How to conduct an effective Literature Review TEQIP Short Term - - PowerPoint PPT Presentation

How to conduct an effective Literature Review

TEQIP Short Term Course on Research Skills and Methods 19th - 21st February 2016 Ashish Garg IIT Kanpur ashishg@iitk.ac.in

Bibliography

• “A Gentle Guide to Research Methods” by

Gordon Rugg and Marian Petre, McGraw Hill

• Many other on-line articles

2

What is “Literature Review”?

• C. Hart (1998) “Doing a literature review:

Releasing the social science research imagination”, London, UK, Sage Publications.

– The use of knowledge/ideas/approaches in the literature

• To correctly iden8fy the problem (if vaguely known

beforehand)

• To jus8fy the par8cular approach taken to the topic
• To assess the selec8on of methods used, and
• To understand that the research to be done contributes

something new

– Quality of the review means appropriate breadth and depth, rigor and consistency, clarity and brevity, and effec8ve analysis and synthesis

3

What is “Literature Review”?

• J. Shaw, System, 23(3), 325-335 (1995).

– Process of the review should “explain how one piece

• f research builds on another”.
• Webster and Watson, MIS Quarterly, 26 (2),

13-23 (2002)

– An effec8ve literature review creates a firm founda8on for advancing knowledge. It facilitates theory development, closes areas where a plethora of research exists, and uncovers areas where research is needed.

4

Literature Review

• An effective literature review should include

the following characteristics:

a) Methodologically analyze and synthesize quality literature, b) Provide a firm founda8on to a research topic c) Provide a firm founda8on to the selec8on of research methodology d) Demonstrate that the proposed research contributes something new to the overall body of knowledge or advances the research field’s knowledge-base.

• Helps you to create a good impression on the

supervisor that you do understand the state

• f the art in your chosen field J

5

Stages of Literature Review

6

• 1. INPUT
• 2. PROCESSING
• 3. OUTPUT
• i. Knowledge of the Literature
• ii. Understanding of the Literature
• iii. Applica8on
• iv. Analysis
• v. Synthesis
• vi. Evalua8on

Quality of input (Source and Search tools)

Knowing and Understanding the Literature

7

Example: One is doing a literature review on “CO2 emission and global warming”

Case-I It has been showed in various studies that CO2 emission leads to increase in the global mean earth temperature (ABC et al, DEF et al…). Case - II It has been showed in various studies that CO2 emission leads to increase in the global mean earth temperature. For example, ABC et al. conducted a detailed study and found that CO2 emissions lead to an increased of 2-5 degrees in the mean temperatures depending on the loca8on and level of pollu8on. In another study, BEF et al. showed that the increase in the mean temperature was between 1-2 degrees and the devia8on was also dependent on the local mean temperatures at a specific loca8on.

BETTER

Application

• Typically a two step two-step process

– Iden8fy the major concepts or topics or subtopics relevant to the subject under considera8on, and – Place the cita8on in the correct category of concept.

8

Concept 1 Concept 2 Concept 3 Concept M ArKcle 1 X ArKcle 2 X X ArKcle 3 X ArKcle N X X

Analysis

• Analyze the data

– Understand hypotheses – Understand models used or experimental condi8ons used – Make connec8ons – Compare and contrast – Find out strong points and loopholes – …..

9

Analysis: An Example

10

Case – I Data mining is the analyzing and interpreta8on of large amounts of informa8on. Through analyzing vast amounts of data it is possible to find paeerns, rela8onships and from these discoveries it is possible to make correla8ons (Chen & Liu, 2005). Case II Data mining is a process of discovering new knowledge by using sta8s8cal analysis to iden8fy previously unsuspected paeerns and clustering in large data sets (Chen & Liu, 2005). Why it would be of any interest or value to find paeerns and rela8onships in order to make correla8ons?

Informing Science Journal, 9, 181 (2006)

Synthesis and Evaluation

• Bringing out something new out of the

literature review

– Iden8fica8on of unsolved problems or issues – Problems in the models or experimental designs used in past – A new idea – Provide recommenda8ons and conclusions

11

Synthesis: Example

12

Example CO2 emission is believed to be responsible for increase in the global mean

• temperatures. The problem is associated with emissions due to coal burning, biomass

burning, vehicular pollu8on and other human ac8vi8es. However there are studies which contradict this view and suggest that global warming has no rela8on to the atmospheric CO2 levels. Example Various human ac8vi8es related to burning of hydrocarbons lead to CO2 emissions in the atmosphere which are widely believed to be responsible for increase in the global mean temperatures. At the same 8me, there are studies which contradict this view and suggest that global warming has no rela8on to the atmospheric CO2 levels. It would thus be interes8ng to examine this in detail by modeling the data as well as by carrying out experiments as well as data collec8on in areas which have experienced large increase in the CO2 levels due to industrializa8on and find out why these discrepancies occur. √Beeer

How do you actually do it?

• Exploit the power of web in a positive way
• Rely on refereed articles published in

scholarly journals, not just on some news items or random articles

• There is a method to it.

13

Method

• Specify your research question.

– It is tricky, should not be too narrow or too broad – Learn by experience – Example: How to search for informa8on related to global warming by using key words

• Global warming
• Global warming and pollu8on
• Global warming and air pollu8on
• Global warming and air pollu8on and carbon di-oxide
• ……………

14

Use of Bibliographic Databases

• Identify ones which are of your use

– Web of Science (or ISI web of knowledge) – Scopus – Google Scholar

• Build a tentative list of terms that you would

use for search

• Narrow your search
• Refine your search

15

Further work

• Thorough reading of the papers/articles

– Forward and backward search

• Read critically, make observations of the salient points

and create summaries

• Compile the information, compare and contrast
• Go back to step-1 if needed
• Determine what is done, what is the relevance and the

importance of the work done, what are the missing links and the remaining challenges on the topic you have chosen.

16

Important aspects

• An effective and quality literature review is

based upon a concept-centric approach rather than chronological or author-centric approach

– J. Webster & R.T. Watson, MIS Quarterly, 26(2), 13-23, 2002.

• Authors of literature reviews are at risk for

producing mind-numbing lists of citations and findings that resemble a phone book – impressive case, lots of numbers, but not much plot.

– D.J. Bem, Psychological Bulle8n, 118(2), 172- 177, 1995.

17

Output of a literature review

18

Hart, C. (1998). Doing a literature review: Releasing the social science research

• imagina5on. London, UK: Sage Publica8ons.

Structure of an argument

Key points

• Literature review is not a one time thing.

– It is a con8nuous and cyclic process that one may need to go back and forth 8ll the end.

• Important to comment upon the comparisons

between numerical and experimental studies.

• Reasons/arguments must be constructed to

explain the disagreements/discrepancies like some of the assumptions made in numerical/ theoretical studies are not at all only partly met in experiments.

• Conversely, experimental error analysis must be

performed to ascertain whether the differences are significant or well within the error band.

19

Writing a literature review

• A difficult task to write a good review
• Unlike an introductory chapter in a thesis or

report, a literature review is usually more comprehensive and elaborate and is well- supported by the evidence.

• Topics are usually covered in substantial

detail and supported by references to research

20

Structure of a literature review

• Various methods of writing a literature

review

• One can either go by the authors or by the

topic

– Usually, second approach is beeer than the first

21

Process of writing the literature review

• Read and understand the articles

– Understand the purpose, assump8ons & hypotheses made and the claims made in the form of results and analysis

• Organization of the topics

– Synthesize the informa8on in the form of notes, take references

• Create a set of topics and sub-topics
• Iden8fy key references for each topic

– Create a story in the mind

22

Process of writing the literature review

• Jot down a framework in the document

– For example, make a list of various sec8ons

• Introduc8on of the area
• Statement of the problem
• Topic and sub-topic wise list of sec8ons
• Provide a discussion
• Provide an account of what needs to be done
• Summary and outlook
• References
• Fill in the details
• Read it thoroughly and get rid of the mistakes

– Pay special aeen8on to factual mistakes, spelling mistakes, assump8ons and claims

23

Reading the articles

24

How to read a scientific article

• A complex task
• Try not to read it like a text book
• Identify its structure and features
• Look for author’s main points
• Ask questions before, during and after

reading the article

• Draw inferences based on your own

knowledge and experience

25

Structure of an article

• Most article follow a conventional structure:

– Abstract – Introduc8on/Literature Review – Methods/Experimental Design/Calcula8on Details – Results – Discussion – Summary/Conclusions – Acknowledgement and References

26

Abstract

• Purpose of the study
• Very brief account of Methodology
• Key results
• Most important message(s)

27

What next: You may move to figures to get a feel of the contents!

Introduction Section

• What is the purpose of introduction section?

– Provide background Informa8on – Expose the lacunas or gaps in the exis8ng body of knowledge and state the problem undertaken – Show mo8va8on of the study

28

Introduction

29

Roy et al., Am. J. Phys. 75 8, August 2007

Introducing the topic

Introduction

30

Roy et al., Am. J. Phys. 75 8, August 2007

Build up to the problem with examples of works done in past

Introduction

31

Roy et al., Am. J. Phys. 75 8, August 2007

Highligh8ng the problem

Introduction

32

Analyzing the Problem

Roy et al., Am. J. Phys. 75 8, August 2007

Introduction

33

How the authors have tackled the problem exis8ng in the literature Analysis of the problem

Roy et al., Am. J. Phys. 75 8, August 2007

Experimental/Calculation Details

• Experimental Details:

– Design of Experiments – Explana8on of the type of experiments conducted and their details for one to be able to reproduce the results

• Explain the calculation approach

– What kind of soqwares or algorithms were used, what were the assump8ons made etc

34

Details of the Experiment

35

Roy et al., Am. J. Phys. 75 8, August 2007 Need to be enough to allow repeKKon

• f the work

Results and Discussion

• Results

– Repor8ng of various kinds of data in the form of figures/graphs and tables – Observa8ons made with appropriate references to the figures and tables

• Discussion

– Comparison of the findings with the literature – Explana8on of any new finding and propose possible reasons by means of either or all of analy8cal/ schema8c/textual methods – Inferences drawn in a broader context – Answers to the ques8ons posed in the introduc8on

36

Results and Discussion

37

Roy et al., Am. J. Phys. 75 8, August 2007

Results and Discussion

38

Source of error

Results

39

Roy et al., Am. J. Phys. 75 8, August 2007

Analysis of the data

Results

40

Roy et al., Am. J. Phys. 75 8, August 2007

Analysis and Novelty

Discussion

41

HighlighKng the important features LimitaKons Possibility of further studies

Roy et al., Am. J. Phys. 75 8, August 2007

What are your critical observations?

• On Page 1, why should the drag force consist
• f a damping like behaviour?
• How correct is the assumption about

dependence of damping coefficient k on permeability of vacuum, magnetic moment, conductivity and inner radius?

• What will this happen if the magnet was not a

good metal?

• What if the magnet is not cylindrical?
• What is magnet gets titled?
• What if the gap between magnet and the tube

wall is tiny?

42

What else?

• Extension of this idea to some other problem

– For example how do shock absorbers work? – Can there be any applica8on of this experiment or Lenz’s Law?

43

Summary

• Have some idea of the topic in mind.
• Choose right sources and be methodical.
• Get enough literature relevant to your

problem.

• Read thoroughly and be critical.
• Ask questions at every stage.
• Make notes.
• Remember it is a continuous process even you

have found your problem and have started experiments/calculations/modelling.

• Have fun while doing all this J

44