The Scientific Paper: A Template with Tips for Working with LATEX

D.P. Jackson, K. Browne, and J.Q. Student

Department of Physics and Astronomy, Dickinson College, Carlisle, Pennsylvania 17013 USA
(Dated: November 11, 2004)
This paper should serve as an outline for writing scientific papers. It contains all of the important
sections that should be included in scientific paper as well as descriptions of what should be included
in each of these sections. It also contains some useful tips on how to use LATEX to write scientific
papers. The easiest way is to use this document as a template and insert your text and figures
as described in the text below. This section is the abstract. The abstract should contain a brief
description of the project including relevant description of the problem, data collection procedures,
and a summary of results as well as a brief description of how this information fits into the overall
field. The abstract may contain equations like E = E0 cos(k · r − ωt + φ) and, as you will notice,
inline equations look much better in LATEX than they do in MS Word. An abstract is usually quite
short. Often, the length is limited to between 200 and 400 words.

I. INTRODUCTION make this clear, no matter whether or not you are right,
you have not done your job as an author and will find
This paper contains a general outline of the informa- that you have a hard time convincing anyone that what
tion that should be included in a scientific paper. It you have done is valid. Finally, every paper should end
provides a good template within which you can easily with a references section. A scientific paper without any
write a paper. When you start out writing papers, you references, indicates that the author believes that every
will likely include most of these sections and utilize this thought conveyed in the paper is original. Any informa-
fairly standard format. As you gain experience, you may tion that you obtain from another source should be cited.
choose a different ordering or different sections as you find The only exception is for material that is considered com-
appropriate. Remember this is just a template to help mon knowledge. As a student, your common knowledge
you get started. You will have your own style of writing. will often be somewhat more limited than the average
Your audience and the content of your paper should be author in a scientific journal. As such, you will often
the most important guiding influence when writing any reference information from class notes or textbooks that
paper. The writing process will go much more smoothly other authors may not. When in doubt, make a reference.
if you take some time to answer a few questions before This eliminates any possibility that you will be accused
you begin writing. For example, before you begin writ- of plagiarism, a very serious transgression indeed.
ing, ask yourself, “Who is my audience?”, “What do I An introduction generally contains a brief introduction
want them to get out of this paper?”, and “What are the to the material that will be presented. Relevant informa-
most important ideas to convey with this paper?” There tion includes a clear enunciation of the questions that will
are lots of other questions you could ask, but these three be addressed in the paper, background information rel-
will help you generate a document that is pitched at the evant for understanding the paper, basic theory needed
right level and contains information that is useful to your to undersand the contents of the paper, etc.
audience. It is important to take into account your audience when
You should keep in mind that a good scientific paper writing the introduction. The purpose of an introduction
always introduces the reader to the subject material and is most often to give your audience enough information
provides the appropriate background information that so that they will be able to understand the rest of your
the author thinks the reader will need. A good scien- paper and put it into a larger context. Depending on
tific paper will always make the experimental, compu- your audience, this context may vary. For example, if
tational, or theoretical methods clear enough so that a you are preparing a paper with other physics students
competent reader should be able to reproduce the work. in mind as the audience, you will write the introduction
A clear description of how any data was collected should so they see how their previous physics knowledge will
be included as well as a description of the raw data, gen- be useful in understanding this paper. If on the other
erally in graphical format. Any analysis performed on hand, you are writing this paper for a narrow selection
the data should be outlined clearly. Analysis and con- of researchers, you will not need to include as much in-
clusions drawn from the analysis should generally be de- formation. Rather, you will present them with enough
scribed separately from raw data. A paper should end information so that they can see how this paper fits in
with a set of conclusions based on the analysis. with relevant research.
It is the responsibility of the author to carefully lead Because you may not be familiar with LATEX, you will
the reader from the experimental design through the con- undoubtedly have many questions about how to do cer-
clusions tying each piece together. For example, it should tain things. This document will serve as a template
be clear to the reader explicitly how your analysis leads for producing professional looking papers in LATEX. Be-
from your raw data to your conclusions. If you do not fore you begin to modify this document, make sure you
 2

have a copy of it saved somewhere so that you can re- III. EXPERIMENTAL METHODS
fer back to it if needed. In addition, there are lots of
places to get help with LATEX (including asking profes- This section is often called experimental design or
sors in physics and math), but a useful place to begin is methods. It contains information about how you went
to visit http://www.giss.nasa.gov/latex/. All the com- about your experiment. The purpose of this section is
puters in the physics labs are equipped with a program to convince your reader that your experimental methods
called TeXshop that runs the LATEX engine. were sound and thorough. That said, if you have made
If you have any questions about the appropriate experimental errors that you did not correct, or if you
style for a scientific paper, you should refer to the made errors along the way it is your responsibility to
American Institute of Physics (AIP) Style Manual at report them here. If you do not clearly report your ex-
http://www.aip.org/pubservs/style/4thed/toc.html. perimental methods, you run the risk of having someone
else try your experiment and get other results. This then
brings into question the validity of your conclusions and
II. THEORY
your reputation as a scientist. In addition, if you made
errors along the way that you corrected before collecting
Often, if the theory needed to understand a paper is your final data, it may be worth presenting them here so
somewhat extensive, a separate section containing a de- that others can benefit from your mistakes.
scription of the theory will be presented. This section Often you will include a diagram of the experiemtal
should contain enough theoretical detail to make it pos- setup. This is shown in Fig. 1 (note that I didn’t have
sible for a member of your target audience to be able to to worry about the figure number). Of course, LATEX is
reproduce any results you come up with. Obviously, the a typesetting program and is not a graphics program,
amount of detail that you include will depend on space so you will have to make your graphics in a different
constraints and the expected level of expertise of your program, say, Adobe Illustrator or Xfig. Fortunately,
audience. including the figures into a LATEXdocument is a pretty
In the context of a paper written by an undergraduate simple matter.
for a class, you should include all non-obvious steps and
be sure to reference material that is not “common knowl-
edge.” If you just learned the material in a class, you
should include references to where the basic derivation
comes from. If you start with a non-trivial expression
that you had to look up somewhere, either in a book,
a paper, or your notes, you should definitely include a
reference.
All equations should be incorporated into the text us-
ing a program designed to properly format equations.
LATEX is designed to handle equations, equation num-
bering, and cross referencing to sections, equations, and
figures with ease. In fact, you do not need to worry about
numbering any sections or equations, that will be done
for you automatically. You may want to refer back to
an equation, figure, or section. To do so, you simply la-
bel the appropriate item and then refer back to it when
needed. For example, to refer back to the introduction,
I can type something like “this is discussed in Sections
\ref{sec:intro} and \ref{sec:theory}” to get “this is dis- FIG. 1: A sample schematic diagram for an experiment.
cussed in Sections I and II.” Notice that I didn’t have
to worry about the sections numbers. This is a life saver Any diagram you include should contain a fairly de-
when you are writing a paper with lots of equations and tailed figure caption. A good rule of thumb is that if
figures. Equation numbering is automatic only in “dis- someone reads the abstract and looks at all the figures
played math” mode, which is illustrated here, and captions, they should have a reasonable idea what
your paper is about. While this isn?t always possible, it
E = E0 cos(k · r − ωt + φ), (1)
is a good thing to shoot for. That said, this document
and here, doesn?t even come close to meeting that requirement,
but it also isn?t so much a scientific paper as a how to
B = B0 cos(k · r − ωt + φ). (2) manual on writing one.
As mentioned before, you should include enough in-
Of course, I can easily refer back to Eqs. 1 and 2 without formation in your experimental design to make it possi-
having to remember the numbers. ble for someone else to reproduce your experiment. You
 3

should generally outline what you did with enough detail thing in the world to do in LATEX, but they’re not all that
so that it is clear how you setup your experiment and difficult either. Basically, if you have to make a table, it
how you collected your data. is best to look for some help in a book or online and then
It is particularly important to include anything out fiddle until you get it looking the way you want. Table I
of the ordinary. Often we make experimental errors in shows an example of a table that compares two sets of
our setup. It isn’t fun, but it happens. If one clearly temperature data. As you might expect, simpler tables
articulates her setup, it is possible for others to identify are easier to make.
these often subtle experimental errors.
TABLE I: Conventional and syringe thermometer readings.
The highest and lowest readings were used for calibration.
IV. RESULTS
Conventional Syringe
Your paper should contain a section describing your
20◦ C 1.8cc 20◦ C
raw results. Often this will be done by including graphs
27◦ C 2.4cc 28◦ C
and/or tables of data. This data should generally not be
heavily processed. Rather, one should include results in 42◦ C 3.9cc 46◦ C
an understandable format that are a good representation 55◦ C 5.0cc 59◦ C
of the data obtained by your experiment or computation. 67◦ C 6.0cc 72◦ C
You will have a chance to show processed results in the 84◦ C 7.0cc 84◦ C
analysis section, but in this section you need to present
the reader with your raw data so she can clearly judge
the quality of your analysis and conclusions. In general, you should never include a table in a paper
Often you have far too much data to include it all. when a figure/graph will do a better job. It is quite rare
In this case, you will include a sample of raw data with to see tables in scientific papers. You should never in-
tables or graphs containing straightforward compilations clude a long list of data or an excerpt from a spreadsheet
of this data. unless the particular values in the list are very important.
It is generally best to make all figures only a single Long lists are hard to read and generally confuse or bore
column width, as shown in Fig. 2. You generally have your reader.
three choices of where to place the figures in LATEXḢere Most often tables are used to show a few numbers de-
(meaning right here if possible), top (meaning top of the rived from a larger dataset. This is a good use of tables
page if possible), and bottom (meaning bottom of the but should generally occur in the analysis sections be-
page if possible). You may still have to do some fiddling cause the numbers are derived from the data.
at the end to get them exactly where you want them. Here is another table. We can reference this table in
the same way mentioned in Section 2. Table II shows a
slightly simpler table.

HaL
TABLE II: Force, area, and pressure data for the experiment
0.1 shown in Fig. 1 and described by Eq. 2. Agreement is typically
M0
€€€€€€€€€€ =2.0 within five percent.
0.05 H0

F3 Piston 1 Piston 2
Avg. Force (N) 4.40 2.25
Area (cm2 ) 6.16 2.25
0.01
F/A (N/cm2 ) 0.714 0.717
0.005

1.5 2 3 5
L V. ANALYSIS

After you have clearly described your results, you will

FIG. 2: Force as a function of length for a particular exper- describe how you will analyze these results, that is, how
iment. The dashed curves represent the nonmagnetic case you will process the data you collected to obtain informa-
while the solid curves show the magnetic effects. tion that will help you answer the questions you brought
up in the introduction.
There are also times when it is appropriate to include a It is critically important that the analysis section of
table of data. Unfortunately, tables are not the simplest a paper is clear. Your job in the analysis section is to
 4

convince the reader that the methods you used to get Including References
from your results to your conclusions are sound. If your
analysis section is incomplete or unclear, your reader may You must also include a references section in any sci-
not trust the conclusions you draw. entific paper. To omit the references section is to almost
This is another section where you will often have equa- certainly commit plagiarism. As mentioned before, you
tions, graphs and tables. Remember that whenever you should include references whenever you have used infor-
use an equation, graph or table, it should be referred to mation from another source. This might be a professors
in the text. Any equation, graph, figure, or table should notes or a textbook. As you advance in your studies,
fit into your explanations. If you include a graph but your references will come more and more from journal
make not mention of it in the text, the graph either has articles since these articles generally present more recent
not reason to be included, or you have omitted important results.
information from the text. In LATEX, references are handled very easily in a sec-
tion called “thebibliography.” Thus, you won’t actually
make a section called references, you will have something
VI. CONCLUSION called thebibliography. All you do is add a “bibitem” to
thebibliography and give it a label. Then, whenever you
Your conclusions section should be brief, but long want to refer to it, you use a \cite{} command. The order
enough to refocus the reader. The conclusions section in which you put the items in “thebibliography” is the
describes your assertions based on your data. In essence, order of the numbering of those items. Therefore, make
it contains the answers you?ve come up with for the ques- sure you put these items in the order that they appear in
tions you asked in the introduction. your paper. Here is an example of a book citation [1], an
You should also make it a point to place your con- article citation [2], and a comment that might make an
clusions within a context. That is, you should discuss important subtle point but one that would detract from
the possible implications of your conclusions or how they the main text [3].
might be relevant to other researchers. This is often hard That is basically all the sections that are normally in-
to do as a student, but not impossible. Some questions cluded in a scientific paper, but there are still some issues
you can keep in mind when writing this section are. Why that might help you regarding LATEX. I have put these
are these conclusions important? Who might these re- into an appendix as an example of how you might use an
sults affect? What could these results be useful for? appendix to put material that is essential to include but
It is important to keep in mind that you should not inhibits the flow of the paper.
overstate your conclusions. A common error authors
make is to over generalize ones conclusions. For exam-
ple, if I find that a particular type of crystal behaves Acknowledgments
non-linearly within certain parameters, it is an overgen-
eralization to conclude that all crystals of that structure You should always have a short acknowledgements sec-
will behave the same way. If the author suspects this to tion. This is where you thank people who helped you
be the case, she can state her prediction, but should not with the project. These can be people that assisted with
assert that it is a fact just because she has a hunch based construction, people you talked with that gave you good
on her experiment with this one crystal. This leads us to ideas, people you had an email correspondence with, basi-
a final sections that you may or may not want to include. cally anyone that contributed in some way to the success
of the project. You would also list fundint agencies in
the acknowledgements section.
VII. SUGGESTIONS FOR FURTHER
RESEARCH
APPENDIX: MORE LATEX INFORMATION
This section contains a listing of the directions that
the author thinks it will be possible to extend this re- This appendix is here to give you a bit more of an
search. It can be a list of possible future experiments or introduction to LATEX. At this point, it is very short and
questions one might ask that are based on the results of only includes the most basic items, but I will expand it
the research presented. This section gives the author the in the future. If there is something you learned about
opportunity to be somewhat more creative. That said, LATEXthat was very valuable, please let me know and I
it should be clear in the paper that the statements made will put it in here.
in this sections are suggestions, conjecture and or gut re-
actions. It is good to include this kind of information,
because it helps one to refine her intuition and practice 1. Getting Started
asking interesting scientific questions. Often, this kind
of information goes in the conclusion or a section called The first thing you have to do is open TeXShop on one
“discussion.” of the macs in the lab and then open a tex file (this one,
 5

for example). Then you need to typeset the document. as F and a. This works fine except when there is no
Depending on the options in TeXShop, the file will com- non-math-mode character to make bold. An example of
pile and produce a PDF file that you can then read or a math-mode character that does not have a non-math-
print out. mode equivalent is ∇, obtained by typing \nabla. If you
want this as a vector operator and you want it to be
bold, you must use the command \bm (boldmath) to get
2. Fonts ∇. This allows you to write that in general,

In a typical scientific paper, you might want to use ∇ × A 6= ∇ · A. (A.2)

italics or bold fonts occasionally. In LATEX, these are ac-
Note the use of \times for ×, \cdot for · and \ne
complished by using the \textit{} and \textbf{} com-
for 6=. You might also be interested to know that
mands. The text you actually want italicized (or in bold)
unit vectors can be written using the \hat{} com-
would be placed inside the curly braces.
mand. For example, \$hat{\textbf{r}}$ results in r̂
and $\hat{\textbf{e}}_{\theta} results in êθ .
3. Math Mode
One more quick topic that is sure to be useful is how
to break equations. It is quite common to have equa-
tions that are too long to fit on a single line. In these
In LATEX, you enter math mode by typing $ and then instances, you must break the equation into multiple
you leave math mode by typing another $. Thus, to type lines. This is done using the equationarray environment:
an equation, you place it between two dollar signs. For \begin{eqnarray}· · ·\end{eqnarray}. In an equation-
example, typing $F=ma$ results in F = ma. Greek letters array, each line needs to be separated by \\ and items
are made by typing a backslash and the name of the greek surrounded by ampersands (&) will be aligned on sepa-
letter. For example, $\alpha-\beta+\gamma$ results in rate lines. Also, each line will be numbered separately un-
α − β + γ. Superscripts and subscripts are handled by us- less you specify\nonumber (which is typically what you’ll
ing ^{} and _{} in math mode respectively. For example, want to do). The following shows an example of a mul-
2
typing $A_{1}=e^{-x^{2}}$ results in A1 = e−x . tiline equation:
So far, all of these examples have been for inline equa- (
tions that occur right in the paragraph you are typing. Z
3η ḃ 2
More often, you will want to put equations on lines all Fj = dA (R − r2 ) + [Ψj (R) − Ψj (r)]
b3
by themselves with an equation number. This is called )
a displayed equation and is accomplished by using the 1 h 2 2
i
equation environment (an environment in LATEXis some- + µ0 Mjr (R) − Mjz (r) . (A.3)
2
thing that you begin and end such as the abstract en-
vironment that was used to create the abstract of this Notice that I have added some space in Eq. A.3 to make
document). Thus, to create a displayed equation that it look a little nicer. For those that really want to make
has an equation number, you type \begin{equation}, things look great, fine tuning math equations with a lit-
then your equation (and a label), then \end{equation}. tle spacing here and there can really make a difference.
Here is an example: In math mode, you can add space with the following
(1) commands: \, small space, \: medium space, \; large
dEm
Fm = − . (A.1) space, and \! negative small space. These can be quite
db useful in a number of situations. For example, compare
√
You will have noticed that to make the derivative in \sqrt{2}x
√ which gives 2x with \sqrt{2}\,x R Rwhich
Eq. A.1, I had to make a fraction. The fraction com- gives 2 x, or \int\int dx dy which gives RR dxdy
mand is \frac{}{} (in math mode); the numerator goes with \int\!\!\int \!dx\,dy which gives dx dy. The
in the first set of curly braces and the denominator goes in differences are subtle, but for those with a discerning eye,
the second set of curly braces. I also used the command it is wonderful to have such control over your equations.
\label{eq:deriv} in the equation environment so that Incidentally, the commands \quad and \qquad add even
I can refer to it simply by typing Eq. \ref{eq:deriv} larger and larger amounts of space.
to get Eq. A.1. Well, that’s all for now. I think that about covers the
You may also find the need to write vector equations. basics. LATEXis an extraordinarily powerful program that
There are different methods of writing vectors in a sci- is capable of probably anything you can imagine. How-
entific paper. Most textbooks and scientific papers opt ever, it is not always obvious exactly how to accomplish
to put vectors in bold: F = ma. This was accomplished what you want to do. Fortunately, most of the “basics”
by typing $\textbf{F}=m\textbf{a}$. Note that the are fairly easy and you should have no problem figuring
command \textbf{} essentially takes you out of math them out. For more advanced techniques, you may want
mode and places a regular boldface letter in the equation. to consult a book or one of the online manuals (there are
This is traditionally how vectors are written in textbooks lots of them). Have fun and let me know if you need any
with the corresponding magnitudes for F and a written help!
 6

[1] R. E. Rosensweig, Ferrohydrodynamics (Cambridge Uni- include. This is usually a comment about something very
versity Press, Cambridge, 1985), and references therein. subtle that might be important to include but generally
[2] D. P. Jackson, R. E. Goldstein and A. O. Cebers, Phys. gets in the way of the regular text.
Rev. E 50, 298 (1994).
[3] Here is an example of a comment that you might need to