Understanding Psychophysics: Methods and Thresholds in Cognitive Psychology

Hello and welcome to the course basics of experimental design for cognitive psychology. I am Dr. Arma from the

department of cognitive science at ID Kpur. We are in week six of uh the course and from this lecture onwards I

am going to be talking about the methodologies that are typically used in experimental research. So in today's

lecture and the lecture for uh the next day I will be talking about psychopysics.

Now what is psychophysics? As you can see psychopysics is basically a combination of two words psychology plus

physics. Psychopysics basically involves the determination of the psychological reaction to physical events. Say for

example if there is a change in the light how do do we psychologically react to it? If there is a change in the

loudness of the sound how do we perceive it? How do we psychologically react to it? Measurement of the psychological

reaction to whatever energy change is happening in uh the environment. That is basically what is the subject of

psychophysics. Boring for example claims that the introduction of techniques to measure the relationship between

internal impressions or psychological reactions and the external world that is the physical world basically marks the

onset of scientific psychology. You know there is a there's a very simple thought experiment that uh you know we do in a

perception class or I'm sure in in several philosophy classes as well that suppose there is some you know there is

a there's a jungle which is isolated and uh nobody is there and somewhere in that jungle a tree falls whether there will

be a sound or not. Okay, now you can pause and think and come back and match your answers. But the answer to this

will be that there will be no sound because sound is basically the psychological or internal impression to

the vibrations that might have been created with that tree falling. If there is no ear there to receive those

vibrations, there is no ear on which the vibrations actually fall and they get registered as a sound, then there is no

sound. All right? So that basically this internal impression of whatever is happening in this outside world whatever

changes in the physical energy that are happening that is called psychopysics. All right.

So the problem of psychophysics actually is slightly interesting. It's slightly paradoxical because it asks us to

objectify our subjective experience. Whatever say is happening in the outside world. Say for example there's a change

in the lighting of this room. How do I perceive that change in the lighting of the room? Now when I'm saying

perceiving, it is basically through my senses, basically my eyes. In this case, I will register that the level of

illumination in the room has changed. That will make me feel something and I will say, oh, I am feeling that there is

dark you know it is dark outside. Obviously the feeling of darkness will correspond to something that is

happening in the physical world as well. When the light increases, I will perceive that the brightness is

increasing. If the light decreases, I will perceive that the brightness is decreasing. This correspondence between

what am I feeling inside and what is the physical change that is happening outside is basically what psychopysics

is trying to measure. Okay. This subjective experience of whatever I am feeling inside or the psychological uh

you know impression that I'm getting is called sensation. All right. Now there are two issues. How do we

measure sensations? We cannot measure sensations. We cannot exactly measure what is how the neurons of the brain for

example are reacting because they are not open to overt measurement. They are basically can be measured by some kind

of reporting some kind of way that I will tell you that oh I am feeling that there is a considerable change in the

illumination of the room. Suppose there are five bulbs that are switched on and suddenly you switch off three bulbs and

now only two bulbs are there. I'm registering the change in the illumination and but how will you know

that I'm registering it unless I tell you if I'm not able to tell you and this translation between what I am feeling

and what I'm telling if that is not perfect you will not be able to know what is uh really whether I felt that

change in illumination or not okay also something that is very interesting here is the internal judgments how dark or

how uh bright I am feeling this outside room to be is not directly corresponding it is not directly identical to the

amount of physical energy changes that are happening in the room and that are affecting my eyes or that are affecting

my sensory apparatus. For example, it can happen that the light in uh you know reduces from 100% to uh 10%. Uh I am uh

perceiving it as much more darker. If the light has changed just from 100% to 95% I am perceiving no change in the uh

light of the room or in the illumination of the room. So the way the internal judgments happen, the way the internal

impression happens of these physical changes is not a onetoone basis and it is not absolutely identical. So

identical changes in physical energy will not call cause identical changes in the uh psychological perception of these

changes. Let's take an example. For example, you have this uh you know amplifier or radio

dial. Suppose you are you know in your car and you're trying to change the volume of the music system of the car.

Now if you're trying to change the uh you know volume of the music system of your car uh remember that it does not

bear a onetoone relation between the movements of the dial. Nowadays you can have a dial or a remote and the

increases in physical energy that your sound equipment is producing. For examp I mean more importantly what needs to

happen here is that the dial will have to be will have had to be calibrated so that its movements increase intensity

proportional to increments in loudness. Loudness is what you are perceiving. Loudness is not the property of the of

the music system. The property of the music system is the increase in energy. How does that increase in energy

correspond to your per perception of loudness? that basically is calibrated uh you know and that is what you are

trying to measure. For example, doubling the volume level on the dial will probably have had to increase the

physical energy about 10 times. So if the output of the sound system has increased 10 times you might only

perceive that okay the volume has just multiplied by two it has just doubled. All right so that is how this uh

relationship really works. Also there is something interesting that the psychophysical relations between the

stimulus and the judgment depend on the particular sensory modality that is being uh tested. For example uh the way

the sensory modality of vision works may be different from the sensory modality of uh audition may be different from the

tactile sensory modality where you are you know manipulating touch or the old factory or the gustatory modality. Okay.

For example, pain judgments in response to increases in electrical intensity of shocks applied to the skin uh can

actually grow much more rapidly than loudness judgments uh you know increase in response to increases in sound

energy. So as I'm increasing let's say the electrical shock by uh 5% and then 10% and then 15% maybe I'm experiencing

that my pain is increasing exponentially on this case in terms of the sound when I'm increasing the energy by 10 times

and by 100 times and by thousand times for me only then it is going to double all right so uh for one shock to be

judged to be twice as painful as another the intensity of the shock will need to have increased by about 1/3. So there is

a larger increase that is required uh there. Actually there's a smaller increase that is required in the shock

and we feel it that oh it has happened much more. Okay. Psychopysics tries to solve this problem

by closely linking perceptual experience uh with physical stimula. The idea is that you use the physical stimulus as

the reference system. So you basically calibrate your experience to certain levels of the physical stimulus. The

stimulus characteristics then are carefully varied uh you know manipulated uh up and down and observers are asked

to report their perception. Say for example uh if you are playing with let's say sound at this point I will go with

the level that has zero sound and I'll ask you to report it and then I'll increase it in some uh steps and then

I'll ask the person to report how they are feeling and I will try and develop I'll try and find out a relationship

between how am I increasing the uh energy in the sound uh instrument and how the person's experience of sound is

increasing. Once I have deciphered that relationship, I will be able to uh estimate what kind of changes in

physical energy here will lead to what kind of changes in the perception of loudness. So the idea in psychopysics is

to formulate a question that is very precise and that is very simple uh and that basically obtains a convincing a

yes no kind of an answer without uh you know having a lot of thought uh you know going in. So the simplest question can

be can you hear the tone at this point or not? You will say yes when you can barely detect the tone. Can you hear can

you tell me the to which tone is being played? You will say yes only when you have recognized oh this is the song that

is being played. All right. Now one of the things that sort of is an

extraneous variable if I may use the word from the uh you know ongoing discussion uh in terms of how we

perceive these signals is the noise in the environment. So for example you're trying to uh play me a sound and I

you're trying to check whether I'm hearing that sound or not can be affected by both external noise there

are other sounds playing in the environment and internal noise my mind is moving here and there and I'm

thinking about any number of things. So there is obviously uh always going to be noise with whatever we are trying to

measure happens in psychophysics as well. Now in such cases the task is one of discrimination of stimulus. So

basically what you're trying to do uh when you're trying to detect a particular stimulus is to discriminate

the stimulus from the noise. Discriminate this particular signal from the ambient noise in the environment.

And this task is always performed under uncertainty because obviously there is a degree of error in the mental sort of

measurement also in the way your neurons are firing and so on. So there is systematic error as well as there is

random error possible in both kinds of measurements. So how do you determine uh whether we

can detect a signal at a given point or we can discriminate a signal from noise? That is basically what is done in

psychophysics and that's what we are trying to do here. The first and the most important concept

here is this concept of threshold. For us to be able to detect any physical stimulus, the stimulus has to be above a

certain threshold. So the idea is mental events will have to be stronger than some critical amount uh in order to be

consciously experienced. So if my if the neurons of my brain are firing in response to uh the changes in the sound

uh and they are say for example they maybe they start firing much earlier much earlier than when they are uh you

know in my conscious experience but for me to be able to report they will have to be uh you know happening at a

particular threshold where I can consciously experience that oh I'm detecting something some kind of lossing

probably is still happening but I will only be able to report once I I have consciously detected that particular

sound that you let's say you're trying to play. Now the most basic function of any

sensory modality or any sensory system that we have you have our five senses is to detect energy changes or detect uh

you know changes of energy in the environment. So how do we hear uh sound? there is changes in the this vibrations

in the air changes in the uh you know air molecules energy that we register in the tempanic membrane of the ear and

then we transduce that back into listening whatever is there. So this energy in some cases can consist of

chemical agents so as in for example taste or smell where there are these particular receptors uh electromagnetic

vision mechanical that happens in audition propriception touch or thermal stimulation. So all of these things that

we uh you know experience through our senses are actually changes in the physical energy of stimula in the

environment. So that is something that is very basic. Now in order to be noticed in order for

us to be able to notice a change in light a change in sound so on the stimulus has to contain at least a

certain level of energy. If it is under that level of energy if it is under that threshold we will not be able to detect

that change. Okay. So this minimal amount of energy that is there you know for us to be able to register uh that

sound or register that amount of light is called absolute threshold. Okay. According to Feckner uh this is

basically uh you know sound makes itself felt when it lifts it sensation over the threshold of consciousness. Let's try

and understand this. This is for example the proportion of responses that are detected. to around

0.5 and stimulus intensity is increasing here in steps at the level of 0.5 when you are you know you till you are not

sure and when it crosses this 0.5 and you can say oh uh yes I detect that particular sound that level is called

the absolute threshold so this is the level of stimulus intensity that you that must be there in that stimulus be

it sound be it light for you to be able to be able to detect it over a level of chance see if there is a two uh you know

is if there is a two-choice decision yes versus no then obviously it has to be above chance for you to be able to say

oh I detected it any number of times out of 10 so here you can see the stimulus intensity is your absolute threshold

let's say number five after which you have been consistently able to say that oh I detected the stimulus so this

becomes an example of absolute threshold now the absolute threshold therefore can the intensity of the stimulus which the

observer can barely detect can basically know okay this is happening. On the other hand there is also the concept of

what is called difference threshold. So the difference threshold basically refers to the minimum intensity by which

a variable comparison stimulus there's a comparison stimulus must deviate from a certain standard stimulus to produce a

perceptual difference. Say for example I've given you two uh balls to hold and I'm asking which one is heavier. Now

suppose this is one uh kilogram and this is 1.05 kg. If the this difference of 05 is able to be registered I can register

this difference of 05 g only then that 05 g will be treated as a difference threshold. Suppose I cannot detect the

difference between let's say this is one this is 1.10 1 0 kg at least 100 g difference that is unless that is

achieved I will not be able to detect the difference that which is heavier so then that will become my difference

threshold so here you can see here is your reference stimulus and here is your

comparison stimulus the difference at which you can actually answer that oh greater this one is greater that

difference is referred to as the difference threshold or the difference limit now how do we determine threshold

so how do we so psychophysics is basically all about How do we you know quantify these uh subjective

experiences? Now, one of the methods to determine threshold is called the method of adjustment. What do we do? The

simplest and the quickest uh way to determine absolute and uh difference thresholds is that you let the subject

uh adjust the stimulus intensity unless it is just noticed or unless until it becomes just noticeable. So that

basically uh will measure the absolute threshold of a particular uh you know sensory modality or it appears to be

just noticeably different from a standard similar. Say for example you have one and you have the other and

you're basically adjusting it to a level that you can just you know 1.1.05 or 1.01 or 1.02. What is the level at which

you can just uh tell that this one is heavier than the other? So how do we do it? Let's go into the

steps. The observer is typically provided with a control of some sort that can be used to adjust let's say the

intensity of the sound until it becomes just audible. So you start from zero and you you basically uh give the uh you

know the participant let's say a remote or a dial that is used to increase sound. you would set it all the way down

to zero and then you ask the uh participant to increase it in particular steps till it becomes uh detectable till

the participant can hear oh this is uh where I can hear it this intensity is recorded to provide an estimate of the

observer's threshold the intensity at which it becomes just detectable alternatively what can happen is that

the observer can adjust the sound from being very clearly a audible so you start at the top and you start reducing

it and you reduce it to the point where it goes out of detection range. So in one case you start from the bottom and

uh uh match around when this is detected and the other case you start from very well audible to and you start reducing

it to the extent where it becomes just not audible. So at at that level somewhere these uh two thresholds will

coincide. Okay. Typically these two kinds of adjustments are alternated several times and the results are

typically average to get a estimate of the threshold. Okay. So this is called method of adjustment. There is another

method that we can help us determine absolute threshold which is called the method of limits. Now a major difference

between the method of limit a method of adjustment and method of limits is that here you're not allowing the observer to

control the stimulus directly. It is the experimental who is adjusting the levels and the experimental is doing it in

discrete steps. So you increase by five another five another five another five and then you basically see where the

participant is going to be able to detect the absolute threshold or is going to be able to register the

difference which is called the difference threshold. So what happens in the method of limits? A single stimulus,

let's say a single light is changed in intensity in successive discrete steps and the observes response to each of

these stimulus presentation is recorded. Increase it to five. Did you detect it? Yes. No. 10. Did you detect it? Yes. No.

15. Did you detect it? Yes. No. And so on. So here just as an example, one can

start when the stimulus is too weak to be detected and then increase the uh intensity in discrete steps. So let's

say every five or every 10 or every 20 whatever the case may be till the stimulus is still visible. Okay. So it

becomes visible sorry or you can start from a clearly visible stimulus and you can start reducing it in discrete steps.

So let's say it is at 100 uh particular 100 units you start going by 95 90 85 80 75 70 and so on. Okay. So the last one

is seen and the uh uh and the last seen and the first not seen that becomes your uh you know uh interval of uncertaintity

which you will measure several times and then you will average out to figure out the absolute threshold. Okay, look at

this here. So in one case you are going in a uh in a way that you are basically uh uh starting from uh 200 reducing it

by 20. So here you get at uh the level of 80. So this is the difference of 20. Here you start with it is not visible

and you increase increase increase increase and at around 140 you get it visible. The third trial you again start

with uh something that clearly visible and go till it is not visible and the trial you also start with when it is not

visible to it is visible and then you get these values. So here it became the average of 180 is 90. Here it is 140 120

so it's 130. Here it is happening at again 140 and 120 is again 130. And here it is happening between uh let's say 100

and 120. So the uh average is 110. You take the average of these measurements that becomes the mean absolute

threshold. Now you can use the same method of limits to uh determine the difference

threshold as well. For example, uh difference threshold as as we've said it is based on relative judgments in which

a constant unchanging comparison stimulus. So the first stimulus or the standard stimulus will be unchanging. So

that is let us say 1 kilogram. Now you have another stimulus whose weight you will vary. So for example you will start

with uh 1.01 1.02 1.03 1.05 05 and so on and then at the point where you can actually say oh I detect that this one

is heavier that will become the difference threshold you can start uh with on the other hand again with

something that oh this is 1 kg and this is uh 1.5 kgs and then you can start reducing till the point when it becomes

oh it is now almost the same so it's the same that you've done the question that you're asking is again going to be very

simple you can ask the question that oh how different must the two stimuli be before they can be reliably

distinguished wished. The traditional way again is to ask the observer to lift pairs of weights. One is constant, the

other is changing. This one you keep changing and to judge if the new weight is heavier or lighter or equal to the

standard weight. Okay, the method is very similar uh you know to the last instance. You can start from a weight

which feels clearly heavier and go till it becomes almost equivalent or you can start with a weight which is ext which

is clearly lighter and go till it becomes almost equal. The upper threshold is the average uh

point at which the observer changes from heavier to equal. The lower threshold is the average point at which the observer

changes from equal to lighter. Uh this basically is called the interval of uh uncertaintity. And the mean of this

upper and lower thresholds is called the point of subjective equality. So here you can see that. So let's say the

standard stimulus is something and you started with uh the one which is heavier. So 350 feels severe, 40 feels

severe, 30 feels severe, 20 feels severe, 310 feels equal. So you basically take the average of 20 and 10

it becomes uh 315. Uh in the other one you basically uh and you can take this lower value also 290 and 280 where it

becomes from equal to uh lighter. So actually here it will be this level. Similarly you start from when it is

considerably lighter to when it becomes equal. So you have this uh mean of 315. Uh the upper threshold is uh 315 and the

lower threshold is 295 and it becomes lighter to equal. So there are two thresholds possible heavier to equal and

lighter to equal. These two thresholds. So the upper threshold is heavier to equal the lower threshold is equal to

lighter. You do that across four measurements. You get the upper threshold four times. You get the lower

threshold four times. You make the subtractions and you basically get the interval of uncertainty to be around 20

g. Okay, here you can see that 310 minus 290 is around 20 g difference is needed. So if the two stimula are different by

at least 20 g only then the participant will be able to detect that one is heavier or lighter. So that is all about

the initial uh you know uh in uh instance into psychopysics. We will continue our discussion on psychophysics

and signal detection in the next lecture. Thank you.