This page lays out the main sections of a research paper and also gives you
mini-outlines for what you need to say where. Use this to generate a
first draft of your paper and to make sure you have covered all the right stuff. Your
reviewers will be very helpful in telling you what a bad job you've done, so
your second draft will be a lot better than this one.
Part I: Introduction
Imagine that you are participating in a lively conversation
with several people talking at once. You want to respond to something
one person said, but you can't squeeze your comment in before someone else
speaks. So when you finally get a chance to speak, you have to clarify
that your comment is in response to something that was said earlier.
This is what you have to routinely do when you are introducing your papers. There
are always multiple conversations going on in your field, so when you insert
your own comment, i.e., your paper, you need to tell readers what conversation
this responds to. Say that you are responding to the debate about retaining
students in grade, or to recent efforts to improve teachers' induction experiences.
One way to think about your introduction is as a way of clarifying who you
are responding to—to alert readers that you are responding to the induction
issue or to the grade retention issue.
It actually takes some time to do this. You can't really do it as you
would in an oral conversation, you have to carefully build up your introduction
so that readers can follow what you are doing and where you are going. And
of course you will want to refer to other authors as part of your effort to
show who and what your paper responds to. Think of the introduction as
something like a paper trail, showing readers the trail from other literature
to yours. Below I offer a sketch of what the main parts of your introduction
should look like. For
starters, try writing one paragraph for each of these points:
Here is an Issue
Your first paragraph should introduce the ISSUE, not the study. You
should point out that there is concern about an issue, or that there is disagreement
in the field about the issue, or that the issue is important because policy
makers need to guidance. You will need some citations to demonstrate
this urgency to your readers. This is the first step in showing how
your paper responds to others. Do this in a way that
segues into the next point:
Here is how Others have Dealt with it
Next you need to lay out how other people have addressed the issue. How
you do this will depend, of course, on how they have addressed the issue. You
may want to point out that there are two schools of thought, and then describe
each one and give some examples of each. Or you might want to show how
thinking has evolved on this issue, and give examples of early thinking and
more recent thinking. Or you might want to refer to a recent literature
review that did a nice job of summarizing where we are as a field on this issue.
Here is a Remaining Question
Once you have an ISSUE laid out, you can raise a QUESTION. You can say,
even though Jones found this and Smith found that, we still don't know
about this other thing. You can use your introduction to the ISSUE to
raise a QUESTION, asking why are other study findings have been so discouraging,
why the issue has generated so many conflicting ideas, why the solutions
that have been tried have yielded a particular pattern of results. You
might also offer a HYPOTHESIS, speculating about the answer to your question.
Here is the Study I did to Address this Question
Now that you have an issue and a question (and maybe a hypothesis), you
can introduce your own study, saying that you did this study in order to
address this issue and/or to answer this question and/or to test this hypothesis.
Then say more formally what the research questions were that motivated
your study or literature review or book review or whatever.
OK That's a good introduction. You've established
that there are people who are thinking about this issue, you've examined
their thoughts and findings about it, you've pointed out that there are
still unanswered questions, and you've said you aim to address one of those
questions. If you've written at least one paragraph on each of these points,
you have an introduction to your study and you are ready to tell about
what you did during the study.
Part II: Methods
can introduce your method. You tell people how you approached this problem.
This is often even more difficult than the introduction because you often don't
realize how many decisions you made that are relevant to your findings. If
your paper is a literature review, for instance, you have to say where you
got your literature. Did you include dissertations, or only journal publications? Did
you do broad searches or just review the literature that your advisor recommended? (Puleeeze!)
So what do you need to tell people?
How you found the schools, teachers, classrooms or students you studied. Were
they people you already knew, or did you have to persuade them to participate? How
did you persuade them?
Say something about the school context. Don't just say it's an urban
school or it's a low-income school. If the district allows it, give
the name of the district. This is the best way to be clear about the setting. If
you are not allowed to do this, give details about enrollment size, grade
span of school, percent of students on free or reduced lunch, ethnic mix,
and any special circumstances that may be relevant, like that the school
just got an award for something, or it is near the university and most of
its teachers graduated from the university, or that it faces stiff competition
from a local charter school, or that it's enrollment is dropping because
a prominent local industry recently relocated.
How many schools, teachers, classrooms or students did you study, and are
you presenting data from the entire group or from a subset? Why? What
What kind of data did you collect from them? Interviews? Surveys? Observations? Did
you gather artifacts like lesson plans, homework assignments, etc? When
did you visit, when did you get each kind of data? Lay this out clearly
because the meaning and value of each datum depends in part on when you got
it. If you used any standardized questions, as in an interview protocol
or a questionnaire or a test or structured task of any sort, you need to
show readers what this looked like. Insert some sample questions.
did you do with the data? I don't mean how did you transcribe
it or what software you used, I mean what analytic work did you do? What
categories were created, how did you decide that one response was a category
A and another was a category B? Show some examples of categorizing
decision rules so readers can understand what you mean by your groupings,
your concepts, your scales and so forth
OK. By the time you lay out all of this, your readers should be clear
on what you did. This is important for two reasons. First, of course,
you want to reduce confusion. But in addition to that, each of these
details can affect your findings. Your readers need to evaluate the trustworthiness
of your findings, and that depends on the trusworthiness of your evidence. If
your entire case rests on Mrs Jones, and she happens to be your mother, readers
may be skeptical about what you've really learned. If you studied the
role of tests in teachers' curriculum decisions, but you limited
your study to the two weeks just before the test is administered, your evidence
is likely to exaggerate the role of tests. You need
to be honest about these details so that others can decide how to interpret
your evidence. Evidence that is gathered two weeks before the test will look
different from evidence gathered six months before the test.
Note that you are also free to comment on the meaning of these decisions. If,
for instance, you did visit teachers just before annual testing began, you
can explicitly address that fact and can say something about whether you think
this timing really compromised your findings. You may have good reasons
for this decision and you should lay those out here and now. You don't
want to be dishonest, but you do want to persuade readers that you can make
a plausible case with the evidence you have and you want to acknowledge any
limitations inherent in the study.
Part III: Findings
I'm not giving any help here. You're on your own. Findings will
vary depending on what you learned. However I can give a few admonitions:
Don't be afraid to organize findings around the main things you think you've
This is not a discovery learning exercise, where you have to be shy about
the main point, laying out zillions of details and hoping your reader sees
the same pattern you see. You can say something like, "A big theme in the
interviews was X," or "the majority of people surveyed said Y," and then
provide the supporting evidence for that claim.
Always walk readers through any tables you present.
Your discussion should go something like this: "Table 2 shows
the number of teachers who responded positively t0 each of the scenarios
row represents a particular scenario. The columns separate out elementary
and secondary teachers. The table shows two important patterns: First,
elementary teachers were generally less positive than secondary, across all
Second, the first scenario received more positive evaluations than any of
the others. Notice that 42% of elementary teachers and 38% of
secondary teachers had a positive response to the first scenario" This
orientation to the table tells readers what each row and each column represents
and it also summarizes the main findings. Even if you think your table
is self-evident, you should always accompany any table with a descriptive
paragraph like this.
Try give both qualitative and quantitative data for any main finding.
If you have studied just one teacher, you can still say things like, "Mrs
Jones mentioned this concern on seven different occasions during the three
interviews I had with her. " Then you can give some examples of
the specific things she said. The reference to the number of times she brought
this up helps readers see how important it was to her, and the quotes help
people see the nuances of the issue as she perceived it. If you have conducted
an observation study, try to tally up the number of times you saw A occur
and the number of times you say B occur.
Address alternative interpretations of your evidence.
This helps readers evaluate your particular interpretation in relation
to other plausible interpretations. For instance, suppose you did interview
teachers a week before an annual test and you asked them about how the test
influence their teaching. One interpretation of your findings would be that
everything they said was influenced by the time of year. But suppose you
had the foresight to ask them about earlier curriculum decisions. Then you
could have a section in your report that described what they claimed they
had done months earlier. You might still need to concede that the anxiety
they expressed was time-dependent, but you might be able to make a plausible
argument that their decisions even months earlier were anticipating this
Be descriptively precise with your language.
Try to be refer precisely to what you actually saw or heard, not what you
inferred from it. For instance, don't say, "teachers were worried
about their success," say "these teachers
were worried about their success." In fact, you can go one better
and say, "These teachers claimed they
were worried about their success." Always make it clear
that you are referring not to teachers in general, but to the specific teachers
in your study, and that you are relating what they said, without inferring
that it was necessarily true. These careful sentences remind readers
that all you have is what a specific group of teachers has said; you are
not a mind-reader and you can't be absolutely sure other teachers would have
said the same thing.
OK. Let's say you actually generated a summary of the patterns in
your data. You're done. Now you get to close out the paper.
Part IV: Discussion and Conclusions
There are many ways of closing out a paper. Here are some basic distinctions:
A summary needs to give a brief repeat of your entire argument. A
summary should repeat your introduction, saying what the issue is, why you
did the study, what it's major weaknesses are, what you think you learned
and so forth.
A conclusion doesn't have to summarize the entire argument, it can just
say what you think you learned from the study. It is an extension of
the findings in the sense that the findings are more descriptive and the
conclusions are more interpretive. If a finding says "Mrs Jones
felt overwhelmed by the new requirements," your interpretation might
say something like, "these new requirements appeared to create a great
deal of stress for Mrs Jones, and her response raises important questions
about whether we can expect other teachers to respond appropriately."
A discussion is a broader ranging treatment of the issue. It allows
you to summarize and to conclude, but it also allows you to discuss the issue
more generally, to raise new questions, to say how your study fits into the
pattern of other literature out there, to maybe suggests new research directions
or new policy directions. My human caretaker tells me she likes discussions
best, because they allow the most flexibility. But she also suggests
that novices should begin by writing summaries and conclusions because you
can't really write a discussion until you have those in mind anyway.
Part V: References
Whoo-hoo! You've got a first draft of the whole dang thing. Go
outside and breath some fresh air. If you read my page on handy software,
you should now be able to use EndNote to format your bibliography. Then
you can call it a day and call the waiter to bring you something special.