Adonis Diaries

Posts Tagged ‘systematic biases

Efficiency of the human cognitive power or mind

Written in March 6, 2006

Cognitively, human is excellent in simple detection tasks or null indicator such as whether a sensation exist or not;.

The mind is fairly good in differentiating the direction of strength of a sensation such as bigger or smaller than a standard, but he is bad in evaluating whether a sensation is twice or three times stronger, and he is worst as a meter for exact measurements.

Human is more accurate in feeling than when relying on his mind. That is why a subject is forced to make a choice between two stimuli rather than responding that the sensation between a standard and a variable stimulus is equally strong.

Man is Not a good observer of complex events;.

Even when viewers are forewarned that they are to see a movie about a crime and that they are to answer questions about details later on, the accuracy of the observers are very low.

The mind is unable to be an objective recorder of the events that transpire because he gets involved in the scene actions.

The mind has a very narrow range of attention and barely can satisfactorily attend to a couple of stimuli.

This observation deficiency is compounded by our sensory differences and illusions. For example, one in sixteen subjects is color blind, many suffer from tone deafness, taste blindness and so on.

The mind does Not think of himself objectively but rather has convictions, feelings, and explanations based on very restricted experiences, hearsays, memories and he tends to generalize and develop a set of beliefs concerning the operation of the mind.

The human usually expects to see and then see what he wants to see and hardly deviates from his beliefs, and sometimes, even when faced with facts.

Many scientists have overlooked obvious data because they clanged to their hypotheses and theories.

Human has to generate an abundance of reliable information and assimilate them before he could eliminate a few systematic biases that he acquired from previous generations and his personal experiences.

This lack of objectivity in human is referred to by the term “common sense”.

The fact is common sense ideas change and are undergoing continual revision, mainly because of the results of research and controlled experimentations and paradigm shifts. away from traditional knowledge

For example, common sense says heavy objects cannot fly until airplanes are common realities.

Common sense says that human cannot see in the dark until infrared goggles have been tested.

Common sense says that it is laughable to use earplugs in order to hear people talking in very noisy backgrounds, until it is experimented and proven to be correct.

The fact that your father or forerunners have always done something in a particular way does not prove that this is the best way of doing it.

The fact that famous people purchase a product from the best known firm does not permit the manufacturer to state that there cannot be very much wrong with the product since the famous people have bought it.

Idiosyncrasy in “experiments”; (Dec. 30, 2009)

Idiosyncrasy or cultural bias related to “common sense” behavior (for example, preferential priorities in choices of values, belief systems, and daily habits) is not restricted among different societies: it can be found within one society, even within what can be defined as “homogeneous restricted communities” ethnically, religiously, common language, gender groups, or professional disciplines.

Most disciplines (scientific or pseudo-scientific) have mushroomed into cults, with particular terminologies and nomenclature:  They want to impress the non-initiated into believing that they have serious well-developed methods or excellent comprehension of a restricted area in sciences.

The initiated on multidisciplinary knowledge recognizes that the methods of any cult are old and even far less precise or developed than perceived; that the terms are not new and there are already analogous terms in other disciplines that are more accurate and far better defined.

Countless experiments have demonstrated various kinds of idiosyncrasies. Thus, this series on idiosyncrasies.  I have already published one on “conjectures” in mathematics.

This article is intended to compare the kind of controlled experiments that are applied by scientists in (natural science), such as physical natural phenomena, engineering… and those developed by scientists dealing with the behavior of people or employing human participants in the experiments (psychology, sociology, economics, or education).

Although the physical sciences, such as all the branches in physics and chemistry…, used controlled experimentation long time ago, in order to develop the huge body of knowledge on the natural phenomena, it was the social and psychological sciences that tried to develop the appropriate and complex statistical modeling packages in order to study the more complex and more varied human behaviors.

It appears that the restricted and countable number of variables in studying the physical nature, and their relative lack of variability with time, did not encourage the physical scientists to contemplate sophisticated statistical models for their controlled experiments, or even to teaching the design of experiments in the engineering curriculum.

Before we expand on the variability of human behaviors it might be more appropriate to analyze the most critical difference in the two sciences. Knowing that any concept is synonymous with the corresponding necessary set of operations in order to be able to measure it scientifically in experiments, we can understand the big leap forward of the body of knowledge in natural sciences compared to the social and psychological sciences.

Whereas the physical scientists can define the concepts of force, moment, power and the like through the relationships of measurable variables based on length, time, and mass the scientists investigating human behaviors have to surmount that hurdle before seriously contemplating to measure human concepts.

Human behavior and the cognitive concepts of attitudes, mental abilities, and moods, problem solving mechanisms, perception, and the like cannot be measured scientifically until sets of operations are agreed on, for each one of these concepts, through the study of human activities or the things that people do while performing a valid task or a set of purposeful tasks.

For example, saying that color blindness is a deficiency that confuses colors will not cut it; what is needed are a set of instances that could define this illness such as what exactly are the colors of the spectrum with mixtures of two primary colors can a “protanope” (color blind individual) match that are different from normal people, he will confuse a blue-green color with white or gray, will confuse red, orange, yellow, yellow-green, and green when suitable brightness and saturation of these colors are used, and has reduced visibility in the red end of the spectrum.

Two decades ago the air force in the USA contracted out groups of psychologists and human factors professionals to specifically establish a set of operations that could be submitted to potential airplane fighters to measure and evaluate their capabilities for the mental and perception workload needed for the job.

This set of ten or twelve operations measuring short term memory capacity, reaction times, computational abilities, attention span, and types of errors committed in each operation is the kind of hurdles that the study of human behavior have to surmount.

The operation measurements of a single human concept may be a life project for a group of scientists that require secure and continuing funding from concerned parties who have vested interests in thorough study of the concept.  It is obvious that a few human concepts will enjoy deeper and more complete investigations than others.

Maybe because the physical scientists did not face the problems of establishing sets of operations that the method of controlled experimentation was not deemed essential enough to rigorously teach in high school programs, and ultimately failed to initiate the students to the experimental methods.

Social sciences made significant in-roads into the educational programs in the last decade.  This lack of early initiation of students to experimental methodology might also be the main reason why rational thinking and the experimental mind is not that widespread throughout all societies and are just confined to the privileged who could afford higher education at select universities.

Physical scientists rely on equipment to “objectively” observe and measure, and the more the equipment are precise the more accurate are the data.  Scientists of human behavior have to rely on people’s responses and observations.

It has been proven that man is Not a good observer of complex events; even when viewers are forewarned that they are to see a movie about a crime, and that they are to answer questions about details later on the accuracy of the observation, subjects were discovered not to be that accurate.

Man is unable to be an objective recorder of the events that transpire because he gets involved in the scene actions.  Man has a very narrow range of attention and barely can satisfactorily attend to a couple of stimuli. This observation deficiency is compounded by our sensory differences and illusions; for example, one in sixteen is color blind, many suffer from tone deafness, taste blindness and so on.

Man does not think of himself objectively but rather has convictions, feelings, and explanations based on very restricted experiences, hearsay  memories and he tends to generalize and develop a set of beliefs concerning the operation of the mind (idiosyncrasies).

Man usually expects to see, and then see what he wants to see, and hardly deviates from his beliefs, even when faced with facts.  For example, many scientists have overlooked obvious data because they clanged to their hypotheses and theories.

Man has to generate an abundance of reliable information and assimilate them before he could eliminate a few systematic biases that he acquired from previous generations and his personal experiences.  Consequently, experimenting with people is more complex and more difficult than the physical scientists or engineers have to cope with.

First, there are no design drawings for people’s mind and behavior as engineers are familiar with because the structure of human organisms is approximately delineated and the mechanisms are imperfectly understood.

Second, people vastly differ in anthropocentric dimensions, cognitive abilities, sensory capabilities, motor abilities, personalities, and attitudes.  Thus, the challenge of variability is different from physics where phenomena behave in stable fashions, are countable, and can be controlled with minimal management.

Third, people change with time; they change in dimensions, abilities and skills as well as from moment to moment attributable to boredom, fatigue, lapse of attention, interactions among people and with the environment.  People deficiencies in senses, physical abilities and cognitive capabilities changes with time and thus, the techniques of selecting subjects have to account for the differences in age, gender, specific deficiencies, training, educational levels, communication skills, and incentives to participate in an experiment.

Fourth, the world is constantly changing and systems used by people are changing accordingly.  Thus, interfaces for designing jobs, operations and environment have to be revisited frequently to account for new behavior and trends.

Fifth, everyone feels is an expert about human behavior on the basis of common sense acquired from life and specific experiences and we tend to generalize our feelings to all kinds of human behaviors but not so expert in the fundamentals of natural sciences such as physics or chemistry.

We think that we have convictions concerning the effects of sleep, dreams, age, and fatigue; we believe that we are rather good judges of people’s motives, we have explanations for people’s good memories and abilities, and we have strong positions on the relative influence of nature and nurture in shaping people’s behavior.  Consequently, the expertise of psychologists and human factors professionals are not viewed as based on science.

Six, physical scientists may enjoy the possibility of “testing to destruction” of prototypes or the materials under study, a luxury that experiments on people forbid or are impossible to do outside the safety range allowed by moral standards, laws, and regulations.  Research on people has to circumvent this major difficulty by using dummies, animals, or willing subjects thoroughly aware and educated to the dangers of the procedures.

Seventh, research on people is regulated by privacy laws and concepts such as consciousness, mental images, fatigue, and motives are highly personal experiences and not open to public inspection while science must be a public affair and repeatable by other researchers.

Consequently, human and social sciences developed terminologies that natural scientists cannot comprehend.  For a experimental natural scientists a variable is a variable.  What is on the left hand side of an equation is the data and what are on the right hand sides are variables and coefficients.

For social scientist you have dependent variables (data), independent variables (factors, manipulated variables, within group variables, between group variables, confounding variables, control variables, treatment variables, sub-group variables, and on).

Controlling an experiment in social sciences is a major project that requires months in preparations to eliminate biases related to people selections and material used by the subjects and the experimenter.

Social sciences have developed many “sophisticated” statistical analyses packages and each discipline prefers its own set of “experimental design” because the members are familiar with the interpretation of results and not because the experiments are pertinent or useful for practical usage.

Multidisciplinary studies are important for a university student to get clear on the many idiosyncrasies of disciplines and start reflecting seriously on what is objective, what is experiment, how valid are research results, how biased are research, and how to correctly interpret results and read scientific studies.

Producing a good reflecting “scientist’ is not an easy task; we are not searching for the appropriate equation but for a good formed scientific and empirical mind. Courses in experimental designs are fundamental even for philosophy students, especially in religious schools.

“Psycho-physical methods in Human Factors”

Article #37 in human factors in engineering, (Written in March 29, 2006) 

In the 19th century, a few physiology and neurophysiology scientists such as Gustav Theodor Fechner (1860), von Helmholtz, and Weber endeavored to study the functional relations between body and mind which resulted in the development of controlled experiments for reaching common reference points that would relate the psychological feelings of the senses to their corresponding physical measurements.

For example, the relationship of the perception of brightness and color versus amplitude and frequency of radiant energy waves or the feelings of loudness and pitch versus amplitude and frequency of mechanical or sound waves.

Fechner demonstrated that the biology of the human does record the variations of the environment by attenuating them to logarithmic values.

The sound intensity of 10 violins in an orchestra is actually heard as one violin. If the conductor wished the audience to hear the sound of ten violins then he will have to increase the number of violinists to one hundred.

If the cord of an instrument vibrates at 80 Hz, then if the cord is pinched in the middle it will vibrate at 160 Hz, at 1/4 it will vibrate at 320 Hz and so on.

Thus, the pitch sensation and identification of octaves in a sound is attenuated logarithmically too.

As the physical stimulant increases geometrically, the biology of the human senses the increase as an arithmetic progression.

Weber discovered that the ratio of the differential threshold to the initial stimulant for the subject to sense a modification in the variation is a constant.

In the case of the pitch the constant is 1.0594; so that for a frequency of 440 Hz (the first La in the octave), then the next frequency to be differentiated would be 440 * 1.0594 = 466 Hz, the next frequency to be sensed would be 494 Hz.

The absolute threshold or the minimal excitation to hear a pitch is 3 Hz.

The psychophysical method have been applied extensively in telephone, TV, radar, illuminating, sound, and geophysical industries who wanted to provide the end users with a satisfactory psychological comfort level without undue expenditure on top notch physical circuitry or infrastructures that would not add to the requisite human performance and capabilities.

Recently, there have been attempts to adapt this method to studying complex cognitive concepts such as attitudes, mental abilities, reliable speech quality, and naturalness of voice in telephone by measuring these concepts.

The measurements of interest are:

1. the absolute thresholds (limens) that mark the boundary between sensation and no sensation,

2. the differential thresholds or the smallest detectable change in sensation corresponding to the amount of changes in the physical stimuli,

3. the equality effects values in sensations produced by two different kinds of physical stimuli like when you filter a green color to make feel as bright as a red color.

Subsequently, whenever stimulus is mentioned it means a physical value and response of a subject refers to his answer for a psychological sensation.

The psychophysical method capitalizes on the capabilities of human:

1. being excellent in simple detection tasks or null indicator or whether a sensation exist or not;

2. being good in differentiating the direction of strength of a sensation such as bigger or smaller than a standard,

3. human is bad is evaluating whether a sensation is twice or three times stronger,

4. he is worst as a meter for exact measurements;

5. human is more accurate in feeling than when relying on his mind and that is why a subject is forced to make a choice between two stimuli rather than responding that the sensation is equally strong.

For example, it is harder to hear speech in a noisy environment than in a quiet one;  and the problem is to investigate how loud should a person raise his voice in order to be heard in a noisy room?

The development of this method generated three basic procedures for measuring the equivalence between psychological sensations and the physical parameters which are:

1. Procedure of adjustment or average error,

2. procedure of limits or just-noticeable difference, and

3. the procedure of constants which is the most accurate and widely applicable.

Consequently, we will focus on the third procedure in the subsequent sections.

The procedures of adjustment and of limits are slightly different but generally the experimenter shows the subject two series of stimuli,

The first series starts with a stimulus well above threshold and the subsequent values diminish by small and equal steps; the second series starts well below threshold and values are increased in equal amounts.  The final responses of the subject are averaged whether the response is an indicator of a sensation or expressing direction of strength such as more than, equal, or less than.  The experimenter either takes control of varying the stimuli or he may allow the subject to control the stimuli until he reaches a definite sensation response.

The procedure of constants starts with the experimenter performing an exploratory study to determine the upper and lowest thresholds and then selects between 4 to 7 stimuli spanning the range of sensations.

These selected stimuli are everytime randomized when presented more than 50 times to the same subject.  If equality or differential thresholds are being measured then a standard value is compared to the subject responses.

Usually, the more intense the stimulus the higher the percentage of correct identification and the choice of threshold criterion vary between 100% correct answers and 50% which basically represents chance happening depending on the behavior of responses or the graph trend and shape; the stimulus value adopted corresponds frequently to a threshold criterion of 75% correct.

An example is in order.

A telephone company is planning to provide TV viewers with signals transmission that give sharp images without the need to invest in expensive top notch circuit infrastructure.

The bandwidth for transmission should be more than adequate for a satisfactory picture; a bandwidth of 4 megacycles transmits blurred images while 10 megacycles is more than excellent.

A cheaper simulator is achieved by using an optical defocusing of movement of the projection lens corresponding for the range of zero to 10 mils.

An observer views a standard scenes compared to an identical but partly blurred scene. The amounts of blurring are set up in steps of 2 mils and the subjects are forced to guess which scene is the sharper.

By selecting a threshold criterion of 75% correctly guessed the company was able to decide on an effective bandwidth of less than 7 megacycles that provide an excellent sharpness psychologically and save on irrelevant costly hardware.

There are many pitfalls for inexperienced investigators conducting psychophysical experiments and many of these problems are the results of systematic biases such as:

1. Observers tend to quickly receive cues from the control devices they are handling and their responses become biased;

2. the space order of the standard and variable stimuli can be a source for error biases, for example, it is well known that a vertical line looks longer than an identical horizontal line;

3. the second of two equal stimuli is judged louder or brighter than the first;

4. a series of “yes” responses for example encourages a subject to persist in the same kind of response, thus the experimenter should vary the direction of responses;

5. observers tend to get cues from the experimenter through his voice, gestures, or instructions.

Most of these errors can be counterbalanced when the order of stimuli presented is randomized at each trial.

The psychophysical method is being used in the study of sensory performance of non-human species as well.


adonis49

adonis49

adonis49

October 2020
M T W T F S S
 1234
567891011
12131415161718
19202122232425
262728293031  

Blog Stats

  • 1,427,446 hits

Enter your email address to subscribe to this blog and receive notifications of new posts by email.adonisbouh@gmail.com

Join 774 other followers

%d bloggers like this: