticated experimental designs required to measure the human experience accurately.

Defining Psychophysics: From Souls to Sensors

Prof. Read opens by tracing the origins of the field to Gustav Fechner in 1860, who ambitiously defined it as an "exact science of the functional or dependence relationships between bodies and souls." While modern definitions are less metaphysical, Prof. Read notes that we now view it as the "analysis of perception by studying the effect on experience of systematically varying the properties of a stimulus." More succinctly, she describes it as the study of the "relationship between stimulus and sensation."

The Practical and Theoretical Importance of Measurement

The ability to measure human perception precisely is not merely an academic exercise. Prof. Read illustrates this through the lens of modern technology:

• Visual Displays: Designing an 8K television requires knowledge of the human "contrast sensitivity function." Prof. Read explains that if a display's detail exceeds our visual system's capacity, the physical improvement remains "imperceptible to me given my visual system."
• Color Theory: She highlights that the concept of "trichromacy"—the idea that three primary colors can synthesize all others—was established via psychophysics in the 19th century long before the biological discovery of the three cone types in the retina.

Weber’s Law and the Logarithmic Mind

A central pillar of the lecture is the "Just Noticeable Difference" (JND). Prof. Read explains Weber’s finding that the JND is not a fixed amount but a "constant fraction of the baseline." Using weight as an example, she notes that while you might notice a 1-gram difference in a 20-gram weight, you would not notice that same gram if you were holding a kilogram. This proportionality, she argues, suggests a "logarithmic encoding of stimulus intensity," where our internal neural signals relate to physical properties sublinearly.

Navigating Experimental Pitfalls: The Decision Criterion

One of the most critical sections of Prof. Read's lecture addresses why simple "Yes/No" tasks can fail. She warns that a subject’s response is often confounded by their "internal decision criterion." A "gung-ho" participant might say they saw a light even when none was present, while a conservative one might wait for a massive signal.

To circumvent this, Prof. Read advocates for the Two-Interval Forced Choice (2IFC) task. By asking a subject which of two intervals contained a flash, the researcher eliminates the need for a subjective criterion. As Prof. Read explains, "the performance is always at 50% for zero signal... and it approaches one as you increase the stimulus intensity," allowing for a pure measure of sensitivity.

Measuring Thresholds: Constant Stimuli vs. Staircases

Prof. Read compares two primary methods for finding the threshold of perception:

• Method of Constant Stimuli: This "gold standard" involves many repetitions of fixed values. While highly reliable, Prof. Read points out it is "very time consuming" and "inefficient" because many trials occur at intensities far from the threshold.
• Staircase Techniques: This adaptive method makes the task harder after a correct answer and easier after a mistake. Prof. Read notes that while a staircase is "much faster"—potentially reaching an answer in 30 trials compared to 300—it can be "thrown off" by early button-press errors.

Bias and Subjective Equivalence

The lecture concludes with a look at discrimination tasks and the "Point of Subjective Equivalence" (PSE). Using optical illusions as examples, Prof. Read demonstrates how two physically identical stimuli can appear different based on context. She underscores the importance of randomizing responses to ensure researchers are measuring "perceptual bias" rather than a simple "response bias" (such as a tendency to press a specific finger).

Final Reflections

Prof. Read concludes by situating psychophysics within the broader goal of neuroscience: relating human experience to neural activity. She argues that because we must measure both sides of this equation with precision, "psychophysics as a field has developed very nice rigorous techniques to measure quantities of interest... while avoiding the confounds which are all too easy to fall into."