Abstracts for scientific articles. (Writing Professionally).

Scientists and other technical researchers must prepare summaries, or "abstracts," of their work for a variety of purposes, most notably as nutshell renditions of journal articles. Besides writing abstracts for articles submitted for publication, scientists also prepare them for conference papers, poster presentations, formal reports, graduate theses, and even lectures. A common way to classify abstracts is by whether they merely describe the overall purpose and methods of the research presented in the document, providing a sense of its main topics, or, instead, inform readers of specific details of the research, especially the results and conclusions. Writing an abstract, whether descriptive or informative, is in principle rather straightforward, but in practice something of an art that matures with experience. For an outstanding and extensive discussion of article abstracts, including their content, range of format, and typographical features, see biologist Antoinette M. Wilkinson's The Scientist's Handbook for Writing Papers and Dissertations (1991). One can also refer to publications that contain national standards, which can easily be located and purchased online, such as the BIOSIS Guide to Abstracts and the American National Standard for Writing Abstracts. The latter publication is prepared by the National Information Standards Organization (NISO). A standard that is developed and approved by NISO becomes an American National Standard once the American National Standards Institute (ANSI) verifies that the approval process has met the ANSI criteria (Council of Biology Editors, 1994; Ebel, Bliefert, & Russey, 1990). The focus here will be on a few key points, including the basic usefulness of abstracts, differences between descriptive and informative abstracts, and other basic elements that make abstracts critical tools for readers.

Purposes Served by Article Abstracts

Naturally, before writing the abstract for an article, one should look at the instructions and examples in the journal to which the article will be submitted. There are a few differences among journal abstracts, such as typographical features, length (generally a 100-to-250-word range), and internal format (e.g., enumerated sentences, subsectioning, outlining, use of keywords). The abstract for any article will, however, serve various purposes in the interest of readers:

* to indicate to readers quickly whether the full article would be useful to read;

* to be extracted (or abstracted) from the article for separate publication (e.g., by Biological Abstracts or Chemical Abstracts); and

* to provide terminology to assist in literature searches by individuals or by Literature retrieval specialists for indexes and computer banks.

Given that an article abstract may be read by many people who may not read the article itself, abstracts must be written to stand alone, independently of the article, and still make sense to the reader. For indexing purposes, in some journals the abstract may incorporate the full bibliographic citation. Besides being brief and concise, a sound abstract must be accurate, precise, objective, and readable. The fact that these characteristics are not easy to accomplish is affirmed by the reliance of major abstracting services on professional "abstractors" in place of exclusive dependence on authors' own abstracts.

Descriptive Abstracts

A descriptive abstract, sometimes also termed a "topical" or "indicative" abstract, in effect acts as a paragraphic or prose table of contents for the article and typically emphasizes the research problem and methods. It is written about the article, rather than transmitting the information contained in it. It functions primarily to tell readers about the kind of information the article contains, focusing on the research problem and providing an abbreviated and indirect description of the methods. Since a descriptive abstract tells about the paper rather than encapsulating the actual results and conclusions, its sentences convey the different aspects of discussion, as in the following examples (Wilkinson, 1991):

* "The parameters affecting pollution from industrial solvents are discussed."

* "Workers' exposure to toxic agents is reviewed and case histories are given."

* "The temperature dependencies of luminescence intensity and duration are described."

* "A model for how such mutagens can be more reliably detected is introduced."

Wilkinson provides the following example of a descriptive abstract, which clearly focuses on the study's purpose (first sentence) and methods (middle sentences), with a reference to the discussion (final sentence): This study explores the influence of demographic factors on the attitudes of individuals living in non-metropolitan, natural-resource-oriented environments toward their community. Interviews were conducted in a rural area of Missouri, and data were subjected to factor analysis and analysis of variance. Strength of feeling scores were calculated. Hypotheses are offered regarding the nature of natural-resource-oriented communities and individual-community relationships. The investigator suggests implications of his findings for adult education programs.

For articles, research scientists only rarely write descriptive abstracts. The exceptions are abstracts prepared for papers that are descriptive, mathematical, or theoretical, or for review articles. Descriptive abstracts are written primarily for those who may be interested in retrieving the article rather than in getting the information from it--for example, librarians, bibliographers, and research scientists searching the literature.

Informative Abstracts

Unlike descriptive abstracts, informative abstracts report the details of the research and do not just describe what the paper contains. The content of an informative abstract should focus on the objectives, methods, results, and important conclusions of the research. All of these features (labeled with added italics) are included in the following example from Wilkinson (1991):

Some parts of the article should not be represented in the abstract, either because they provide no new information or because they take would take up too much space. The abstract does not need to include background, literature, or discussion, since these aspects of the paper will be covered sufficiently in the article itself. These principles are, however, violated in many abstracts. An author may waste space at the start of the abstract by repeating introductory or background information that is provided at the beginning of the paper itself, whereas that space is better used simply to announce the research objective. Similarly, references to the literature should be avoided, although sometimes abstracts may refer to earlier studies, either generally (e.g., "This phenomenon has earlier been reported by other authors") or by mentioning names of other investigators. If such references are really essential, they should be cited as briefly as possible.

The changes in adipose depot weight, cell size, cell number and body composition during pregnancy, lactation and recovery were studied in Osborne-Mendel rats fed standard or high fat diets [objective, materials]. Rats were killed on day 21 of pregnancy, after 21 days of lactation, and after 21 or 22 days of a post-lactational recovery period. Nonpregnant control groups were killed at the beginning and at the conclusion of the experimental period [methods]. The high fat-fed, mated group was always fatter than similarly treated animals fed standard diets throughout pregnancy and lactation. However, by the end of the recovery period, carc ass composition of the animals fed high fat or standard diets and nonpregnant groups were not statistically different. The weight of the parametrial, retroperitoneal and subscapular depots was higher in the high fat--fed animals at the end of the recovery period, and in the latter two pads, this increase was statistically significant [results]. Thus, despite the extensive lipid mobilization that occurs during lactation, the high fat--fed animals appear to be predisposed to postpartum obesity [discussion, conclusion].

In the process of writing the abstract, every word should be examined carefully to see if more words than needed are being used to narrate the story of one's research. Also consider the fact that editors and reviewers, seeking initial orientation, are likely to first read a paper's abstract and that this first impression is "perilously close to a final judgment of your final manuscript" (Day, 1994). In an informative abstract, the closest attention must be given to the language and the sentences that report the research results, understandably the most important and probably the dominant part of the abstract because they hold the new scientific information.

Closing Remarks

The goal in writing the abstract, as Wilkinson (1991) states, "is to convey as much new information as possible to scientists in the same or related discipline in as few words as possible-- accurately." Since the objective is the most content possible per word, the abstract is the densest part of the article, Therefore, authors must first reduce drastically the parts to be included, and then they must eliminate wordiness and repetition. It should also be underlined that there are other elements involved in the process of writing abstracts that are not covered here, but that require attention. These aspects include certain structural and typographical features, the importance of using third person (versus first), and the consistent use of past tense (with few exceptions). It is worth reiterating how critical it is that authors write the abstract with especially close attention and utmost care, since for most researchers the abstract is the primary source of information in their discipline. Physicist Michael Al ley (1996) quotes Winston Churchill as having said: "Please be good enough to put your conclusions and recommendations on one sheet of paper at the very beginning of your report, so that I can even consider reading it."

References

Alley, M. (1996). The craft of scientific writing, 3rd ed. New York: Springer-Verlag.

Council of Biology Editors, Style Manual Committee (1997), Scientific style and format: CBE style manual for authors, editors, and publishers (6th ed.). Chicago, IL: Council of Biology Editors.

Day, R.A. (1994). How to write and publish a scientific paper (4th ed.). Phoenix, AZ: Oryx Press.

Ebel, H.F., Bliefert, C., & Russey, WE. (1990). The art of scientific writing: From student reports to professional publications in chemistry and related fields. Weinheim, Germany; New York: VCH.

Wilkinson, A.M. (1991). The scientist's handbook for writing papers and dissertations, Englewood Cliffs, NJ: Prentice Hall.