Psychophysics Research Paper

This sample Psychophysics Research Paper is published for educational and informational purposes only. If you need help writing your assignment, please use our research paper writing service and buy a paper on any topic at affordable price. Also check our tips on how to write a research paper, see the lists of psychology research paper topics, and browse research paper examples.

Measurement is an important catalyst for the advancement of science. As measurement techniques are developed, so is the ability of scientists to ask and answer important questions. Nowhere is this observation clearer than in the development of psychophysics. Psychophysics comprises a set of methods that allow for the precise determination of how a person’s internal experience relates to external stimuli. The term psychophysics can be broken into two parts: psycho-, taken from psychology, refers to a person’s internal experience, and physics refers to the features of the natural world to which the person is being exposed—that is, the stimulus. The combined word is thus defined as the measurement of how a person’s experience is related to changes in the physical world. Psychophysics is not just the statement that such measurement is possible; it is also an approach to this problem that makes psychophysics a fundamental advance in the field of psychology. The key to psycho-physics is in how the questions are asked of the observer. It is an old saw to ask if the red you see is the same as the red some other person sees. But in a way, it is this problem of trying to access the ephemeral nature of our experience that is at the heart of psychophysics. Psychophysics does not seek to ask a person to explain his or her experience of red, but to see if different people experience red in regard to the same physical stimuli. This research-paper covers the basic history and methods of psychophysics—traditional methods and the method derived from signal detection theory. Then, I discuss some of the basic psychophysical laws and illustrate the power of psychophysical methods in a few of the many applications that use psychophysics or have been built on psychophysically collected data.

A Very Brief History

The founding of psychophysics dates to 1860, when Gustav Fechner published Elements of Psychophysics (Boring, 1950). In his book, Fechner described the basic methods for measuring psychological experience in relation to physical phenomena. In addition, he proposed that these experiences can be reduced to scientific laws.

The next major advance in psychophysical methods came with work by S. S. Stevens (1956). Psychophysics up to that time emphasized the measurement of thresholds and matches. Important as these techniques have proved to be, they lacked the ability to measure stimuli and differences between stimuli that are easily detected. Stevens, based upon a conversation in an elevator with a colleague, developed magnitude estimation that allows such measurements to be made directly.

Another important development in psychophysics comes with the advent of signal detection theory (Green & Swets, 1966). Signal detection theory is a wholly different approach to how to think about our sensitivity and detection of events in the environment. Whereas many psychologists using psychophysics will happily slide between traditional methods and signal detection theory, they actually propose fundamentally different understandings of how observers detect sensory events.

General Approach

Psychophysics is characterized by a general approach to collecting data. The methods tend to be very repetitive, even tedious, from the perspective of the observer. As in any psychological method, the goal is to elicit responses from the participants that are as clearly interpretable as possible. Given that so many psychological phenomena are mental experiences, this is a difficult goal. Often researchers must validate the results indirectly, much as studies of molecular phenomena must do because atoms and molecules cannot be observed directly. The approach of psychophysics is to ask questions or seek responses that require very simple decisions on the part of the observer. For example, a researcher could present a stimulus and ask the observer if he/she heard the stimulus. In other cases, the researcher might ask the observer to adjust one stimulus until it was identical to a second stimulus.

Given such simple questions and responses, it may be clear what the observer means. Still, the researcher has learned little from any single response. Thus, the experimenter needs to change the stimulus in some fashion and repeat the question(s). Usually there are many trials (questions) in the experiment. Each time, the observer is asked the same question, with the stimulus changing in systematic and often very minor ways. This feature of psychophysical experiments gives them both their power and their repetitive nature.

Basic Measures

Using these procedures, researchers have measured several basic features of psychological experience.

Thresholds

A threshold, in common parlance, is a boundary. There is a threshold at a door that defines the boundary between being in the room and being outside the room. There is a direct analogy between the threshold of a room and a sensory threshold. In sensory terms, the threshold defines a boundary between perception and nonperception. Just as there is an ambiguous point as you enter a room where it is not clear whether you are in or out of the room, there is usually a small range of intensities where people will sometimes report that they perceive and at other times report that they do not perceive the same stimulus.

Absolute Threshold

The absolute threshold is the minimal level of energy that the sensory system reliably detects—for example, the dimmest light or least intense sound that can be detected. Thus, studies measuring absolute thresholds are detection experiments.

Difference Threshold

A difference threshold is the minimal difference between two stimuli that an observer can reliably discriminate. The ability to tell that one sound is just louder than another is a difference threshold or just noticeable difference (JND). Whereas absolute thresholds measure detection, difference thresholds measure discrimination.

Point Of Subjective Equality

The point of subjective equality (PSE) is reached when two stimuli appear identical in some manner to the observer. For example, there is the classic illusion of length in the Muller-Lyer illusion (Figure 20.1). In this illustration, both lines are the same length, but most people say that the figure on the right with the arrowheads pointed in toward the line looks longer. To measure the PSE, the length of the line on the right could be adjusted until it appeared to the observer to be the same length as the line on the left.

Basic Psychophysical Methods

In this section we discuss several of the most basic psychophysical methods.

Traditional Methods

The traditional psychophysical methods comprise some of the most basic and oldest methods that psychological researchers still use. Although most modern studies use more recently developed methods that avoid some of their limitations, researchers still use these methods. They provide a clear description about how psychophysics approaches the questions of measurement, which makes them valuable to cover even if they are not as commonly used as they used to be.

Figure 20.1    The classic Müller-Lyer illusion.

Method of Limits

The method of limits proceeds in a very direct fashion. The researcher preselects a set of stimulus intensities. For example, a researcher is interested in determining the finest pressure that can be felt on the forefinger (i.e., the absolute threshold for pressure on the forefinger). To begin, the researcher selects a set of intensities. A set of five to nine intensities ranging from above to below the threshold that is being measured is most common. The greatest intensity ought to be felt all of the time and the lowest intensity ought never to be felt.

The researcher then presents the stimuli to the observer in an ordered series, called a staircase. In most experiments, the researcher uses two staircases, one going from greatest to least intensity (a descending staircase) and the other going from lowest to greatest intensity (an ascending staircase). In the descending staircase, the greatest intensity stimulus is presented first and the observer is asked if the stimulus is perceived. If the experiment is set up correctly, the observer ought to be able to perceive this first stimulus. Then, the next most intense stimulus is presented and the observer is again asked if the stimulus is perceived. This procedure is repeated with each successively less intense stimulus. Eventually, a point is reached where the observer will report that the stimulus is not perceived. At this point, the next step depends upon the rule used for determining that the stimulus is no longer perceived. Two simple, common rules define the stimulus as no longer able to be perceived: (a) the first time a stimulus is reported as not perceived, or (b) the first time that two successive stimuli are reported as not perceived. The first rule, which could be called the one-stop rule, is useful in situations when the observers have very reliable responses or for rough determinations of the threshold. The second rule, which could be called the two-stop rule, is useful if finer precision is needed. With either rule, the intensity where the observer can no longer detect the stimulus is called a turnaround. The intensity of the turnaround is the stimulus the observer does not detect for the one-stop rule or the first of the two stimuli the observer does not detect for the two-stop rule. After this staircase is completed, the next procedure—usually an ascending staircase starting with the least intense stimulus—is begun.

The ascending staircase is just the opposite of the descending staircase; the researcher uses the same rules only now they refer to the intensities when stimuli are first detected. There are usually several sets of ascending and descending staircases to determine any one threshold. The more staircases the researcher conducts, the more precise the estimate of the threshold. The threshold is defined as the average of all of the turnarounds.

Method of Constant Stimuli

The method of constant stimuli is initially set up just like the method of limits. The researcher selects a set of five to nine stimulus intensities for the experiment. It is desirable that the most intense stimulus be easily detected and the least intense stimulus be rarely detected. The method of constant stimuli differs from the method of limits in the order of the stimulus presentation; in this method the researcher presents the set of stimuli in a random order a set number of times. Using the skin pressure example, each pressure in the stimulus set might be presented 30 times with the order completely random. Each time researchers present a stimulus, they ask the observer if the stimulus was perceived.

After running all of the trials in a method of constant stimuli experiment, the researcher will have collected the percentage of time that each stimulus in the set was detected. A sample set of data for the skin pressure example is shown in Table 20.1 and plotted in Figure 20.2. This function in Figure 20.2 is called a psychometric function, and it represents the change in detectability as stimulus intensity increases. However, it is still necessary to determine a threshold from this data. As you can see, there is no clean step in stimulus intensity from where the stimulus is never detected to where it is always detected. Traditionally, researchers define the stimulus that is detected 50 percent of the time as the threshold. As can be seen in Table 1, it is rare that observers actually will detect any stimulus in the set exactly 50 percent of the time. In that case, it is necessary to use some method of estimating the stimulus intensity that would be detected 50 percent of the time. There are many such methods, the simplest being to do a linear interpolation between the two intensities that were detected just below and just above 50 percent of the time. This method identifies the intensity associated with where the line between the 4th and 5th stimuli in Figure 20.2 crosses the 50 percent level on the y-axis.

Method of Adjustment

Method of adjustment is just as its name implies: The observer adjusts or directs the researcher to adjust the stimulus to the threshold.

Table 20.1       Sample data from a method of constant stimuli experiment

Figure 20.2      Data from Table 20.1 plotted to show how these data can represent a psychometric function of the observer’s responses to all of the stimuli tested.

To use method of adjustment to measure a threshold, the instructions to the observer are to adjust the stimulus until it is just detectable, in the case of an absolute threshold, or just discriminably different from a comparison stimulus, in the case of a JND. However, the method of adjustment is often the method of choice to measure the point of subjective equality. Whereas researchers can use the method of limits and method of constant stimuli to infer the point of subjective equality, with the method of adjustment, the researcher can ask the observer to directly match two stimuli. To take the Muller-Lyer illusion example in Figure 20.1, the researcher can instruct the observer to adjust the length of the line on the right until it matches the length of the line of the left.

Whether measuring a threshold or PSE, the observer repeats the adjustment several times. It is important to have the starting value of the stimulus to be adjusted—for example, the length of the line on the right in the Muller-Lyer example. In this way, the observer must make a fresh adjustment based upon his or her perception instead of making adjustments based on the memory of previous trials. The threshold or PSE is the average of the final settings from each trial.

Limitations of Traditional Psychophysical Methods

The primary criticism of these methods is that they are dependent on the report of the observer. There is no direct way to assess whether the observer’s report is accurate or not. It is possible that observers may seek to make their sensory abilities seem greater than they are and will sometimes report that they detect a stimulus even when they do not. Other observers might be cautious and want to be completely sure that a stimulus is present before they report that they detect it.

The problem remains that these methods rely on the accuracy of the response of the observer. One correction was to put in blank trials. In an absolute threshold experiment, a blank trial would be a trial without a stimulus; in a difference threshold experiment, a blank trial would include two stimuli that are identical. From the response to these blank trials, the results from the actual trials could be adjusted. However, as you will see, researchers sought better ways to deal with this issue. Forced-choice methods and signal detection theory, discussed below, are different ways of dealing with this problem.

Another limitation is that these methods measure only thresholds or PSEs; they cannot allow direct measurement of how perceptible a stimulus is that is easily perceived or how different two stimuli are that are easily discriminated. Magnitude estimation provides a means to make these measurements.

Newer Methods

Forced-Choice Method

In order to obtain a clearer measure of the observer’s sensitivity to a stimulus, the forced-choice method changes the way the researcher structures a trial and the question asked of the observer. I will illustrate a forced-choice method with two alternatives. Using the traditional psychophysical methods to measure the absolute threshold, the researcher presents the stimulus alone and asks the observer if he or she detected the stimulus. The researcher using a forced-choice method has two options: a temporal forced-choice trial or a spatial forced-choice trial. In a temporal forced-choice trial, there are two time periods—Time Period A is shortly followed by Time Period B. The researcher randomly presents the stimulus during one of the two time periods and asks the observer during which time period the stimulus occurred. If the observer is unsure, then he or she must guess; this is the forced feature of the forced-choice method. In a spatial forced-choice trial, there are two locations—for example, the left or right ear in an auditory study. The observer must indicate in which ear the stimulus is presented.

For the rest of the experiment, the forced-choice method can be combined with either the method of limits or the method of constant stimuli. If the forced-choice method is combined with the method of limits, researchers define the turnaround as the moment when the observer guesses incorrectly when or where the stimulus is presented. In the method of constant stimuli, the threshold has to be defined differently. Guessing correctly 50 percent of the time in a forced-choice method is what happens when the observer cannot detect the stimulus at all in a two alternative, forced-choice situation. So, when researchers use a two alternative, forced-choice method combined with the method of constant stimuli, they adopt 75 percent as the value for the threshold.

Magnitude Estimation

The essence of measurement is the assigning of a number to a quantity. For the measurement to be valid, the assignment of the number must have a meaningful relation to the phenomenon under investigation. Temperature assigns a number related to the amount of heat in a system, and the values have proved meaningful. Generally, finding a useful measure is a difficult task. Every now and then, the simplest procedure works to provide useful measures of phenomena of interest. Magnitude estimation, developed by S. S. Stevens (1956), is one such simple method. The psychophysical methods discussed previously have specified elaborate means for determining a measure such as a threshold. In magnitude estimation, the researcher presents a stimulus of a given intensity and observers assign a number that represents how strongly they experience the stimulus. For example, in a measure of perceived intensity of a sound stimulus, a sound is presented and the observer replies with a number that represents how loud that sound is. It does not matter that different observers might use different number ranges; researchers use basic range correction procedures, akin to changing from Fahrenheit to Celsius, to put the observers’ magnitude estimates all in the same range.

However, if the researcher desires to restrict observers’ estimates to a single range of values, it is possible to take one stimulus intensity and make it a standard intensity. This intensity is presented to the observer and given a value, say 100 or 50. This standard intensity is called a modulus. The use of the modulus tends to restrict observers to using more similar ranges of estimates, though researchers often still range correct these estimates (e.g., Silverstein, Krantz, Gomer, Yeh, & Monty, 1990).

Despite the simplicity of this approach, it yields reliable and remarkably consistent results (Krantz, Ballard, & Scher, 1997; Silverstein et al., 1990). Subsequently I will discuss how this method has changed ideas on the perception of stimuli of different magnitudes. Reflecting one of the hallmarks of a good method, magnitude estimation has changed ideas and led to a more flexible understanding of how senses handle stimuli of different magnitudes.

Signal Detection Theory

Signal detection theory, like forced-choice methods, is interested in measuring the sensitivity of the observer to the stimulus separate from any response bias, but it takes a very different approach. In essence, the discussion of thresholds has assumed that the detection of a stimulus, at its basic level, is merely an automatic response to the stimulus. However, anyone who has been in a situation where they have tried to detect a weak stimulus knows that there is nothing automatic about detecting such a stimulus; consider listening for a strange sound a car motor might be making while traveling down the road. It is ambiguous whether that sound is there or not; different people in the car might disagree on the presence of the sound. Even if everyone agrees that the sound is there, it is still possible to disagree about what type of sound it might be and what it might portend about the condition of the car. Detecting a stimulus is not automatic and involves a person’s cognitive abilities.

Signal detection theory agrees with this intuition and describes stimulus detection as a cognitive event. In signal detection theory, the stimulus is called a signal. Although this change in terminology is a bit confusing, it recognizes the history of signal detection theory, which comes to psychology via engineering. In this field, the to-be-detected event often was a phone signal transmitted over phone lines; this terminology carried over to psychology.

Signal Detection Theory Experimental Description and Trial Outcomes

The signal is not presented on every trial. On each trial, the observer’s task is to report whether the stimulus was presented or not. Because the stimulus is not presented all of the time, the observer might or might not be correct. Table 20.2 presents the possible results of any trial in a signal detection theory experiment. First, consider the trials in which the researcher presents the signal. The outcomes of these trials are shown in the left-hand column of Table 20.2. If the signal is presented, the observer may report that the signal was presented. In this case, the observer is correct, and this type of trial is called a hit. Even though the signal is presented, the observer may report that the signal is not presented. This situation is represented in the lower left cell of Table 20.2. The observer’s response is incorrect, and the trial is referred to as a miss. Next, consider the trials in which the signal is not presented. The outcomes of these trials are shown in the right-hand column of Table 20.2. When the signal is not presented, but the observer responds that it was, the trial is called a false alarm. This is the boy-who-cried-wolf type of trial. It is also an error like the miss trials. If the signal is not presented and the observer responds that the signal has not been presented, then the trial is called a correct rejection, as shown in the lower right cell of Table 20.2.

For theoretical reasons, the responses to trials in which the signal is presented and the responses to trials in which the signal is not presented are tracked separately. Thus, the percentages of trials that are hits and the percentage of trials that are misses add up to 100 percent of trials in which the signal is presented. In the same way, the percentage of trials that are false alarms and the percentage of trials that are correct rejections add up to 100 percent of trials in which the signal is not presented. One practical advantage of this method of counting trials is that it is possible to just track hits and false alarms and know how the observer responded to misses and correct rejections. Similarly, it creates fewer numbers that have to be tracked by the researcher.

Table 20.2       Stimulus presentation, participant response, and outcomes in signal detection theory

Underlying Theoretical Explanation for Outcomes in Signal Detection Theory

Signal detection theory begins its explanation for the way people will react in these experiments by arguing that in any perceptual system there is some level of activity, even in the absence of a signal. Sometimes this activity is greater; other times it is lesser; it is not constant. According to signal detection theory, the activity in the perceptual system that occurs when there is no signal is called noise (see Figure 20.3) and varies around a mean level according to a normal distribution. Noise is always present and never goes away, even when the signal is presented.

The signal simply adds a fixed amount of activity to the current level of noise in the perceptual system. For example, if there is no noise in the perceptual system when the signal is being processed, the perceptual system will respond with a level of activity in exactly the same amount as the effect of the signal. If there is a large amount of noise in the perceptual system when the signal is being processed, then the perceptual system will respond with a level of activity equal to the sum of the noise and the effect of the signal. The result is that the effect of the signal on the perceptual system is not fixed but varies exactly to the same degree as the noise, but with a mean level of activity defined by the sum of the mean activity of the noise and the effect of the signal. Researchers call the curve in Figure 20.3 that represents what happens when the stimulus is presented the signal + noise curve.

The distance between the noise curve and the signal + noise curve indicates the sensitivity of the observer to the stimulus. When this distance is measured in the number of standard deviations between the signal + noise curve and the noise curve, this measure is called d’ (Figure 20.3). When the observer’s sensitivity to the signal is not great, the noise curve and the signal + noise curve overlap, as shown in Figure 20.3, and there are some levels of activity in the perceptual system that could be caused by noise alone or by the presence of the signal. Only one of these events can happen at a time. The noise alone, as you will recall, happens on a trial when the signal is not presented and the signal plus noise happens when the signal has been presented. The observer in this experiment does not know which trial has been presented. Still, there are levels of activities that could come from either type of trial. In this circumstance, the observer must decide which type of trial occurred and it cannot be based on any completely objective evidence in the sensory system.

This necessity of making a decision leads to the last feature of signal detection theory, the criterion. The criterion represents a cutoff in the decision making. If the level of activity in the perceptual system is below the criterion, the observer will report that the signal was not presented, whether it was or not. Conversely, if the level of activity of the perceptual system is above the criterion then the observer will report that the signal was presented, again whether it was or not. One possible position for the criterion is shown in Figure 20.3. From this diagram, the four possible outcomes of experimental trials just discussed can be observed.

1 If the signal is not presented and the level of activity in the perceptual system from the noise is below the criterion, the observer will report that the signal is not present. This is a correct rejection.

2 If the signal is not presented and the level of activity in the perceptual system from the noise is above the criterion, the observer will report that the signal is present. This is a false alarm.

3 If the signal is presented and the level of activity in the perceptual system from the signal plus noise is below the criterion, the observer will report that the signal is not present. This is a miss.

4 If the signal is presented and the level of activity in the perceptual system from the signal plus noise is above the criterion, the observer will report that the signal is present. This is a hit.

Figure 20.3     The theoretical underpinnings of signal detection theory.

One important feature of the criterion is that it is set by the person making the observations and can be set at any desired level of activity in the perceptual system. For example, if the observer really wants to maximize the number of hits, the criterion can be set to the left of its position in Figure 20.3. This change will increase the probability that the perceptual system activity will be above the criterion when the signal is present. However, it also will increase the chance for the perceptual system activity from a trial when no signal is present to be above the criterion. As a result, false alarms will also increase. This is called a lax criterion. Conversely, it might be that the observer wishes to minimize false alarms by moving the criterion to the right in Figure 20.3. The rate of hits would go down as well. This is called a strict criterion. According to signal detection theory, then, the proportion of hits and false alarms will depend upon the criterion and sensitivity. One measure of criterion is called beta and is the ratio of the height of the signal + noise curve over the noise curve at the criterion.

Psychophysical Laws

Psychophysical methods have lead to psychophysical laws, summaries of findings from psychophysical studies. We will examine three important laws: Weber’s law, Fechner’s law, and Stevens’s law.

Weber’s Law

Weber’s (pronounced Vaber) law is the oldest and simplest of the laws. Weber’s law deals with the description of the size of a difference threshold relative to the background intensity. Weber’s law is given in Equation 20.1 below.

Equation 20.1     k = ∆I/I

In Equation 20.1, I refers to the intensity of a stimulus. The I alone refers to the intensity of the stimulus against which the change in stimulus intensity is being measured, sometimes called the background intensity. The AI refers to the change in the intensity that reaches the threshold. This measure is the difference threshold or JND. The k is a constant. Translating into English, Equation 20.1 states that the ratio of the difference threshold to the background intensity is a constant. The implication of this law is that observers lose sensitivity to changes in intensity as the background becomes more intense. Not only does sensitivity decrease, but it decreases at a rate proportional to the increase of the background intensity.

Although Weber’s law does not hold for all situations (e.g., very weak and very intense background intensities), it does seem to hold for most situations. Moreover, this law has many important implications. Most people do have an intuitive sense that Weber’s law holds, at least in a general manner. If one candle is lit in a dark room it is easy to perceive the change. But if one additional candle is lit in a room with 500 candles already lit, it is unlikely that anyone would notice the change even though the change in light intensity (AI) is the same in both cases. However, the AI/I in the case of the dark room is very large and very small in the case of the room with 500 candles lit.

Fechner’s Law

Recall that the classical psychophysical methods are mainly useful for measuring thresholds, either absolute or difference. Fechner was interested not only in our threshold perception, where observers are at their limits of ability, but also in the more common suprathreshold perception, where detection or discrimination is easy. Fechner’s law is given in Equation 20.2.

Equation 20.2   S = k x log(I)

In this equation, I stands for the intensity of the stimulus, S stands for the strength of the sensation as experienced by the observer, and k is a constant. In fact, k is the constant from Weber’s law for the given sensory system (e.g., the experience of light intensity). Fechner’s law, by using logarithms, follows along with Weber’s law that we lose sensitivity to all stimuli as their intensity grows. Fechner simply applied Weber’s law about thresholds to all levels of intensities, subthreshold, near threshold, and much greater than threshold. However, there was no direct evidence to support this law, as the psychophysical methods of the day did not directly measure the magnitude of experience of suprathreshold stimuli.

Stevens’s Law

Data from magnitude estimation experiments gave S. S. Stevens direct measures of observers’ suprathreshold experiences. In several cases, Fechner’s law seemed to hold reasonably well, as when observers determined how intensely a light or a sound was experienced. In other cases that Stevens and others tested, however, Fechner’s law did not seem to hold too well, as when observers estimated line length. The relation seems to be more linear than logarithmic. What that means is that if one line is twice as long as another line, it looks twice as long. If perceptions of line length were logarithmic, the longer line would look less than twice as long. In addition, there were other examples in which sensory experience increased even faster than physical intensity. For example, when using low levels of electrical shock (at levels that do not cause harm), if the experimenter doubles the intensity of the stimulus it more than doubles the experienced pain.

Although Fechner’s law seemed adequate in some cases, it is always preferable to seek a more general description that applies to all of the situations described. Stevens found one law that applied to all of the situations so far described (see Equation 20.3).

Equation 20.3  S = cI^b

This is an exponential function. The symbols S and I are the same as in Fechner’s law: strength of sensory experience and intensity of the stimulus, respectively. The symbol c is a constant but has no special meaning. The important symbol that describes the nature of the relation between S and I is b. The exponent b describes the shape of the relation between the physical stimulus and the sensory experience (see Figure 20.4). If b is less than 1, sensory experience increases more slowly than physical intensity, similar to Fechner’s law. In Figure 20.4, which has an example of Fechner’s law on it as a reference, the b < 1 curve becomes ever flatter, without ever reaching an asymptote. The shape is not identical to that for Fechner’s law, though it can be made more similar than is indicated on Figure 20.4, but the similarity of the two types of relations is clear. The perception of light intensity and sound intensity follows this type of relation, which is why Fechner’s law worked reasonably well.

When b =1, Equation 20.3 becomes the equation of a line as shown in Figure 20.4. Thus, the perception of line lengths would follow this form of Stevens’s law. For b > 1, the form of Stevens’s law is the curve that has an ever-increasing slope to it. The perception of pain to electric shock follows this pattern. Given the observation that Stevens’s law can be used to describe a wider range of results, it is generally the preferred law for describing these relations.Figure 20.4      A comparison of the possible relationships between stimulus intensity (I) and sensory experience (S) in both Stevens’s law and Fechner’s law.

Applications

Since their introduction, researchers have applied psychophysical methods to countless practical problems; as a result, many technologies have been impacted.

Development of the Color Matching System (CIE)

One of the major technological developments of the 20th century was the development of color reproduction technologies. The ability to reproduce color in film and print seems standard and everyday, but the development of this technology was a concerted effort during the early part of the 20th century. Although many disciplines play very important roles in this development, at the center is the need to clearly and precisely describe how color mixing and matching happen in the human visual system. Understanding how the human visual system deals with color mixing is vital because it forms the basis of color reproduction technologies. Rather than re-create the exact physical complement of wavelengths in the original object, these technologies produce a set of wavelengths that causes the human observer to perceive the same color (see Chapter 23). Researchers used psychophysical methods, especially the method of adjustment, to define how color mixing occurs in what was defined as the standard color observer. This standard color observer was turned into a mathematical system that allows color matches to be made without the need for a human observer. The first such system was brought out in 1931 by the Commission Internationale de l’Eclairage (CIE) in France, and has been updated since (Silverstein & Merrifield, 1985). This system led quickly to practical applications such as color movies, which were first produced later in the same decade.

Development of Seat Belt Alarms

As cars have become faster and driving more dangerous, the safety of the vehicle has achieved paramount importance. Although new devices are developed every year, seat belts are still the principal safety device in the car, and finding ways to encourage their use has been a concern for many years. One recent method to try to get people to use the seat belt is a high-pitched dinging sound that indicates that the driver is not buckled. The dinging sound is relatively quiet and not too unpleasant to hear. This sound contrasts with the sound some car manufacturers use as a seat belt alarm, which is a very loud, unpleasant noise. The loud noise seemingly makes sense because it is desired that the sound never be covered up by any background noises.

However, psychophysical experiments on how noises mask other sounds revealed that this loud noise was unnecessary and only served to irritate drivers (Zwicker, Flottorp, & Stevens, 1957). Psychophysical experiments have revealed that noises can mask sounds only in the same frequency range, a concept known as a critical band. Background sounds made by car engines tend to be low-frequency sounds. Our voices and any music we might play tend to be of a middle range of frequencies. So by using a high-frequency but relatively low-intensity tone, the sound is unlikely to be covered up by the other noises in the car. For this reason, car manufacturers could abandon the loud, unpleasant sound for one that is far more tolerable.

Cockpit Displays

In the late 1970s and early 1980s, there was a major revolution in airplane instrumentation. Starting with the Boeing 757/767 and the McDonnell-Douglass DC9-80, all of the dials that pilots had been reading were replaced with electronic displays. In these first instances, the displays were CRTs—the same type of device as the standard, deep-back television. There were many engineering and safety goals in putting these devices into the airplane, including reliability and flexibility in what can be displayed, such as radar information. However, there were many barriers of a perceptual nature that had to be overcome to allow the use of these devices.

The main difficulties with these devices are twofold and interrelated. First, CRTs are not nearly as bright as the sun, which means that when flying into the sun, the pilot’s adaptation state is so great, reducing the pilot’s visual sensitivity, that the CRT can be very hard to see. Second, the surface of the CRT reflects a fair amount of light. Because the most powerful light source in the sky is the sun, it tends to wash out everything on the screen. All of the colors become more desaturated (see Chapter 23) and more like white; hence, it is very difficult to tell one color from another. This fact can be quite a problem because the colors mean different things to the pilot. For example, red, amber, and cyan each indicate different levels of problems with the plane (red is for the most severe problems; Silverstein & Merrifield, 1985).

A series of psychophysical experiments conducted by Silverstein, among others, identified the intensity needed by the CRT in all conditions to allow the pilot always to be able to read the screen. In addition, these studies identified a set of six colors that could be distinguished even under the worst conditions. Finally, this psychophysical data allowed engineers to devise an automated mechanism to adjust the intensity of the CRT so that it was at an appropriate light level for all different types of lighting conditions so that the pilot did not have to keep adjusting the display intensity (Silverstein & Merrifield, 1985). This work was recognized with a national award by the Human Factors Society.

Quick Sensory Screening

Psychophysical methods tend to be very time-consuming and are unsuited for health-related screening in normal practice. For example, many states require school systems to screen students for vision and hearing problems. The number of students who have to be tested in a timely fashion requires a faster method of testing than the traditional psychophysical methods allow. So these methods need to be adjusted to allow for a more rapid determination of results. With simplifying the methods, some precision in the results will be lost. For many of these purposes, the loss of precision is acceptable. Here are two examples.

Vision Screening and Getting Glasses

Most people are familiar with having their vision tested; it is required whenever you apply for a driver’s license. These tests screen acuity, the ability to resolve fine details. In particular for driving, the acuity for objects at a distance is tested. Many of these tests are based on a method of screening devised by Dutch ophthalmologist Hermann Snellen in 1843. The chart has letters with carefully calibrated features that, when they are too small, cause the letter to become confused with other letters. The observer reads the chart, usually starting with large letters, and continues until the letters start to become confused. This screening is a variation of a method of limits with a descending staircase. At each level there is some repetition of a stimulus of the same size to gain some greater precision.

Auditory Screening

Auditory screenings, when needed to be done quickly, often use either a variation of a method of limits or a forced-choice method. In the method of limits approach, the tone is decreased in intensity until the child no longer hears the sound. Sometimes the tone starts from a very low intensity and the intensity is increased until the child can hear the sound. Often this staircase is not repeated and the next frequency is tested directly. In this way, a profile of the sensitivity across a range of frequencies can be determined in a relatively short time. Although subtle differences in auditory sensitivity may not be detected, large-scale problems that could affect a student’s performance in school can be discovered.

Sometimes the requirement is only to test a single critical intensity at several frequencies. In this case, a modified forced-choice method is desired. The student, wearing headphones, will be asked in which ear he or she hears the sound. By repeating the trial a few times, it is possible to determine quickly if this tone can be heard.

Summary

Psychophysical methods are not a static set of ways to collect data. As in all of science, the methods used are being adapted and modified constantly. In one direction, methods are being adapted for use in measuring a wider range of perceptual phenomena than the threshold. In another direction, these methods are being adapted for use in wider areas of psychology and are beginning to fulfill Fechner’s ideas when he first proposed this methodology.

Thresholds are important but they represent the limits of sensation and perception. We need to learn more about the sensation and perception that occur during the more normal operation of perceptual systems. Object recognition is an area of intense current research interest—for example, what are the perceptual features of a stimulus that allow it to be recognized as a dog? One example from this research area will illustrate this point. Collin, Therrien, Martin, and Rainville (2006) were interested in how blurring (filtering) images would play a role in the perceptual information needed to correctly identify a face. The filtering of the faces was carefully designed to match some theoretical proposals about how the visual system processes faces. For the current purposes, the method is more important. In this experiment, one face was presented that was filtered. There were four other faces, one of which was the same as the filtered face. The observers in this experiment were to adjust the filtering of the original face until they could correctly identify which of the other four faces it was. The adjustment of the filtering uses the method of adjustment; however, neither a threshold nor a point of subjective equality is really being used. Here, the observer is asked to identify a stimulus in a modification of a forced-choice situation using four alternatives.

Psychophysical methods are also being applied to a wider range of psychological areas and are not as exclusively a method of sensation and perception. For example, cognitive and social psychologies have found use for psychophysical methods. Keeping in mind the topic of facial recognition, it is clear that this issue also is relevant to eyewitness identification. In one study, modifications to the face such as adding eyeglasses or removing a beard reduced accuracy of identification. This reduction in accuracy was measured using d’ from signal detection theory (Terry, 1994).

This research-paper has examined the nature of psychophysical methods. These methods have allowed for the precise measurement of sensory and other psychological phenomena. These methods require careful manipulation of the stimuli with many repetitions of the trial. Despite their longevity, the classical methods still find application (Collin et al., 2006). Signal detection theory represents the greatest departure in approach in psychophysics since its inception. Signal detection theory sees even the basic detection of a sensory stimulus as a cognitive event requiring decisions on the part of the observer. The precision of the psychophysical methods has allowed for application of knowledge about how sensory systems work to be used in many diverse areas. As all methods do, psychophysical methods will continue to develop and be applied to an ever-wider range of sensory phenomena.

References:

1. Boring, E. (1950). A history of experimental psychology (2nd ed.). New York: Appleton-Century-Crofts.
2. Collin, C. A., Therrien, M., Martin, C., & Rainville, S. (2006). Spatial frequency thresholds for face recognition when comparison faces are filtered and unfiltered. Perception & Psychophysics, 68, 879-889.
3. Green, D. M., & Swets, J. A. (1966). Signal detection theory and psychophysics. New York: Wiley.
4. Krantz, J. H., Ballard, J., & Scher, J. (1997). Comparing the results of laboratory and World Wide Web samples on the determinants of female attractiveness. Behavior Research Methods, Instruments, & Computers, 29, 264-269.
5. Silverstein, L. D., Krantz, J. H., Gomer, F. E., Yeh, Y., & Monty, R. W. (1990). The effects of spatial sampling and luminance quantization on the image quality of color matrix displays. Journal of the Optical Society of America, Part A, 7, 1955-1968.
6. Silverstein, L. D., & Merrifield, R. M. (1985). The development and evaluation of color systems for airborne applications (DOT/FAA/PM-85-19). Washington, DC: DOT/FAA.
7. Stevens, S. S. (1956). The direct estimation of sensory magnitudes—loudness. American Journal of Psychology, 69, 1-25.
8. Terry, R. L. (1994). Effects of face transformation on accuracy of recognition. Journal of Social Psychology, 134(4), 483-492.
9. Zwicker, E., Flottorp, G., & Stevens, S. S. (1957). Critical bandwidth in loudness summation. Journal of the Acoustical Society of America, 29, 548-557.

See also:

• Psychology Research Paper Topics
•  Psychology Research Paper Examples

Free research papers are not written to satisfy your specific instructions. You can use our professional writing services to order a custom research paper on any topic and get your high quality paper at affordable price.