Foundations and Evolution of Scientific Method in Cognitive Psychology

Hello and welcome to the course basics of experiment design for cognitive psychologists. I am Dr. Ark Warma from

the department of cognitive science at IIT Kpur. As you've seen in the introduction lecture, we're going to

talk about various aspects of experimental design and the quantitative method in cognitive psychology in this

course. But before I start with those things, uh I thought it might be a good idea to basically give us a background

uh give us a bit of a uh context about what experimental design is all about. Where do the principles of experimental

design derive from and related facts that are important not only in terms of measurement in uh you know in psychology

or cognitive psychology but also generally. So we will discuss in this and a couple of other lectures about the

basic principles of what basically makes the scientific method where are these principles coming from and so on.

So this week is going to be a little bit about a brief historical background and some of the things that I just

mentioned. Uh let's first begin uh talking about the scientific method. All of us have uh you know heard this term

the scientific method a number of times and some of us probably are still wondering what this scientific method is

all about. Is it just collecting data uh coming up with observations and basically saying okay this is correct

this is not and so on. Let's talk a little bit about where this this scientific method really come from. Now,

uh the advent of psychology as a scientific discipline uh carries a lot of stories about uh you know the

internal struggles amongst the psychologists themselves uh to basically adopt a particular kind of method to uh

you know to get themselves categorized as a science. One of the initial anxieties of psychologists probably has

been uh you know this thing of calling themsel as a genuine bonafide science much like let's say the physics and the

chemistry and slightly different from uh you know only only philosophical or analytical traditions. So this is

something that has occupied the minds of scientists or psychologists a lot and let us try and see where is this really

coming from. Now what do you call science? What do we say? For example, if we were to talk

about basic definitions, where do we sort of uh you know draw a line and say, "Oh, this is scientific. This is what is

called science and this is not." So, traditionally, uh the scientific method or science can be looked upon as a

continuous accumulation of facts based on objective and importantly valuefree observation and experimentation. So all

of the knowledge that we are gaining through you know objective methods of observation which are valuef free which

are not you know afflicted by our judgments of what is right and what is wrong what do we like what do we not

like what aligns with our ideology what does not align with our ideology science in some sense uh your or in that sense

you can see is a sterile observation of facts facts which can be later verified and woven in form of a theoretical

framework. So let us say that is a loose definition of science that we will start with. Now interestingly uh this uh you

know preoccupation with a definition of science or what scientific method is all about uh is broadly if you see uh you

know absent with a lot of scientists themselves they prefer talking more about what they do rather than you know

dwelling about the underlying principles. More often than not the underlying principles of scient

scientific inquiry are taken for granted. It is basically assumed that we are all following this scientific

tradition, the tradition of the particular method, let us say the hypothetical deductive model or uh you

know hypothesis testing and so on. But it's rather seldom uh and probably only in courses of uh you know uh research

methods and research designs that we wonder and we talk about the underlying principles of this scientific method.

I'll make an attempt to very briefly cover some of these things as we go ahead in this week.

Now uh where does this all start? You know as with most disciplines it starts uh you know in ancient Greek where uh

there were thinkers and you know scientists like plateau and Aristotle. So plateau basically says that reasoning

or uh you know gaining of knowledge is something that is innate. It is it comes within us. It is something that you know

we are endowed with uh to a certain extent. So Plato thinks that knowledge is endowed by the soul. Okay. Uh when

the soul enters the body, it brings with itself some kind of knowledge, some kind of uh you know the ability to

distinguish between what is correct and what is incorrect. In that respect, scientific knowledge or

any knowledge for that matter does not really require careful and objective observation. Rather, it rests within uh

you know innate reasoning and endowed intelligence. We are endowed intelligent beings. So we should know what is right,

what is not. We we do not need to scrutinize what we know. We do not need to verify what we think is correct.

Rather the tendency is to believe that okay uh we are endowed beings children of uh you know the creator and we know

what is correct and what is not. There are other ma methods. For example, however, uh say for example, people like

Aristotle have distinguished between different modes of arriving at knowledge or reasoning. Let's let's look at that

for a little bit. For example, Aristotle distinguishes between uh you know deductive and inductive reasoning. What

is deductive reasoning? Deductive reasoning basically starts from indisputable principles. Things that you

have observed yourself. Let us say firsthand observations uh which we know that are true conclusions. These are new

maybe these are true conclusions and new facts can be derived uh from these observations following the rules of

logic. For example consider the statement all humans are mortal. Uh now you start with all humans

are mortal. Rajes is human. What will be the conclusion B? The conclusion will be Rajes is also model. So basically if you

take premises A and B and you follow the correct principle of logical reasoning premise C or basically the conclusion C

is automatically drawn from premises A and B. That is uh an example sort of a crude example for deductive reasoning.

What is inductive reasoning? Inductive reasoning basically starts from a series of converging observations, lots of

observation picked over time. Uh and these observations are used to arrive at a general conclusion. It is typically uh

you know one of the dominant methods in science to arrive at these scientific laws using observed phenomena. We'll

talk about this in more detail in the next few lectures. Take for example this uh these statements.

The sun has risen in the east every day of my life. So I know that to be true. The sun rose in the east yesterday. The

sun rose in the uh east day before probably today itself. Uh so we have this third statement. The sun rose in

the east today. Now if we have this you know bundle of observations we can conclude from this bundle of

observations that the sun rises in the east every day. So on the basis of these specific observations that the sun has

risen in the east every day of my life. The sun has risen in the uh east yesterday and the day before and even

today the sun has risen in the east. So we can basically put these three or four observations together and we can

conclude from this that the sun rises in the east every day. All right. So this is again uh on the basis of certain

observations we have arrived at a conclusion and we assume that conclusion is correct.

Take another example. Uh for instance in a set of clinical trials let's say A uh 90% of the patients who took drug A

recover from the disease Z. So you know there was a set of clinical trials a particular drug was administered and 90%

of the patients who uh you know took this drug uh you know drug A recovered from whatever disease uh you know we are

calling as Z. Another another statement is in study B 80% of the patients reported a reduction in the symptoms of

the disease Z uh when given drug A. Let's look at another study. In a third study, 85% of a diverse group of

patients who were given the drug A experienced relief from the ailments accompanying disease Z. So now across a

set of three different studies, maybe conducted in different parts of the world by different scientists by

different research groups, we have some evidence that says when the drug A is administered, people experience relief

from the disease Z. So on the basis of putting together these three different studies, we can actually conclude that

drug A is likely an effective drug for treating patients with disease Z. All right. So this is another example of how

uh inductive reasoning may be used to come up with scientific conclusions. Now uh notice the subtle differences in

the two. In both cases, we have premises and we have conclusions. But notice there's something interesting happening

in deductive reasoning. In directive reasoning, the conclusions are guaranteed to be true if the logical

premises if the logical rules are correct logical rules are followed. Inductive reasoning however does not

guarantee that whatever conclusions you are you know taking out are basically correct because it depends on a number

of factors. It depends upon the uh quality of observations. It depends upon the methods being followed and so on.

All right. So inductive reasoning therefore is more steeped in rationalism according to which reality can be known

by reasoning. Now here we are reasoning uh driving from innate knowledge. Uh and it's basically most uh you know uh used

uh you know in more common deductions in daily life. uh the conclusions and the generalizations that we are sort of uh

you know uh building upon uh in case of inductive reasoning are also drawn from a limited set of observations. Say for

example if one were to say that you know uh uh all swans are white. There is no way to you know get and measure you know

or say for example come across all the swans that are present in this world. We see one son, we see two, we see 10, we

see 20. And the basis of this we basically concluded okay probably all swans are white. Something like that.

Okay. This is the kind of reasoning we also adopt in daily life when we notice the similarities between events and

experiences and on the basis of those uh observations we formulate general principles about how the world works.

All right. Such reasoning is typically emphasized in empiricism according to which induct according to which

knowledge is based on observation and experimentation. So you observe something you experiment maybe formulate

some hypothesis test them and basically come up with knowledge uh to explain whatever observed phenomena are there.

Now for the most part both inductive and deductive methods of reasoning have been used as sources of knowledge. So they

are both uh well accepted well practiced uh you know ways of getting at knowledge. Uh but if we carefully

compare the two modes of gaining knowledge deductive and inductive reasoning it seems that inductive

reasoning may underly the nature of scientific thought uh and the scientific method. It was championed however say

for example by British scholar Francis Bacon who opined that knowledge mainly should be based on careful empirical

observations and not on authority. See a lot of times uh the premises that are uh that form the uh you know basis or

foundation of deductive reasoning or deductive knowledge are based on authority uh authority of say for

example somebody who is considered an expert somebody who's considered say for example in cases like say for example

engineering so there are godmen there are holy books you say this is how things are and then you have to accept

them Francis Bacon and people like him basically said that knowledge should be based not on authority but on careful

empirical observation. Began advocated that we should document experimental histories in a systematic

manner which are systematic observations where critical elements could be actively manipulated and their effect

studies. So if you have a premise, if you have an observation, if you want to take out a conclusion, before taking out

a conclusion, you manipulate things. You basically look at their effects before you derive or before you arrive at some

kind of conclusions about these things. This tradition of where does knowledge come from and what are the correct

sources of knowledge is basically what is the broad subject matter of the subject you know of this field called

philosophy of science. Now uh in early 20th century philosophers basically you know uh shifted their attention from

metaphysical questions to questions about the nature and the uh you know the uh methodology uh of the scientific

approach. This was uh taken up by several scholars in Vienna around that time in the you know early 19th you know

late 19th and early 20th century. uh people like Morris Schlick uh and others like Rudolfph Carnap and Ludwig

Vidgenstein uh you know among others which were together uh you referred to as the Vienna circle. This movement that

starts roughly around this time is referred to as the logical positivist movement. Let's talk a little bit about

this to understand the context and the history of the scientific uh method. Now what is logical positivism? The

logical part of this definition or the logical part of this term logical positivism derives from the

preoccupation of the times basically saying that you know there is this preoccupation with language and there is

this preoccupation with where does meaning come from. So logic is supposed to be a method using which we derive

meaning and therefore the logical part of this logical positivism uh comes from uh positivistic or the positivism part

of it basically derives from the high esteem uh that was associated with the fact that we are talking about

scientific knowledge and in some extent uh you know to some extent about the arrogance related to scientific

knowledge that scientific knowledge is the best or the uh only true form of knowledge. So this is where the movement

is termed logical positivism. The way of deriving knowledge and the fact that this knowledge that we are deriving

using scientific method or using logical reasoning is the correct and the truest form of knowledge. Now there is a bunch

of things within this logical positivist movement that are interesting and they basically contribute to the evolution of

scientific thought. Let us pay attention to that a little bit. One of the things that logical positivists uh you know

took uh in their hands or uh you know were really concerned with was uh this uh idea of demarcation of science. So

they're basically uh concerned uh a little bit about setting scientific knowledge apart from non-scientific

knowledge. So for example what is non-scientific knowledge? Knowledge that is derived from tradition, from common

sense, from belief, faith etc. How do you distinguish this from scientific knowledge? Why is scientific knowledge

superior or special as opposed to the knowledge that is derived from our beliefs or our faith or even say by the

method of authority and so on. So they wanted this definition to be very clear-cut. They wanted to be able to say

that okay this is where scientific knowledge starts or this is where common sense ends and scientific knowledge

begins. All right. Now logical positivists postulated that science uh follows through a circle of

observation, induction and verification. The underlying assumption is that there is a direct uh you know correspondence

between reality, how the world is and human perception. Our ability to observe how the world is. The idea somehow is

that uh our means of observing how the world works are in some sense infallible and once we go through that once we

observe carefully note down carefully uh you know analyze the workings of the world we will understand how the world

works better and that will be the best form of knowledge. So once you have these observations what

do you do? The next step would be to translate these individual observations into general conclusions some using

inductive reasoning uh basically which will tell us about how the world works. Okay. So for example something like how

Newton would have arrived at the mathematical laws of physics that eventually could explain you know the

laws of motion that Newton arrived at. Okay. So uh say for example uh we talk about this thing about the apple falling

on uh you know Newton's head and what does it basically indicate and how does he start from there and reach at the

laws of gravity and so on. So things like this this is basically the initial ideas of how scientific knowledge must

be generated. The next step in this once you have made observations because method of

observation is considered sacrosen you've observed a bunch of phenomena you have uh translated these observations

into general uh frameworks or uh you know you can say theories that predict how the world works. What is the next

and the final step? The final and the most important step would be that you should be able to verify these general

conclusions that one has arrived at. So the positivists claimed that experimental verification was one of the

important demarcation criteria of scientific knowledge. Scientific knowledge will be built upon knowledge

or observation of the facts that can be verified to be true. How will they be verified? They will be verified through

experimentation. So you come up with this phenomena. Somebody says this is how things are. You go out and verify

it. Once you have been able to verify it, that will only then constitute scientific knowledge.

Okay. And again how do you verify it? You verify anything as true or false on the basis of valuefree sterile

observation. Okay. So there is a degree of uh you know circularity here which we will

address going forward. Now uh the logical positivists also accepted deductive reasoning based on the

principles of logic as a way of making meaningful statements but not really as a means of generating new knowledge.

Okay. So uh according to the positivists or according in the tradition of this logical positivist movement, the idea is

that deductive reasoning uh which is based on premises and follows uh you know logical reasoning and comes up with

some conclusions is good for uh you know understanding whatever existing knowledge is there but it is not uh uh

you know the preferred way to generating new knowledge. Okay, for that maybe inductive reasoning would be the

preferred method. Now this is a little bit about the positivist movement. I'm not really

going into an extreme detail but I just want to give you a flavor of where things started from. Okay. So let's look

at some of the possible critiques of this movement to understand where do things go from here. Now while the

positivists were uh rather firm with their view of the demarcation criteria uh if you zoom in if you look a bit

closely uh there could be more nuances to how scientific knowledge must be generated. Uh let's let's look at some

of these uh things. Now scientific knowledge does not automatically proceed from perception to understanding and

even verifiable facts for example may lead to erroneous conclusion. This is something that we should keep in mind

and we'll come across several uh examples along the way to you know reiterate this fact. Let's take some

specific criteria. For example, perception cannot be the final destination. Uh logical positivist

basically assume that the facts could be perceived independently of any theoretical framework. See what

perception or observation of the world tells us are isolated unconnected facts. And you observe one thing the other day,

the uh other fact the other day, the uh completely unrelated fact the other day. It is important that we are able to

connect these dots. It is important this this scientific knowledge that we're talking about is structured in in in a

particular way. Otherwise, what we will be doing is we will be having random observations across a particular time

period and we will say okay this is correct and this is correct and this is correct. But we will not have a

framework to connect these dots and to evolve a particular framework within scientific knowledge. function. So the

logical positivists assume that facts uh could be perceived independently uh uh you know of any theoretical

framework and would be the same for all observers. However, it has been pointed out that this may not be the case.

Typically when you start observing uh you know things uh there is a broader frame framework under which you will

weave or you will be able to connect all these observations. This is slightly different to what the logical

positivists were assuming in the beginning. For example, researchers typically

observe huh and this is also something that researchers typically observe reality in fragments. You see something

here, something there uh which you know they must try to you know stitch together and build a theoretical

framework and interpret the facts within that context. So for example, if you have a theory about how you know the

planetary system works or you have a theory about how gravity works or how motion works or say for example as we

are psychologists you have a theory about say for example why do people suffer from personality disorders or say

for example what causes autism anything for that matter you can have certain observations but those observations have

to be seen within a given framework they are not disconnected isolated disorganized facts so you need to be

able to put them together is what the idea is okay. Look at this picture for example.

At this point, if you're seeing this picture, uh you will just see that okay, these are just uh you know fragmented

aspects of you know black and white patterns but they're not making any sense. How will they make sense?

If we have a theory of something, if we have a framework within which we can start uh looking uh towards a particular

pattern, all of these observations that we are finding uh from point A to point B for example or at different points of

time and different uh places, how do we stitch them together? How do we start seeing a pattern? We will not see a

pattern unless we have a framework to look into uh these findings. So theories are very important and a theoretical

framework shall provide an element of interpretation to the observed facts. Now you will start say for example you

have a particular theory of motion or you have a particular theory of uh you know how the planetary system works.

Once you have that theory in place you will now start interpreting these isolated findings from the lenses of

this theory. And some of some will sort of go together, some will not. But it will basically give you a way uh in

which to interpret whatever isolated findings or disconnected findings you are observing.

Also, having a theory helps uh in defining the focal points of the inquiry as well. For example, theories typically

allow scientists to search and research with focus and specific directions. For example, if you have a theory about how

the uh you know the earth and the other planets revolve around the sun, uh this particular uh theory will give you

questions will give you uh you know focus as to what are the important questions underlying this theory and how

will we sort of investigate uh the uh you know findings and phenomena within the uh ambit of this particular theory.

So facts and observations cannot stand alone. They are not isolated. They are not disconnected. They need to be

organized and and stitched together in some form of a framework which we can typically uh you know at least at this

point refer to as a theory. All right. So uh let's take some examples uh uh you know

uh about this uh Brisbane and Russell in in their book uh you know uh they talk about this example uh with you know the

famous astronomer Galileo Galilee they says uh they say Galio initially proposed a method to measure the size of

stars. All right wherein he positioned himself relative to a chord that just blocked a particular star out of uh out

of sight. He assumed that on the basis of the thickness of whatever cord that is so a piece of thread uh and the

distance of the eye from the chord, he will be able to calculate the visual angle subended by the chord and thus the

visual angle subended by the star which can eventually be used to determine the overall size of the star. That is the

idea. That is what he started with. However, at this point he did not realize that the apparent size of the

star appears much bigger than they actually you know that the apparent size of the star appears much bigger than

they are because of the atmospheric scatter of the light rays uh you know the light rays coming. So the star might

be small but because the light coming from the star is dispersed through the atmosphere it may appear much bigger

than it actually is. and also the fact that stars are at different distances from the earth. So the relative

calculation of size must take into account the distance at which a particular star is from the earth. So

these two are important points and these two points if uh ignored or if not factored in will lead to erroneous

calculation of the size of the star. So the measurements of the diameters of the star as derived by this chord method

would be worthless if not corrected for these uh you know sources of errors. Uh similarly a similar example comes

from say for example Paul Buer. Paul Buer in 1963 published his discovery of a substance called phosphoidine

which at that time was touted as a substance that was very important in the chemical reactions that provide energy

to the cells of the body. Now, interestingly, around two decades later after this publication of you know the

the role and the function of phosphoistadine, it was discovered that actually this element this chemical

served a very different function to what Ber had initially assumed in 1981. See the initial publication was in ' 63. So

Buer in 1981 admitted that he had been wrong and he moved on with a new theory of energy provision in cells that did

not require uh the interactable uh you know uh intermediate which he had settled with earlier and eventually for

this new theory he was awarded the Nobel Prize uh for chemistry in 1997. So you can see that science and scientists are

basically rather uh you know uh ready to modify their observations move on. It is possible that once you have a theory,

the theory, you know, allows you to see the initial observations in a different light. Some of them may make sense, some

of them may not. Some of them will basically fall in very nicely into a pieces of a jigsaw. That will tell you

or help you explain a given phenomena. All right, look at this. Now, uh this is the same picture that we were seeing

moments ago. Now, if we have provided it a context, we have provided this picture a background. Now you can see that it

seems like a picture of a actual female as opposed to uh you know slides ago where we saw this without a proper

background. So here it is very difficult to interpret oh this is what we are seeing but once you have a particular

kind of a background let's say a theoretical framework then you are able to decipher oh this is the picture of a

female. Another uh very interesting critique to uh the positivist stance can be found in

the idea that scientific theories typically include also nonobservable variables. For example, force is equals

to mass into acceleration. None of the variables in this equation can actually directly be perceived by the senses. So

the idea that positivism was based on tangible observations being made by people uh sort of uh you know suffers

here a little bit. Now uh while not in its entirety what science is proposed by positivists needs to get needed to get

around this problem of nonobservable entities. Positivists initially pointed out that while certain variables cannot

be directly perceived by their senses they could still be quantified and hence measured and have magnitudes measurably

in relatively simple ways. For example mass as weight in grams, acceleration as the rate of change of velocity over

time. Things like this. The idea was that you would use operational definitions to postulate how a given

variable shall be quantified and hence measured within the confines of a given theory allowing then for the

verification of the same. However, this idea of operational definitions does not by itself guarantee that the variables

of interest can be easily measured because you can very well have a very wrong operational definition of

something. Let's say you want to measure anger and you have the operational definition of anger as the number of

times a person smiles. Now or something like let's say for example number of times a person's heart rate increases or

breathing increases and so on. You can come up with various operational definitions but it is not necessary that

those operational definitions are the correct way of quantifying these non-observable entities.

So even with operational definitions it is possible that variables require complex indirect methods to be measured.

All right. So Carnap here points out that the distinction between observables and non-observables may be an arbitrary

distinction just as a matter of convention and so on. But again this is something that is a bit of a thorn in

the uh fundamental principles of uh generating scientific knowledge that are laid down by logical positivists.

Also uh something that is uh you know important in posterity is that non-observables can actually become

observable over time uh by the evolution of technology. For instance, the bold signal obtained through fMRI can

actually be seen as an index of the brain's functioning. You know, wherever the flow of blood and the oxygen sort of

goes in the brain uh is taken now as an index of that that particular area of the brain is engaged in performing a

particular cognitive function. So in that sense while functioning of the brain per se cannot be directly measured

but with the advent of the bold signal and there are newer techniques nowadays which can uh you know basically act as

in uh you know measures of brain activity. So uh this is one. Finally, it has been pointed out that it can be

logically impossible to verify all the conclusions uh and prove to them to be true beyond doubt. I was giving this

example earlier. For instance, while all swans are white seems like a reasonable conclusion for the most part, uh one can

really never you know go out and verify all of this to be true. you know it's it's not possible for you to meet find

every swan in this world whatever the number be and basically uh say that okay I have verified this particular fact

that all swans are white it cannot be done so verification in that sense cannot be used as an infallible basis of

science it is an important process but is it an infallible basis of science is something that can be questioned

also uh conclusions drawn from inductive reasoning are falsifiable if it even a single counter example. So

maybe we stick with this uh statement of all swans are white but somebody you know brings across us a black swan. By

the way people have found black swan. So as soon as the first example first counter example is presented the uh uh

you know the particular theory is falsified and hence the you know the validity of whatever that knowledge was

uh basically gets questioned. So verification in that sense and this is basically the conclusion of this

exercise that verification cannot be the foundation of science because while it is not possible to verify all the

scientific statements there might be merit to pursue this through falsification. So later theories and

we'll talk about this in the next lecture basically suggest that verification maybe not but falsification

has to be the basis of scientific knowledge. All right. So I'll stop here. I'll

basically continue the same discussion in the next lecture. Thank you.