Chapter 2: Approaches to Studying Sensation and Perception

Introduction

Sensation and perception are essential components of our daily lives, allowing us to interact with the world around us. In this chapter, we will explore various approaches to studying sensation and perception.

The Gestalt Approach

One of the foundational approaches to studying sensation and perception is the Gestalt approach. According to this perspective, we do not perceive objects and stimuli as isolated elements but as wholes. The Gestalt approach emphasizes that perception goes beyond the mere presence of stimuli in the environment. It also involves emergent features, where our minds create meaningful patterns and structures (Koffka, 1935). For example, when presented with four Pac-Man figures arranged in a square, most individuals perceive the parts in addition to a square shape, even though the square is not explicitly drawn (Kanizsa, 1976). Additionally, the perceived brightness of certain regions may differ, with many people seeing the central square region as brighter than the background, demonstrating the emergence of features beyond the physical stimulus.

Figure 1.1

Pacman shaped figures when arranged with the cutout areas facing each other can create the global image of a square.

"Pacman shaped figures." by Kahan, T.A. is licensed under CC BY-NC-SA 4.0

The Gibsonian Approach

In contrast to the Gestalt approach the Gibsonian approach (named after the American psychologist J.J. Gibson) posits that perception is a direct process (Gibson, 1979). It suggests that our perception of the world reflects the rich sensory information available in our environment. The Gibsonian perspective emphasizes that our perception is shaped by the stimulus hitting our sensory receptors. However, it faces challenges when dealing with visual illusions and situations where our perception does not align with the physical world. Similarly, this approach has difficulties explaining mental imagery (where perceptual experiences occur in the absence of sensory information).

The Information Processing Approach

The information processing approach views perception as a multi-stage process influenced by various interconnected factors (Neisser, 1967). This model likens perception to a series of stages, each influencing the next. It emphasizes the flow of information through these stages and how they collectively shape our perception of the world. This approach is characterized by the interaction of different stages, demonstrating the dynamic nature of information processing.

The Computational Approach

The computational approach employs mathematical models, such as perceptrons and machine learning, to simulate how humans perceive the world (Rosenblatt, 1958)^[1]. This perspective forms the foundation of artificial intelligence and is crucial for understanding how computer systems process information. While it is not extensively covered in this class, it plays a vital role in modern technology and the development of AI systems.

The Phenomenological Approach

The phenomenological approach, although less favored among perception researchers today, is rooted in descriptive exploration of human perception. It involves asking individuals to describe their perceptual experiences without imposing experimental constraints. This approach can be valuable for gaining initial insights into perceptual phenomena. For example, it is illustrated by phenomena like the Necker cube (Necker, 1832) (where people perceive an inverted and stationary mask as following you as you move) (Gregory, 1970). The Necker cube can be perceived in two ways; either the lower-right square can be seen toward the front or the upper-left square can be seen toward the front.

Figure 1.2

The Necker cub is a simple wireframe cube that affords two competing perceptions: one face can appear closer to the viewer, then can recede as another face becomes prominent.

"The Necker cube." by Kahan, T.A. is licensed under CC BY-NC-SA 4.0 

Psychophysics

Psychophysics is a method used to quantify the relationship between physical stimuli and our perceptual experiences (Fechner, 1860). Gustav Fechner, credited with establishing this field, introduced the concept of the absolute threshold—the smallest amount of a stimulus required for detection about half of the time.

Measuring Sensory Thresholds

A central goal of psychophysics is determining sensory thresholds. Psychophysicists are interested in two important types of thresholds:

Absolute Threshold: The minimum intensity of a stimulus that can be detected approximately 50% of the time. This threshold represents the point at which a stimulus becomes detectable.

Difference Threshold (Just Noticeable Difference; JND): The smallest change in a stimulus that can be reliably detected. Rather than asking whether a stimulus is present or absent, the difference threshold measures the smallest perceptible difference between two stimuli.

For example, an absolute threshold study might investigate the dimmest light a person can detect in a dark room. A difference threshold study might examine how much brighter a light must become before a person notices the change. Similarly, an absolute threshold for weight perception might involve determining the lightest weight a person can detect, whereas a difference threshold task would determine how much additional weight must be added to an object before the change is noticed.

Psychophysicists have developed several methods for measuring these thresholds.

Method of Adjustment

In the method of adjustment, participants directly control the intensity of a stimulus and adjust it until it reaches a specified criterion. For an absolute threshold experiment, a participant might increase the volume of a tone until it is just audible. Alternatively, they might decrease the volume until it is barely detectable. The threshold is estimated from repeated adjustments.

The method of adjustment can also be used to measure difference thresholds. For example, a participant might adjust the brightness of one light until it appears just noticeably brighter than a comparison light.

This method is relatively quick and efficient, although it can be influenced by participant bias because participants control the stimulus themselves.

Method of Limits

In the method of limits, the experimenter presents stimuli in ascending and descending orders.

In an ascending series, stimulus intensity begins below threshold and gradually increases until the participant reports detecting it. In a descending series, stimulus intensity begins above threshold and decreases until the participant reports that it is no longer detectable. The threshold is estimated by averaging the transition points across multiple trials.

For example, to determine an absolute threshold for hearing, an experimenter might gradually increase the volume of a tone until the participant reports hearing it. In descending trials, the tone would start clearly audible and gradually decrease until it can no longer be heard.

Difference thresholds can also be measured using this method. An experimenter might gradually increase the weight of a comparison object until the participant reports noticing that it is heavier than a standard weight.

Although the method of limits is more systematic than the method of adjustment, participants may develop expectations about when they will perceive the stimulus (or the change).

Method of Constant Stimuli

The method of constant stimuli is considered one of the most accurate methods for estimating thresholds. In this procedure, stimuli of varying intensities are presented in random order across many trials. The participant simply reports whether the stimulus was detected or whether a difference was perceived.

For an absolute threshold experiment, a range of stimulus intensities surrounding the expected threshold might be presented repeatedly. The threshold is typically defined as the intensity detected on 50% of presentations^[2].

For a difference threshold experiment, participants might compare a standard stimulus to comparison stimuli that vary in intensity. The just noticeable difference is estimated from the smallest change that is detected reliably across trials.

For example, an experimenter studying vision might present lights of varying brightness levels and determine the brightness level detected on half of the trials. To measure a difference threshold, the experimenter might present a standard light and comparison lights of different intensities to determine the smallest brightness difference that can be detected.

Because stimuli are presented randomly, this method reduces response biases and anticipation effects, although it typically requires more trials and takes longer to administer.

From Difference Thresholds to Weber's Law & Steven’s Power Law

Once psychophysicists determine the just noticeable difference (JND), they can examine how the JND changes as the intensity of the original stimulus changes. Research conducted by Ernst Weber revealed that the size of the JND is not constant. Instead, the amount of change needed to notice a difference increases as the original stimulus becomes larger (Weber, 1834).

Weber's Law

Weber's Law, a fundamental principle in psychophysics, states that the just noticeable difference (JND), or the smallest perceptible change in a stimulus, is a constant proportion of the original stimulus (Weber, 1834). Mathematically, it can be expressed as ΔI/I = k, where ΔI is the change in stimulus intensity, I is the initial stimulus intensity, and k is the constant specific to the dimension being studied. As an example, Imagine you are holding a small bag of rice weighing 1 kilogram (1000 grams). According to Weber's Law, the just noticeable difference (JND) in weight perception should be proportional to the original weight. Let's assume the Weber fraction for weight perception is 0.05, meaning that the JND is 5% of the original weight.

1. Original Weight (I): 1000 grams
2. Change in Weight (ΔI): To calculate the JND, we multiply the original weight by the Weber fraction: ΔI = 1000 grams x 0.05 = 50 grams.

So, according to Weber's Law, you would need a change in weight of at least 50 grams for you to reliably detect a difference. If someone added or removed 50 grams from the bag of rice, you would likely notice the change. However, a change of 10 grams might not be noticeable because it falls below the JND dictated by the Weber fraction.

Stevens's Power Law

Stevens's Power Law builds on the psychophysical tradition, focusing on the perceived magnitude of stimuli (Stevens, 1957). This law asserts that our perception of stimuli is not always linear but follows a power function. It can be expressed as P = k * S^n, where P represents the perceived magnitude, S is the stimulus intensity, k is a constant, and n is the exponent that characterizes the specific sensory dimension. This model helps us understand how our perception responds to changes in stimulus intensity. Notably, it highlights that our perception of different sensory dimensions may exhibit response expansion or response compression.

Figure 1.3

Stevens' Power Law describes the relationship between physical stimulus intensity (S) and perceived magnitude (P) as P = kSⁿ, where k is a scaling constant and n is an exponent that varies across sensory modalities. When n = 1, perceived magnitude increases in direct proportion to stimulus intensity (e.g., judgments of line length). When n > 1, perception grows more rapidly than the physical stimulus, producing response expansion (e.g., electric shock). When n < 1, perception grows more slowly than the physical stimulus, producing response compression (e.g., brightness). The curves and numerical values shown in this figure are for explanatory purposes; they do not represent data from any specific empirical study.

"Steven’s Power Law." by Kahan, T.A. is licensed under CC BY-NC-SA 4.0 

The exponent (n) determines how perceived intensity changes as stimulus intensity increases. When n = 1, perceived intensity increases in direct proportion to the physical stimulus, producing a linear relationship. For example, line length is often perceived approximately linearly, so doubling the length of a line makes it appear about twice as long. When n < 1, perceived intensity grows more slowly than the physical stimulus, a phenomenon known as response compression. In this case, large increases in stimulus intensity produce relatively small increases in perceived intensity. Brightness perception is a classic example; doubling the intensity of a light does not make it appear twice as bright. In contrast, when n > 1, perceived intensity grows more rapidly than the physical stimulus, a phenomenon known as response expansion. Here, small increases in stimulus intensity produce disproportionately large increases in perceived intensity. The perception of electric shock provides a clear example, as a modest increase in electrical current often feels much more intense than the physical increase would suggest.

1. Response Expansion (n > 1): When the exponent (n) is greater than 1, it indicates response expansion. In this case, the perceived sensation increases more rapidly than the increase in stimulus intensity. Small changes in stimulus intensity lead to significant changes in perception.
2. Response Compression (n < 1): When the exponent (n) is less than 1, it indicates response compression. Here, the perceived sensation increases more slowly than the increase in stimulus intensity. It takes a substantial change in stimulus intensity to produce a noticeable change in perception.

Let's apply this concept to two examples:

1. Perception of Electric Shock (Response Expansion): Imagine measuring the perceived pain of an electric shock. Suppose that a small increase in electrical current results in a significantly more painful perception. In this case, the Stevens Power Law might describe this relationship with an exponent (n) greater than 1. This indicates response expansion, meaning that small changes in current result in a disproportionately larger change in perceived pain.

2. Perception of Brightness (Response Compression): Consider the perception of brightness when varying the intensity of a light source. In this case, you might find that increasing the light intensity by a small amount doesn't lead to a very noticeable change in brightness perception. The Stevens Power Law might describe this relationship with an exponent (n) less than 1. This indicates response compression, meaning that you need a relatively large change in light intensity to produce a noticeable change in brightness perception.

Thus, Stevens's Power Law extends psychophysics beyond simple threshold measurements by describing how the magnitude of sensory experiences changes across a wide range of stimulus intensities. Different sensory systems are characterized by different exponents, revealing that the relationship between the physical world and our perceptual experience is often nonlinear.

Thresholds and Sensitivity

Traditionally, psychophysics has focused on determining sensory thresholds, such as the absolute threshold (the minimum intensity needed to detect a stimulus) and the difference threshold (the smallest change in stimulus intensity that can be detected). However, signal detection theory challenges the concept of thresholds altogether (Green & Swets, 1966).

Signal detection theory suggests that we don't have fixed thresholds. Instead, our perception is influenced by various factors, including the context of the situation and the instructions given. We then set a criterion for responding, which can vary from being a conservative criterion (requiring strong evidence to indicate that a stimulus was present) to a liberal criterion (requiring little evidence to indicate that a stimulus was present). This criterion can change depending on the situation and the individual.

Signal Detection Theory

Signal detection experiments involve the use of catch trials, where no stimulus is presented, to assess how observers respond.

To understand signal detection theory, let's consider an analogy. Imagine you are on a whale-watching trip, trying to spot whales in the ocean. Your decision to shout "whale" or not depends on various factors, including the situation and the people around you. If you're with family, you might be more liberal in your criterion for identifying a whale.

However, with strangers, you may adopt a more conservative criterion to avoid false alarms.

In signal detection experiments, there are two main situations: trials with no signal (just noise; e.g., no whale just waves) and trials with a signal embedded in noise (e.g., a whale in the ocean). Observers must decide whether a stimulus is present or not. This decision is influenced by the amount of evidence suggesting the presence of a signal. Signal detection theory distinguishes between hits (correctly detecting a stimulus), misses (failing to detect a stimulus), false alarms (incorrectly detecting a stimulus when none was present), and correct rejections (correctly identifying the absence of a stimulus). Sensitivity in signal detection theory is measured using "d prime" (d') (for a wonderfully clear and detailed review of how to calculate many different signal detection measures see Stanislaw & Todorov, 1999). D prime represents the difference between the observer's ability to discriminate between signal and noise. A larger d' value indicates higher sensitivity, while a lower value suggests lower sensitivity. The d prime (d') value quantifies how well an observer can distinguish between signal and noise. The numerical value represents the distance between the mean of the signal distribution and the mean of the noise distribution in standard deviation units. A value of zero suggests that there is no separation between the signal and noise distributions (i.e., the observer cannot distinguish between signal and noise). As d' increases the observer can discriminate between signal and noise to some degree. Values between 0.5 and 1.5 indicate fair discrimination and values greater than 2 indicate a strong ability to distinguish between signal and noise (a value of 3 or more is nearly perfect performance).

Figure 1.4

The basics of signal detection theory with frequency (y axis) plotted as a function of sensory information (x axis) for situations when no signal is presented (top curve) as well as when a signal is embedded in noise (bottom curve). When the evidence is above the criterion a yes response is given and when the evidence falls below the criterion a no response is given. This results in hits, misses, correct rejections, and false alarms.

"Basics of signal detection." by Kahan, T.A. is licensed under CC BY-NC-SA 4.0

Additionally, signal detection theory allows for the analysis of an observer's criterion. A liberal criterion results in more hits but also more false alarms, while a conservative criterion leads to fewer hits but fewer false alarms. The choice of criterion can be influenced by factors such as instructions and the consequences of making errors.

Physiological Approaches

Physiological approaches to studying sensation and perception delve into the underlying neural and physiological processes responsible for our perceptual experiences. These approaches focus on understanding how the brain processes sensory information and how physiological factors contribute to perception.

Neural Communication and Physiology

As you will see throughout this course, basic physiology of neural communication play a crucial role in understanding perception. Neurons communicate through electrical signals, involving the flow of ions, particularly sodium and potassium. Neurons have a resting electrical charge, which changes when they are stimulated. Sodium ions rush into the neuron, making the charge more positive, while potassium ions later rush out to restore the resting state.

This process results in an action potential, a brief electrical signal that travels along the neuron. The action potential is an all-or-none response, meaning it occurs at full intensity or not at all. The rate of firing of neurons can change in response to stimulus intensity, rather than the size of the action potential itself (which is unchanged).

Recording Neural Activity

Researchers use various techniques to study neural activity associated:

Modular Organization of the Brain

Physiological approaches have contributed to our understanding of the modular organization of the brain, where specific regions are specialized for particular functions. For example, the fusiform face area (FFA) in the right hemisphere is associated with face processing (Kanwisher et al., 1997). These findings support the idea that different brain regions play distinct roles in perception (Fodor, 1983).

In the following chapters, we will delve deeper into the physiological processes and neural mechanisms underlying specific aspects of sensation and perception, exploring how our brain interprets and makes sense of the sensory world around us.

Conclusion

The study of sensation and perception encompasses various approaches that have contributed to our understanding of how we make sense of the world. From the holistic perspective of Gestalt psychology to the computational models of artificial intelligence, these approaches offer unique insights into the complex processes governing our perception. Additionally, psychophysical principles like Weber's Law and Stevens's Power Law provide a mathematical framework for quantifying our perceptual experiences. As we delve further into the realm of sensation and perception, we will explore how these approaches have been used to study specific sensory modalities, such as vision, hearing, touch, taste, and smell.

References

Fechner, G. T. (1860). Elements of psychophysics. https://psychclassics.yorku.ca/Fechner/

Fodor, J. A. (1983). The modularity of mind. MIT Press.

Gibson, J. J. (1979). The ecological approach to visual perception. Houghton Mifflin.

Green, D. M., & Swets, J. A. (1966). Signal detection theory and psychophysics. Wiley.

Gregory, R. L. (1970). The intelligent eye. McGraw-Hill.

Hubel, D. H., & Wiesel, T. N. (1959). Receptive fields of single neurones in the cat's striate cortex. Journal of Physiology, 148(3), 574–591. https://doi.org/10.1113/jphysiol.1959.sp006308

Kanizsa, G. (1976). Subjective contours. Scientific American, 234(4), 48–52. https://doi.org/10.1038/scientificamerican0476-48

Kanwisher, N., McDermott, J., & Chun, M. M. (1997). The fusiform face area: A module in human extrastriate cortex specialized for face perception. Journal of Neuroscience, 17(11), 4302–4311. https://doi.org/10.1523/JNEUROSCI.17-11-04302.1997

Koffka, K. (1935). Principles of Gestalt psychology. Harcourt, Brace. (see also: https://psychclassics.yorku.ca/Koffka/Perception/perception.htm)

Luck, S. J. (2014). An introduction to the event-related potential technique (2nd ed.). MIT Press.

Necker, L. A. (1832). Observations on some remarkable optical phaenomena seen in Switzerland; and on an optical phaenomenon which occurs on viewing a figure of a crystal or geometrical solid. London and Edinburgh Philosophical Magazine and Journal of Science, 1(5), 329–337. https://doi.org/10.1080/14786443208647909

Neisser, U. (1967). Cognitive psychology. Appleton-Century-Crofts.

Posner, M. I., & Raichle, M. E. (1994). Images of mind. Scientific American Library.

Rosenblatt, F. (1958). The perceptron: A probabilistic model for information storage and organization in the brain. Psychological Review, 65(6), 386–408. https://doi.org/10.1037/h0042519

Stanislaw, H., & Todorov, N. (1999). Calculation of signal detection theory measures. Behavior Research Methods, Instruments, & Computers, 31(1), 137–149. https://doi.org/10.3758/BF03207704

Stevens, S. S. (1957). On the psychophysical law. Psychological Review, 64(3), 153–181. https://doi.org/10.1037/h0046162

Weber, E. H. (1996). E. H. Weber on the tactile senses (H. E. Ross & D. J. Murray, Trans.). Academic Press. (Original work published 1834)