Thursday, February 10, 2011

Writing good papers

Writing good papers is matter of having the contents before ("rem tene, verba sequentur").   So I give for granted that you chose a topic which is important (not just because fashionable, but because you humbly assimilated what your community of scientists, and the best among them, thinks it is important) and you have done the right amount of work with solid methods.

However, here, I want to support the idea that a scientific paper has its own structure. The Nature journal has a page with several useful references (here, last accessed February, 9, 2011, as well as all the links below). Other and many resources can be easily found googling "scientific writing" or similar keywords (and you find, for instance this).
However, I am summarizing here a contribution brought to us a few years ago by our friend  Julie Campbell.
She reminded that the general organization of a scientific paper has some standard parts:

Introduction and Literature Review (Previous Studies)
Approach and Method(s)
Results and Analysis**

Stanley Malloy wrote a small guide where he defines the previous parts as:

"Abstract: An abstract is a succinct (one paragraph) summary of the entire paper. The abstract should briefly describe the question posed in the paper, the methods used to answer this question the results obtained, and the conclusions. It should be possible to determine the major points of a paper by reading the abstract. Although it is located at the beginning of the paper, it is easiest to write the abstract after the paper is completed.

Introduction: The Introduction should (i) describe the question tested by the experiments described in the paper, (ii) explain why this is an interesting or important question, (iii) describe the approach used in sufficient detail that a reader who is not familiar with the technique will understand what was done and why, and (iv) very briefly mention the conclusion of the paper.

Approach and Methods: Or the Materials and Methods section should succinctly describe what was actually done. It should include description of the techniques used so someone could figure out what experiments were actually done. The details of a published protocol do not need to be reproduced in the text but an appropriate reference should be cited – e.g., simply indicate “were done as described by Hughes et al. (4)”. Any changes from the published protocol should be described. It is not appropriate to indicate volumes of solutions added – instead indicate the relevant information about the experiment such as final concentrations used, etc.

Results: Begin each paragraph with an opening sentence that tells the reader what question is being tested in the experiments described in that paragraph. Write the opening sentence in bold font for emphasis. (Sometimes a complete sentence is used and sometimes a short phrase is used – either style is OK but the style should be used consistently throughout the manuscript.) Any results that include multiple data points that are critical for the reader to evaluate the experiment should be shown in tables or figures. However, the results should be summarized in accompanying text. When referring to a particular table or figure, they should be capitalized (e.g., Table 1, Figure 6, etc.) The text of the Results section should be succinct but should provide the reader with a summary of the results of each table or figure.
Not all results deserve a separate table or figure. As a rule of thumb, if there are only a few numerical results or a simple conclusion describe the results in the text instead of in a table or figure. Your paper should focus on what worked, not things that did not work (unless they didn’t work for reasons that are interesting and provide insights).

Analysis (Discussion): Do not simply restate the results — explain your conclusions and interpretations of the Results section. How did your results compare with the expected results? What further predictions can be gleaned from the results?

Acknowledgments:  Projects or Institution that financed the research, reviewers that helped. People with whom the topics treated were discussed appear cited here.

Reference lists: Like citations, a variety of reference formats are used by different journals. For an example of a commonly used example, see “Instructions to authors” on ASM web site ( or examples from published manuscripts. "

Citations: It is essential to credit published papers for work mentioned in your manuscript. - RR Note: There are no many other rewards than being cited. So as you like to be cited, do not forget those from whom you learned something - There are a variety of ways of citing references in the text – the style used depends upon the policy of the journal. In text citations should refer to reference list. Do not rewrite title of references in text. "

Julie provided a nice hand-out in her lecture and here it is an excerpt regarding figures: create Tables and Figures that stand-alone.

"The tables and figures that you place in a scientific paper should be able to be understood mostly, if not completely, whether or not your reader reads the accompanying text. Therefore, contents should be clearly labeled and units of measure specified. Use only abbreviations of terms that will be clearly understood by your readers; otherwise, spell out terms. Keep in mind that even though you have defined abbreviations of terms in your text, including equation variables, the table or figure will not stand alone unless you redefine them in your table or figure, being consistent with the abbreviations that you have used in your text.

For figures, specifically plots, there should be a key for symbols used. Also, there should be fully explanatory axis labels that include the units of measure. Recall that the y axis
often begins at zero, but it if does not, you should consider, depending upon who your readers are and the nature of the quantity, calling that fact to your readers’ attention in your caption."

Example: From Gutierrez-Magness and McCuen.

(Notice that the graphic key is easier to comprehend than a verbal description of the line types and/or symbol types would be. Regarding the issue of standing alone, this figure basically satisfies some of the requirements for standing alone. It does, however, need further notation explaining what kind of data has been analyzed, as well as the region, the year(s), and the type of instrument used.)

Example: From “Distributions-Oriented Verification of Ensemble Streamflow Predictions” by A. Allen Bradley, Stuart S. Schwartz, and Tempei Hashino, Journal of Hydrometeorology 5 (2004): 532-545.

(Note the key for the symbols, as well as the notation that provides information about the data in question and its units, as well as the year. Note also the somewhat unusual abbreviation of September, which is traditionally abbreviated, Sept. In this case, the logarithmic scale must start at 1000. In general, this figure satisfies the requirements for standing alone.)

Example: From “Heavy Rainstorms in Chicago: Increasing Frequency, Altered Impacts, and Future Implications,” by Stanley A. Changnon and Nancy E. Wescott, Journal of the American Water Resources Association 38.5 (2002): 1467-1475.

(Note the detailed information in the title that aids the table in standing alone. Note that in the labels, the numbers 2 through 10 have been spelled out, while those numbers greater than ten are given in their numeric form. Many style guides request this differentiation. In general, the rule goes as follows: “Technical quantities of any amount are expressed in numerals. Nontechnical quantities of fewer than 10 are expressed in words. Nontechnical quantities of more than 10 are expressed in numerals” [Markel p. 644]. Needing clarification, however, are the last two rows—the third row indicates the maximum number of storms in a year, and the fourth row indicates that the specific year in question is 2001, but this information is essentially part of row three.)

For more information, see Ch. 14 in Markel on “Creating Graphics,” especially, “Creating Effective Tables,” pp. 333-335.

And more from Julie's handout

I. Notes from Article Excerpts: “Advances in the Use of Observed Spatial Patterns of Catchment Hydrological Response,” by Rodger B. Grayson et al, Advances in Water Resources 25 (2002): 1313-1334.

--Collective nouns—use a singular verb when the parts of the group act as one unit; use a plural verb when the parts of the group act individually. Example: the jury have/has reached a verdict. Note: the expression “a number of” means “several” and needs a plural verb.

--Parallelism—the repetition of a grammatical structure. It suggests similarity between ideas and creates symmetry and balance. A parallel construction repeats an identical grammatical pattern within the same sentence, paragraph, or passage. Sentence examples: a) In this study we will focus on examining the parameters of . . . , calculating the difference between . . . , and explaining how the different patterns signify . . . .
b) The first goal is to acquire the necessary data. The second goal is to calculate . . . . Finally, we hope to show . . . . c) Our study encompasses the derivation of . . . , the illustration of . . . , and the conclusions that we draw . . . .

--Revision of Second Sentence, Intro. Paragraph:
Original—“But these have served more to consolidate the work of the 1990s and propose new methodological advances, rather than focus on new data sources.”
Revision—“These, however, have served more to consolidate the work of the 1990s and (to) propose new methodological advances than to focus on new data sources.”

--Revision of Third Sentence, Intro. Paragraph:
Note: the comma is required before “and” in this case because the author is combining two “independent clauses” with a conjunction.
“Nevertheless, the calls of the 1980s and early 1990s for more research into representing spatial heterogeneity, the collection of data sets for the testing and development of distributed models, and methods how to best deal with issues of scale, have to some extent been heeded, and it is these on which we will focus in this paper.”

--Revision of Results Paragraph:
--See “trade-off”
--Why is it incorrect to place a comma before “and” in the last sentence?

--Revision of Summary Paragraph:
--How could we rewrite for parallelism here?

II. Punctuation Review, American English

--Most common punctuation errors:
A) The comma splice—links two independent clauses (clauses that could stand alone as sentences) with only a comma.
Incorrect: This report was distributed widely in the U.S., its findings were considered groundbreaking.
Correct: This report was distributed widely in the U.S., and its findings were considered groundbreaking.
This report was distributed widely in the U.S.; its findings were considered groundbreaking.
This report was distributed widely in the U.S.: its findings were considered groundbreaking.

B) The run-on or fused sentence—contains two independent clauses with no dividing punctuation.
Incorrect: The internal temperature reached 30 degrees we stopped the experiment at that point.
Correct: The internal temperature reached 30 degrees, so we stopped the experiment at that point.
The internal temperature reached 30 degrees; therefore, we stopped the experiment at that point.
The internal temperature reached 30 degrees. We stopped the experiment at that point.
Also: When the internal temperature reached 30 degrees, we stopped the experiment.

C) The sentence fragment—is an incomplete sentence; some grammatically necessary part is missing.
Incorrect: The drop in temperature caused by a leak in the canister.
Correct: The drop in temperature was caused by a leak in the canister.

Note: In the “Appendix: Reference Handbook,” in Markel (see bibliography below), there is an excellent section on “Editing Your Documents” that covers most aspects of punctuation and grammar.

III. Use of Acronyms and Initial Abbreviations

--An acronym is a word consisting of initials and pronounced as a word, for example, NATO—North Atlantic Treaty Organization, SAR—Synthetic Aperture Radar, or GRASS—Geographic Resources Analysis Support System

--An initial abbreviation consists of the first letter of each word in a phrase or name, for example, GPS—global positioning system or GIS—geographic information system.

IV. Use of Prepositions

-- Prepositions are the words that link and establish specific relations among words or group of words. In some cases, usage is highly idiomatic—in other words, there are few clear rules!
A. Prepositions of Place (literal and figurative)
a. in—in the picture, in the photo, in the paragraph, in the sentence, in the equation, in front (of), in the process (of), in New York, in fact, in place (of), in existence
b. on—on the map, on the page, on the subject, on the edge, on purpose, on top (of), on site, on the contrary
c. at—at the corner, at the intersection, at a glance, at least, at these coordinates, at one time, at work

B. Prepositions of Duration
a. for—for a while, for three days, for a week
(also: a need for, a desire for, for that reason, for instance)
b. while—while we calibrated the...., while I was watching...., while we measured....
c. during—during class, during the experiment, during the day

C. Prepositions of Time and Date
a. in—in a couple of weeks, in February, in time (for)
b. at—at seven o’clock, at Christmas, at sundown, at first
c. on—on Christmas day, on Tuesday, on schedule

Common Words and Phrases Used as Prepositions: about, above, according to, across, after, against, ahead of, along with, among, around, as, as for, at away from, because of, before, behind below, beneath, beside, between, beyond, by, concerning, due to, except, except for, for, from, in, in addition to, in back of, in case of, including, inside, inside of, in spite of, next to, of, off, on, onto, on top of, other than, out, out of, outside, over, past, regarding, through, to, toward, until, unlike, upon, up to, with, within, without

Note: See sections on prepositions in books listed in the bibliography below.

V. Use of Idiomatic Expressions

--Idioms are words, phrases, or expressions that are either grammatically unusual or they convey a meaning that is different from the literal meaning of the term or phrase. Examples: “the big picture”, “to warm up to someone”, “a cold fish”...

Note: A convenient source for help with idiomatic expressions in English may be found at

These sorts of expressions are interesting and useful to know in general, but have limited use in scientific and technical writing because of their lack of univocal meaning in some cases.

VI. Use of Capitalization
--For help with standard practices of capitalization in English, see one of the recommended handbooks in the bibliography below.

VII. Use of Precise Terms to Avoid Ambiguity
--When writing a technical paper, it is critically important to be as specific as possible. Avoid using imprecise words and terms (e.g., a lot, some, a few, an insignificant amount), and if you must use them, be sure to follow them with the precise explanation of what you mean. Use precise words and terms (e.g., 300 meters to the west, 40 degrees warmer than the previous measurement, two key points) whenever possible.
--See pp. 251-251 and pp. 625-627 in Markel, as noted in the bibliography below.

VIII. (Addendum) Writing “Mathematical English”
--When you incorporate an equation into your writing, you should remember that the parts of the equation function as parts of speech. Therefore, the equation and its parts require appropriate use of punctuation, as well as grammatical consideration.

--Variables used in the equation all need to be defined.

Typical errors include the following:
-- Misuse of the colon (:) to introduce every inset equation, no matter the grammatical structure of the sentence which incorporates the equation.
--Capitalization of linking words following the equation that are actually already in the middle of a sentence, such as where, therefore, thus, so, if, and that.
--Failure to follow the equation with a comma preceding the “where...” phrase because in this case, the “where...” phrase is a nonrestrictive element of the sentence.

Example: From “Recent Developments in Statistical Time Series Analysis: Examples of Use in Climate Research” by F. Godtliebsen, L.R. Olsen, and J.-G. Winther, Geophysical Research Letters 30.12 (2003): 56-1—56-4.

A relevant nonparametric regression problem for this is to attempt to use data of the form
where m(x) is the target curve. Here, we assume that the xi are equally spaced on the range of x, that m is smooth and that the εi are independent Gaussian variables with mean 0 (which makes m the regression curve of yi on xi) and variance Var (εi) = .

(Note that there is no colon introducing the equation, and “where” is not capitalized. A comma, however, should be placed after the equation.)

Example: From “Accuracy Evaluation of Rainfall Disaggregation Methods” by
Angelica L. Gutierrez-Magness and Richard H. McCuen, Journal of Hydrologic Engineering ASCE 9.2 (2004): 71-78.

(Note that there is no colon after “by.” Note also that elements which could stand alone as “mathmatical sentences” are set off with semicolons.)

For more information on “Mathematical English,” see the American Institute of Physics Style Manual at


--Anderson, Paul V. Technical Communication: A Reader-Centered Approach. 5th ed. Thompson/ Heinle, 2002.
--Fulwiler, Toby and Alan R. Hayakawa. The Blair Handbook. 4th ed. Upper Saddle River, NJ: Prentice Hall, 2003.
--Hacker, Diana. A Writer’s Reference. Bedford Books, 2003.
--Hacker, Diana. A Pocket Style Manual. Bedford Books, 2003.
--Markel, Mike. Technical Communication. 7th ed. Boston: Bedford-St. Martin’s, 2004.
--Strunk, William and E.B. White. The Elements of Style. 4th ed. Pearson Higher Education, 2000. See also:

External links

See the beautiful presentation by Jeffrey McDonnell

And find here the clarification of co-authorship responsability

Here the Thomas Hengl guide to write a paper

The mononota song as a paradigm for writing a good paper


  1. I recently found this site:
    that is a sort of blog on scientific writing.

  2. Particularly interesting the links @