1

20.109Writing Introductions andDiscussions

Neal LernerOffice: 68-150a; 617-452-2939nlerner@mit.edu

Macrostructure of a Research Article

• Introductionprovides generalfield or context.

• Methods follows aparticularized path.

• Discussion movesfrom specificfindings to widerimplications.

2

What is the form and function ofan Introduction?• An introduction is a method to

familiarize and orient yourreaders.

• The content of an introductiondepends on its purpose and theaudience.

• All models share a directapproach. Don’t hide your mainpoint or save it until the end of thepaper.

Introductions across disciplines containthe essential elements of context, focus,and justification.

Swales (1990)

Context: Orient your reader to the publishedliterature related to thetopic and to essentialbackground information

Focus: Define the researchspace, stake out territory.What questions are youaddressing? What is yourhypothesis?

Justification: Show howyour work fits into andextends previous work. Argue for the importance of your work.

3

CARS Model ofIntroductions

Create a Research Space1. Re-establish significance of

research field.

2. Situate actual research inthese terms.

3. Show how this niche will beoccupied and defended.

Swales (1990)

What are Some Common Pitfalls ofan Introduction Section?

• Including unnecessarybackground or being repetitive.

• Exaggerating (or understating)the importance of your work.

• Using lackluster openers andweak follow-through in the bodyof your introduction.

• Not grounding the work in acontext that will be important toyour reader.

• Not focusing on a clear andcompelling research question orhypothesis.

4

Tips on Writing Introductions©2003 UW-Madison Writing Center

Move from general to specific:The problem in real world/research literature--> your experimentEngage your reader: answer the questions,“What did you do? and “Why should Icare?”Make clear the links between problem andsolution, question asked and research design,prior research and your experiment.Be selective, not exhaustive, in choosingstudies to cite and amount of detail toinclude. (In general, the more relevant anarticle is to your study, the more space itdeserves and the later in the Introduction itappears.)

What is the problem?• Describe the problem investigated.• Summarize relevant research to provide

context, key terms, and concepts so yourreader can understand the experiment.

Why is it important?• Review relevant research to provide

rationale. (What conflict or unansweredquestion, untested population, untriedmethod in existing research does yourexperiment address? What findings ofothers are you challenging or extending?)

What solution (or step toward a solution)do you propose?

• Briefly describe your experimenthypothesis(es), research question(s);general experimental design or method;justification of method if alternativesexist.

REQUIREMENTS, ADVICECONTENTS

Guidelines for Introductions from TwoScientific Publishers:From the International Committeeof Medical Journal Editors:

State the purpose of thearticle and summarize therationale for the study orobservation. Give onlystrictly pertinent referencesand do not include dataor conclusions from thework being reported.

From the American Society forMicrobiology:The introduction should supply sufficientbackground information to allow thereader to understand and evaluate theresults of the present study withoutreferring to previous publications on thetopic. The introduction should alsoprovide the hypothesis that wasaddressed or the rationale for thepresent study. Use only those referencesrequired to provide the most salientbackground rather than an exhaustivereview of the topic.

Guidelines for Introductions are consistent acrossjournals and, often, scientific fields.

5

An Example of An Introduction fromthe New England Journal of Medicine

• Morbidity and mortality among patients with chronic obstructive pulmonary disease are related in large part to acute exacerbations,which occur one to three times per year.1,2,3,4,5,6 Our understanding of the cause and pathogenesis of these exacerbations isincomplete, and the role of bacterial pathogens is controversial.7,8,9,10

• In studies performed decades ago, investigators followed patients with chronic obstructive pulmonary disease longitudinally, withperiodic collection of sputum samples for culture, to determine whether there was an association between the isolation of bacterialpathogens in sputum and the occurrence of exacerbations.5,6,11 In these studies, the rate of isolation of potential bacterialpathogens from sputum samples during stable disease was identical to the rate during acute exacerbations. This finding led to theconclusion that bacterial pathogens do not cause exacerbations and that their presence in sputum is due to chroniccolonization.7,12

• An increased understanding of the genetic heterogeneity among strains of a bacterial species exposes a major limitations of theolder cohort studies.13 At the time of these studies, it was not possible to differentiate among strains of a pathogenic bacterialspecies. Therefore, all strains isolated from sputum over the course of the study were regarded as identical if they belonged to thesame species. This approach did not allow for the detection of changes in strains over time. More recent studies have shown thatthe immune response to bacterial pathogens after exacerbations of chronic obstructive pulmonary disease is characterized byconsiderable strain specificity, suggesting the importance of differentiation among strains of bacterial pathogens isolated over timefrom patients with chronic obstructive pulmonary disease.14,15,16

• We hypothesized that the acquisition of a new strain of pathogenic bacterial species in a patient with chronic obstructivepulmonary disease who has no preexisting immunity to the strain leads to an exacerbation. To test this hypothesis, we conducted astudy in which we obtained sputum samples monthly and during exacerbations in a cohort of patients with chronic obstructivepulmonary disease. Bacterial strains isolated from sputum obtained during periods of stable disease and during exacerbations weresubjected to molecular typing. This report represents the results from the first 56 months of this study.

• New Strains of Bacteria and Exacerbations of Chronic Obstructive Pulmonary Diseas Sanjay Sethi, M.D., Nancy Evans, R.N., Brydon J.B. Grant, M.D., and

Timothy F. Murphy, M.D. NEJM Previous Volume 347:465-471 August 5, 2002

context/focus

hypothesis/focus

justification

context

Writing the Discussion and Conclusion

6

What is the Purpose of a Discussion Section?

• Summarize findings presented in theresults section

• Cite supporting literature.• Explain discrepancies between your

findings and previous reports.• Point out shortcomings of your work and

define unsettled points.• Discuss theoretical and practical

implications of your work.• End with a short summary or conclusion

about the work’s importance.

Questions You Will Address in aDiscussion Section:

1. What did you expect to find, andwhy?

2. How did your results comparewith those expected?

3. How might you explain anyunexpected results?

4. How might you test thesepotential explanations?

7

Tips for Writing a DiscussionSection

“This is the place to interpret your resultsagainst a background of existingknowledge. Explain what is new in yourwork, and why it matters. Discuss boththe limitations and the implications ofyour results, and relate observations toother relevant studies. State newhypotheses when warranted, clearlylabeled as such. Includerecommendations, when appropriate.”

• Suggest the theoretical implications of your results.• Suggest practical applications of your results.• Extend your findings to other situations or other species.• Give the big picture: do your findings help us

understand a broader topic?

How do your results fit into abroader context?

For each major result:• Describe the patterns, principles, relationships your

results show• Explain how your results relate to expectations and to

literature cited in your Introduction.• Explain plausibly any arguments, contradictions, or

exceptions.• Describe what additional research might resolve

contradictions or explain exceptions.

What conclusions can youdraw?

•Summarize the most important findings at thebeginning.

What do your observationsmean?

How to address them:Questions to address:

More Tips from the UW-Madison Writing Center

8

Eight Common Components of aDiscussion Section

1. Background information2. Statement of results3. (Un)expected outcome4. Reference to previous

research5. Explanation6. Exemplification7. Deduction and Hypothesis8. Recommendation

What is the Purpose of aConclusion?

“Besides presenting an analysis ofthe key results in the conclusionsections, you also give a futureperspective on the work. In somedocuments that future perspectivemight be recommendations. Inother documents that futureperspective might be a nod to thedirection in which your research willhead. A third kind of futureperspective is to mirror the scopeand limitations that you presentedin the beginning of the document.”

9

What are the Pitfalls of aDiscussion/Conclusion Section?• Including too much

information (wordyarguments, not focused,meandering, etc.).

• Failure to follow argumentsset up in the introduction.

• Failure to focus on thecurrent results.

• Speculating too much or notenough.

• Improper tense (Discussionlargely in present tense).

• Hedging excessively.

Excessive Hedging

“The cause of the degenerativechanges is unknown but possibly onecause may be infection by apresumed parasite.”

Rule of thumb: One hedge word persentence!

10

Common Hedging Words

appearpostulatesuggestseemmay bespeculate

presumablyprobablypossiblyapparentlynot unlikelyseemingly

suppositionideaspeculationconjecturepossibilityinference

verbsadverbsnouns

11

12

Data Set: Annual Deaths in the UnitedStates from Substance Abuse, 1988

346,000125,000 4,000 4,000 2,000 75

TobaccoAlcoholAlcohol & DrugsHeroin/MorphineCocaineMarijuana

Task: Draw three conclusions from these data.