 Good afternoon. So, today afternoon session we are going to devote to reading scientific literature. Well, these slides are prepared by keeping science and engineering students you see certain terminology being used, but notice that there is a large fraction of faculty from humanities here and I want to make it clear that it is not that this word science should not alienate them, should not separate them. In today's context actually science has a very very broad meaning in particular the scientific method. The scientific method has percolated to every branch of knowledge. So, actually when in literature if you are doing a PhD actually you are being trained to use scientific method to analyze the literature that is what I would claim. Just reflect those of you who have a background in humanities and let us say you have done your PhD. You have some hypothesis and then you collect evidence from the literature or you know by conducting certain experiments and use those you know use those experiments to support your hypothesis. So, what you do is you actually end up doing a scientific study of the topic that you have undertaken. So, the method of science actually has percolated every discipline. So, you shouldn't feel that this is not for me this is for people who are doing engineering and science. Just do not go by this word science here or the scientific method does not necessarily mean you are going to apply it only for doing some you know the so called natural sciences. Social science it could be any other investigation that is undertaken scientifically. When I say scientifically you have to follow the method that was shown to you. You have to have a hypothesis. Your hypothesis should explain your basic you know observations and then those observations you could come up with more predictions using a hypothesis and you should actually corroborate those predictions using tests and so on. So, those of you who are from humanities background probably your department has been handling the course of technical communication. So, whatever I am talking is equally relevant in any domain. It's not necessarily only in natural sciences or engineering. So, technical work implies you are trying to solve a problem and just a quick overview of what we have done and you have to have a question and you have to have a crisp answer for this question at the end of your endeavor. So, a question would involve clearly stating a more objective of your work will actually give and it could actually formed as an interrogative question to what where why and so on and your answer in short in crisp form will give you what is it that what is it that you achieved through the scientific endeavor. So, this it could be that it's your own scientific endeavor which you are trying to summarize through the question and answer. It is possible that you are trying to read somebody else's work and then come up with the question that person or that group of researchers have posed and then the answer they have found. The second thing that we emphasized in the morning was that this is how a scientific method works. So, first of all you since you have a question you start doing experiments you may have observations. If you have interesting observations which nobody else has seen before you end up communicating those observations. You go back to literature and see if some old hypothesis is not useful to explain some new phenomena. Then you come up with a new hypothesis and revised hypothesis and then you know you come up with predictions using your new hypothesis then come up with a test for that hypothesis and the process goes on. So, this is an iterative process it develops it improves itself and all of us indirectly or directly contribute to this exercise. When I say directly when you are writing a report or you are writing a paper or you are writing a thesis indirectly when you are training your students to do this. So, we are actually contributing to this entire process of maybe our contribution to a to some solving some bigger problem could be very very small. Maybe you have just verified some small hypothesis small applied it to some small problem does not matter. So, you know everyone's contribution matters. So, what are the types of scientific literature? So, at the forefront of any scientific research in any domain are the general articles. So, again and again I want to stress scientific as in it could be a scientific investigation taken up in the area of literature. So, are the general articles and then in every domain there are reading general articles, journals that publish research articles. If you look at if you try to classify these general articles they could be letters. So, letters are typically short communications which are typically published you know they are named differently in different domains. In some cases they are called as short communications in some domains they are called as letters. So, depends upon your sub domain where you are working. So, these are typically some quick results that or some small improvements that you want to publish for the wider audience. Then there are original articles. The original articles are typically full length papers. So, in the area in which I work it is called full length paper or regular paper. So, this would have you know detailed treatment of a hypothesis or development of a new model or conducting experimental verification of some existing model and so on. So, this could be a fairly involved thing but it is specific to one particular small topic and the third kind of articles that are published in journals are review articles. So, review articles are the ones in which experts are invited to collect development in a particular area over a period of time and then you know come up with a sort of a coordinated presentation of the developments over last few years. This could be in different areas. It could be some field of mathematics. It could be some field of science. It could be you know we have this department of education technology in our IIT where people look at what are the effective methods of teaching students, undergraduate level, post graduate level and so on. So, maybe some expert collects developments in effective ways of teaching over last 10 years and acts his or her own interpretation you know comments on what is important, what is not important and then comes up with a review article. So, all these three classes of papers are important. They if you want to know where the research is actually happening, it is these two the original articles and the letters. The review articles is something like trying to screen and collect what is good, what is relevant and what is worth you know taking forward by some experts. Then you see these monographs. These monographs are typically collected body of work on some specialized topic, specialized area. Let us say somebody is working on wireless communication using certain kind of antennas and then there is some development over last 10 or 15 years. One expert or maybe a group of experts come together and write a monograph or research monograph. A research monograph is typically you know an advanced level book. I would not call it a textbook, advanced level book which collects the recent developments in a fairly detailed manner so that an expert can benefit from this. The target audience is typically doing research or you know practicing research group in industry or you know so it is meant for experts and experts were involved in modifying things and developing things and so on. At the next level are handbooks and encyclopedias. So this is where experts are again invited to summarize technical developments for a wider audience. So this is not necessarily for a you know technical expert or researcher. This is for practitioner. A lay practitioner who is in the field would need to use the accepted theories and encyclopedias do that task. So of reaching out to these people and at the bottom of this you know technical or scientific literature are the text books. The text books are meant for beginners for educating undergraduates whom you are training to be later on to be experts in the field. So actually if you trace the development okay in time this is actually evolution of a hypothesis into a theory. So this is where you present theories which have been very well accepted. The theories or the set of hypothesis and models that have stood the test of multiple test of falsification and everyone agrees. All the experts as of now agree that this is the way you explain a particular topic. So this is the level which is the most calm okay. If you want me to give you a comparison this is like a probably all of you know whether you know fluid mechanics or not is that the water at the bottom is not flowing. It is mostly stagnant particularly next to the you know the river bottom. There is a layer which is completely stagnant it is not changing it is if it is changing it is changing very very slowly okay. But as you go up in a river bed in a river okay you see there is lot of turbulence okay so same is true same is true at this level okay. People or the research groups researchers are arguing or you know it is the same thing that happened today morning okay except what you are arguing about is different okay. So you have some research problem and then different research groups are arguing about what is the correct way of explaining it what is the best model what is the best hypothesis and that is where there is no agreements there is no unique way of explaining you are questioning everything. At this level at this level you do not expect a student to come and question at this level at textbook level you want the student to accept and reproduce whatever is accepted theory he or she should be without questioning should be able to. So these two levels are the ones in which you never question you just accept and reproduce whatever has been said by the experts. But this is where the turbulence begins and this is where there is maximum turbulence there is no agreement in many many times. So if you just to connect with a with any technical document okay any paper in any domain whether in literature or in science or in technology or engineering okay it is roughly organized into these sections that I am going to talk about. This is something that has evolved over a period of time and maybe what we have listed here with one or two differences is true in every domain so just correct us if you have. So of course first thing that you look at in the title the title tells you about it is a short sentence that tells you what is being investigated in a particular paper. Well why are we going to talk about discussing why are you going to discuss about how to read a paper okay. Because the other thing is how to read a textbook all of us have already been trained how to read a textbook okay. What we are not trained for is how to read a paper. And many times when you start doing any independent work any independent project okay you have to cross the boundary of you know textbook and encyclopedia and start entering into monographs research papers and you are forced to read papers you are forced to read the literature where things are not things of fluid they are not completely crystallized into a theory okay. That is why you should know how to read a paper and we are going to focus and spend some time on how to read a paper. So you should look at author and affiliation an important thing you should know where the paper is coming from that gives credibility to the paper you know it's coming from a very sound school if your paper is coming from TIFR or it is coming from Tata School of Social Sciences. Well you know that the researcher is you know well acknowledged and you know you can believe his or her opinion and or researchers group of researchers. The next thing is of course the abstract which gives you what is being done in this technical work. This is followed by introduction. The introduction is typically mixed with literature review in some cases in some some some domains these two are kept separate is a separate introduction and separate literature review. So this is this is where actually you know this part of the so this is a this is a cartoon that shows how a paper is organized okay. So what this what this is trying to tell you is that there are broader implications of what you're being what is being done in the paper those are discussed in the beginning and in the end okay. This is connecting this is connecting your paper with the literature which is available okay what is the significance what are the what observations are being explained and what are the new observations that you're going to make. Specific problem okay specific hypothesis that you advance the you know evidence that you create is given here and in the end there are conclusions which are broader implications of what you are investigated. So main body of the work consists of the method the results and discussion and then you again connect with your conclusions connect with the broader literature broader investigations being taken by the scientific community. And of course in the end you provide references which is the support for you know which is that you have looked at the literature and then you're trying to build upon the literature. So all these this is a typical organization of a of any research paper in any domain. So is there any correction to this anything to be added this is broadly the organization of a paper in any domain in humanity is in yeah. How to decide keywords so that we will that we will come later at some point in the right so right now I'm not going to talk about it so keywords could be an important important component okay between the abstract. Depends upon the domain it depends upon which even within say chemical engineering I'm from chemical engineering and he's also from chemical engineering but he works in fluid mechanics I work in control systems. So the conventions are different so it could be that some authors like to or some in some in some journals you just write a crisp introduction and then go to literature review in some cases you just mix two into one okay. So this could be together this could be in one but usually the humanities are both are different introduction you are introducing the subject and then you are going for a degree of literature yes so that's what so it within humanities also you might find different conventions in say linguistics might be different from sociology and also it depends upon that but even if so this is for a full length paper if you take a full length paper these are typically the sections typically yeah. I think one more that is objectives or the hypothesis of the paper we should it should be a kind of element there no objectives of the paper those are stated typically in abstract introduction so I'm talking about organization of a paper that is across domains it is they are the broad sections so you explain the objectives in the introduction section and in the abstract yeah yeah acknowledgments we are listing here the technical components of the paper okay so acknowledgement is important well who is funding who who is giving you money is important of course but not every investigation will be funded like this some cases could be so let me continue with the next important that I'm going to talk about a method to read a research paper now question is that who is a target audience so who are expected to read a research paper okay first of all reading something that involves research starts coming when a student undergraduate student comes to his final year in the final year there is normally a project okay and a difference between the courses that a student takes and a project is that in the courses the student is not given a hypothesis or is not expected to come up with a hypothesis to explain some observation in in the courses he's taught a theory that this is how it is you know if Apple is falling this is the law of gravitation okay you're not supposed to question you're supposed to understand how Apple falls using gravitation how moon rotates around earth using you know driven by the gravitation gravitational force and same thing about say earth and the sun you're not supposed to question here okay but when student starts his or her own first project okay that is where they're supposed to carry out some independent investigation okay so so what we're going to present here is not relevant only for people who are doing research it's for everyone who start doing a project in a project you know when you start doing a project it's beginning of it's beginning of the you know doing an independent investigation or it's beginning of learning to use the scientific method okay see in a project you are given certain you know topic okay a topic indirectly states on hypothesis okay and then you are supposed to collect evidence to support the hypothesis okay and develop develop around this hypothesis so that is what is expected from a topic so so don't don't think that this is meant only for research when I say research is any independent investigation in which a student is expected to learn by collecting material from multiple sources okay maybe not just encyclopedia's maybe you know monographs research monographs or maybe from you know research papers so just we just had a we just went through this so I will skip this we went through different parts of a and definitely keywords are important because they quickly tell you what is the broad area which is being investigated now why why we look at this way to read a paper this is a this is a difficult problem that is typically faced by a person who start doing a project independently okay so nowadays you give some topic to the student the student immediately goes to Google and types the keywords and well he either gets some 1 million hits and then comes and says I don't know what to read you know there are so many so many things to read how can I find out so then of course the tendency is to look at the first one which Google gives which may not be a relevant paper and then start reading that okay or the other answer that I typically get is that there is no literature on this why because Google is not giving anything so simple answer there is no library there is no other source of you know finding out so okay so leave leave that aside so now now let's take the first scenario where you know a student is given some topic and he and she comes up with you know find that there are some 30 40 50 papers research papers in this particular area which he or she is expected to read in a short period of time maybe in a semester so if it's a course project in a semester you have to read some 10 20 papers and write a literature review okay so how do you do that one thing that we need to understand when we shift from reading a textbook to reading or reading for a course to reading for a project and scanning literature is that we have to move away from the idea of reading a document from first word to last word okay so we have been you know taught from the day one from our school is that reading a particular thing means open the textbook read the first word and read up to the last word okay that is what is meant by a reading and this this you know this training haunts you for a very long time even after you start doing research sometimes for many years so we have to learn to quickly grasp what is there in the research paper and then make a judgment whether I want to spend more time on it or not okay and well doing this business of you know reading every word of a particular document you might end up doing for only you know a few few few papers and you should be able to first come up with a short list in which you need to read the paper from first first line to last line okay so here is a method given by a professor Kesha from University of Waterloo from Kennedy University well he has given a three pass method to read a paper so first pass he says is observe the second pass is judge and a third pass is understand and I'm going to spend some time talking about these three passes the first pass is when you want to quickly scan the paper okay so here advice is to read the title and abstract and the introduction okay as the case may be maybe you want to go back and also read literature review if literature review and introductions are separate you may want to quickly scan the literature review because this is this is where the broader part of the paper connection with what is existing body of knowledge is discussed okay what is the objective what is the you know what are the method that is broadly going to be used all that is discussed in this part in the abstract and introduction then you should just browse through the headings and skip details skip content okay and then jump to the conclusions well not in the literal sense you shouldn't jump to the conclusion about the paper but jump to the conclusion section of the paper okay and look at the conclusions look at the introduction look at the abstract this is a quick pass okay what you should note you should note the following points they should note the bibliographic details who is the communicating author well you want to know what is the authenticity of this person's work so typically in a in a research group the person who is supervising the research is the communicating author so my if I'm writing a student paper with my PhD student my PhD student will be the first author he's the one who has done the work but I might be the communicating author okay so look at who is the communicating author and then try to quickly classify the paper whether this is a review paper it's a short communication you know adding something to existing body of knowledge by some small improvements or it's a it's a original article which is some detailed investigation on some then we can also categorize what kind of work is reported whether it is experiments have been conducted and mainly it's reporting the experimental work whether it is some computations and simulations some new theory new theories being proposed and developed you know you should probably note things like digital object identifier DOI because it's easy to share this DOI is a unique number for any net document and you can share with your core co-researchers co-workers DOI or a particular you should look at broad area of the work and look for the keywords what are the keywords and and in the abstract in the title in the abstract and this pass should not take too much time from the introduction from the abstract you should be able to formulate what is the main question that is being investigated okay and what is the answer that has been provided okay this should be clear from the introduction abstract and conclusions when I'm saying this I'm indirectly giving you a hint how you should structure your documents when you write any document a project report or a thesis or a paper or a patent okay introduction conclusions and abstract should communicate two most important things what is the question that is being asked for and what is the answer that is provided okay so this should come clearly so well you can actually then categorize whether a new observation has been reported or an old observation is being tested old hypothesis is being tested on some new system so you could categorize the paper quickly without getting into reading it line by line okay so here you should not try to judge in the first pass just try to record what is this what is what is the broad area what is the question being asked and what is the answer provided and try to summarize the paper in less than 20 30 words okay because the Kesheva thinks that it can take five to 15 minutes I would say this is five to 15 minutes for an expert if you are beginning to read papers maybe it will take you 25 to 30 minutes okay read read the introduction abstract and then form the question and try to write the answer which the authors have provided the next pass the second pass is judgmental and again we are not going to read the paper from first first sentence to last sentence okay so now we are going to look at we are going to analyze the paper in more detail for example I might decide to look at you know in a scientific paper in which experiments have been carried out I decide to look at the figures and illustrations okay now we have done any experiments okay if you are if you are a careful researcher you would put these error bars what are the error bars can somebody help me in a figure what are the error bars have you seen figures with error bars maximum and minimum error so so what do you what what what we do typically when you are doing experiments is that you repeat the experience multiple times okay and then you report the mean and you know maybe two standard deviations mean and three standard deviations you give a band around the mean so to tell you that actually my observations were lying in this region okay majority of my observations were lying in plus or minus 3 sigma plus or minus 2 sigma so you give this band error bars if it's a researcher who is doing experimental work is giving a figure with error bars that tells you that this researcher or this research group is very very careful they repeated the experiments they are not giving you one of the results okay so it's it's important to see all these small things just look at whether axis is properly marked whether the figures are neatly drawn so is the problem stated clearly you know again we are not getting into two details of reading the paper but we can start browsing we can start scanning the content and then try to make a judgment whether the problem has been stated clearly in the introduction in the is the motivation for the work stated clearly connection with the literature stated clearly the solution methodology adopted is it justified okay again what is the methodology adopted you should be able to make out just by reading introduction and conclusions and then from your own understanding of the subject you could make a comment well this new approach seems to be interesting it doesn't seem to be you know good idea to use this approach or whatever whatever is your judgment so the second pass and we try to look at look at many other things for example I would want to see has this author done similar work in the past an important consideration because if somebody has been working in the area for quite some time that again gives authenticity to his you know his work but it is also there's also one more one more thing many times nowadays you find many researchers over publishing okay they may not be significant improvement you try to scan papers of particular group or particular author and you see that there are some minor improvements over the last last paper okay so that also that also you can note whether this is a minor improvement over the previous work whether it is a major difference so so no down your judgments and this was a Keshava again things should not take more than 15-20 minutes so overall if you are a beginner you know past one plus past to say should take about 40 to 45 minutes or 40 to 50 minutes okay so here in this 40 to 50 minutes we haven't read the paper from first line to last line okay now if if this paper crosses these two stages that is after having done the judgment okay if you think you want to go further and read the paper okay then you would do what is called as the third pass the understanding pass okay so this for any project whether it is you know your final year undergraduate project or whether it is your PhD thesis you might do it only for few documents for example if you're doing your master's work you might read a paper thoroughly from one first line to last line maybe four or five papers okay let's say in my domain in engineering you might end up reading 40 papers okay in your review to put the work in the context but all the 40 papers I have not read from the first line to last line okay this this reading is like a thorough reading of a paper and it can take few hours to few days it is not it's sort of so read the methodology very very carefully okay then go back and read the literature so this is not reading only one paper we should go back and trace the literature and see just suppose the work presented in this work in this paper with reference to what has been already done in the literature okay so just read the relevant books and literature read every sentence critically so this is where you know you are reading a paper the way you were reading a textbook earlier but there's a big difference there's a big difference in the textbooks were never questioning whenever you never thought that this could be wrong okay now when you're reading a paper you're reading it with a critical mind with a questioning mind this what is being presented may not be correct okay may need improvement okay maybe partially correct okay so it's it depends upon okay it's good to do a mental recreation of the work if what what would you do if you were to do this okay so that helps in understanding a research paper well verify if there are mathematical proofs you should try and work them out yourself this mental recreation of the work that how you would have done it actually it's in much more clarity in what the authors are trying to present you could also you could also mentally try to look at alternate approaches that you would use to solve this problem so don't have to actually sit and solve it but you know if a method A has been used to solve a problem and you know what if method B was used what would happen you could you could actually think about all the possibilities that might lead to new research ideas for example you should look for is there a missed literature quite quite often it happens that many so so the body of knowledge has grown so large that you know one of the reasons that we subject a work to peer review is that because the peers collectively are reading the literature okay and then we hope that you know a similar work has not been done already so that that we try to that we hope to find from these experts from these peers okay so you when you are reviewing a paper or a research paper or a thesis you should try to see whether there's a missed literature something has already been done not been reported okay now when you are submitting a review a report on the thesis or on the paper you should write down the overall purpose as you it what is the problem and what is the thesis the introduction of the paper introduce this look at the body of the work look at the correctness of the work is the paper focused on these goals so we should actually look at this in the introduction is the problem stated clearly and does the does the main body of work focusing on the problem that has been stated or it's rendering over something else so all the reasons given or the proofs given do they justify the claims so this is a this is a involved long process okay and this is something in which we need to train ourselves because you know your coursework doesn't train you to do this that is to look at things critically in a questioning manner okay so we have to we have to train ourselves in doing this so we have to look at conclusions are the conclusions properly bringing out what has been investigated this is a strong finish or it's not stated properly of course you should check grammar and things other things that are important like typos examine citations whether the citations have been correctly given for correctness for completeness or formatting style and you should collect your final comments so this is this is what you need to do when you review a paper or when you review a thesis or when you review your report at whatever level whatever level so we when when when we are reviewing a report here in IIT Bombay for undergraduate project we try to apply the same ideas that we would apply for a m tech thesis or well we may not expect the same rigor from of other undergraduate student but but conceptually it should be similar structure okay because again even if it is an undergraduate project it is under some small hypothesis being tested some small model being used okay and it's an investigation which is independent which is not copying from a book it is some kind of independent investigation so it is subject to all these tests okay so this this exercise you might end up doing for a few papers when you are doing your research or a few documents when you are doing your research or as a role in the role of an examiner you might do this for for your students okay so these are the three passes of and you should read this paper by Keshav it's just a two-page document it's very nice it is nothing to do with science or technology it is just reading a scientific document a document prepared scientifically okay scientific document doesn't mean natural sciences any science any document there has been prepared using the scientific method okay and then this book the craft of research is also a useful document okay thank you