Wade B. Worthen
Biology Dept.
Furman University
This document is designed to help you write a successful research paper in the biological sciences. The first section summarizes the structural and stylistic requirements for research papers submitted in biology courses. Most specifically, this includes Research and Analysis (BGY 222) and Research in Biology (502). However, your professors may expect this format in other courses, as well. Ask about the structure of research papers and lab reports when the assignment is made so you can get started on the right track.
There are several quality texts available that should be used as structural and stylistic guides. Use them! One book, McMillan's Writing Papers in the Biological Sciences, is often a required text for Research and Analysis. It highlights the structure of both primary and secondary research papers. There are excellent examples describing common mistakes. Another manual is Strunk and White's Elements of Style. Remember all those little points about English grammar that you learned in eighth grade and forgot by ninth? Well, they are all in Elements of Style. Remember this: the quality of the writing reflects the quality of the research! Clear, direct prose that communicates your ideas in a logical manner is rewarded. The greatest experiment in the world is worthless unless the reader understands what was done.
Here is an outline of the proper structure of an experimental research paper. Although the sections should be presented in this order, they are not written in this order. Suggestions for writing sequence follow later.
A. Title. The title should be informative, specific and short (13
word max., usually). These objectives are difficult to satisfy concurrently. Typically,
titles include: 1) the species studied (common and/or Latin names as space
permits), 2) the variables addressed, and 3) the site, if it is important.
Obviously, the space constraints will occasionally force you to exclude one of
these parameters. Here are some examples:
1. Phenotypic and demographic
variability among patches of Maianthemum canadense (Desf.) in central
New Jersey, and the use of self-incompatibility for clone discrimination.
Long! 23 words, but includes all three parameters. This is difficult to
shorten, because two different questions were addressed in one study.
2. Fish predation on Notonecta
(Hemiptera): relationship between prey risk and
habitat utilization.
Nice length (12 words), but site has been omitted. Good description of the problem
and mention of the organism (genus, in this case).
3. Foraging behavior of a western reef
heron in North America.
4. Foraging behavior and food of grey
herons Ardea cinerea
on the Ythan estuary.
These include all three aspects within the standard space constraint. However,
in order to fit the site into the title, the topic is rather general (foraging
behavior).
The title should reflect the thrust of the paper. If site is important (maybe
there are geographical differences in foraging behavior in these species), it
should be included. If the study is a broad-based analysis of several foraging
behaviors, then it is more appropriate to say 'foraging behavior' than to list
every variable studied. Site is typically NOT included in laboratory
investigations, as the purpose of the lab is to standardize (and thus exclude) the
effect of environmental variables…. So hopefully, one lab is as good as
another. Good titles are hard to write, and they take more time than you might
think.
B. Abstract.This is a concise summary of the paper. Ideally, it should be short (500 word maximum), and should include at least one sentence describing each of these topics:
Again, the space limitations may force you to be selective. In addition, the methods and result may be difficult to describe completely in single sentences, and may require a larger fraction of the space budget. Certainly, the results are the most important part of the abstract; they represent the 'meat' of the experiment and cannot be slighted. However, you must also include a conclusion sentence; what did the results mean? If published, the abstract will appear in citation sources such as Biological Abstracts and Science Citation Index. It is the first thing someone will read, and it must be descriptive and interesting! The abstract demands clear, direct writing. When readers finish the abstract, they should be so intrigued by the experiment that they decide to read the entire paper. What search strategy do you use when search for articles? You enter keywords and then scan the list of article titles that appear. Like a fish at bait, you 'nibble' at an interesting one by reading the abstract. Here is where the author 'sets the hook'. If it's interesting, you read the article. If it is not, the author has lost you and you start to nibble on other titles again. Abstracts are also very difficult to write; it will take more time to write than any other paragraph in the whole paper.
C. Introduction. The introduction serves two functions. First, it provides the reader with the background information relevant to your experiment. Second, it presents the objectives of your study. These two functions are directly related; the background information that you provide should justify your experiment. After reading the background, the reader should understand why your question is significant. To write a good introduction, think of a funnel. Start with a broad background statement that provides some common ground for readers with different levels of expertise. Then, develop the information in the field that is important to your topic, focusing in on the objectives of your study. Try and maintain the flow from broad to specific.
Perhaps your paper is on foraging strategies of insectivorous birds. You might start off with a general statement on foraging strategy, and then highlight some of the relevant theories applied to animals, in general. Then you might focus on the particular developments and applications of avian foraging behavior. Finally, you could specify the energetic constraints imposed on insectivorous birds, and discuss this material. This section should conclude with some indication about the gap in our knowledge in this area. You then present your objectives, showing how your study attempts to fill this gap. At this point, the reader knows how your study relates to the field and understands why your question needs to be addressed.
There are space constraints on the introduction, too. This space constraint may alter your presentation. You may not be able to start out as broadly as you had intended, or the transition from background to objectives may need to be more direct. Remember: the introduction must demonstrate a logical progression to your objectives. This demands logical transitions. The topics in your introduction must be linked effectively so the reader can follow your argument. Excess information or tangential paragraphs will throw your reader off the track. Don't use the introduction as an information dump to show the reader how much you found on a topic. Show the reader you understand the relevant issues in a field and know how your study complements this information. Typically, the final paragraph of the introduction contains your “purpose statement”. It describes the relationship you are testing, and it may even give a very brief (one sentence) synopsis of what was found. So, the introduction “funnels down” from general theory, to specific theory or examples relevant to your model system, to a presentation of conflicting opinions or gaps in our understanding, to the purpose of your experiment (which should address the conflict or fill the gap). In some cases, your introduction may need to be long to place your study in perspective. In this situation, you would want to present your objectives early so the reader can relate your background information to your question. Finally, unlike papers in the Humanities, do not begin your introduction with a quote or hyperbole. Indeed, avoid using direct quotations! In science, paraphrase the point of the author and cite them parenthetically at the end of the sentence.
D. Methods. This should be the easiest section to write; you simply state what you did. It is written in the past tense; don't write a series of instructions, write a description of the experiment you conducted. In the course of the methods section, you should specify your experimental design, describing the levels of your independent variables and the variables you chose to measure (dependent variables). You should also justify why you performed these methods. Why did you choose these dependent variables to address your question? Why did you choose these levels of your independent variables for manipulation? Be sure to include the equipment that you used (manufacturer and model number, if unusual), and define the environment where the test was performed (temp, light, etc.). In a field study, a brief description of the field sites usually appears first in the methods. Finally, you should specify the statistical tests (and software packages) you used to analyze the data, and any transformations performed.
E. Results. You may think the results should be easy, too; this is
simply the information that your experiment produced. However, interesting results sections are very
difficult to write. You usually have several specific statements that you
want to make (the new 'facts' that you have found), but you also have
statistical analyses to present and figures and tables describing your results.
First, analyze your data. Which independent variables had significant effects
on your dependent variables? Make your figures and tables describe these
patterns (see part VI). Digest these patterns; interpret your results before
you start to write. After completing your analyses, you will have to decide on
an ORDER in which to present them. This
is not necessarily the order in which the tests were done. You are trying to make a logical argument;
presenting your results in the most logical order will greatly assist the
efficacy of the argument. When you know what your results mean, you can try to
explain them to the reader. Read it back
to yourself, out loud. If you stumble over grammar or can't understand it
yourself, the reader won't have a chance! When you are ready to begin,
start out with a declarative statement:
"Hours in sunlight had a significant effect on plant growth."
Now, call the reader's attention to the table or figure that shows this effect
to be statistically significant:
"Hours in sunlight... (Table 1)."
Table 1 might be the Analysis of Variance you performed, which documents that
the hours of sunlight had a statistically significant impact on plant growth at
the p=0.0025 level. As such, you might include the test and the alpha level
parenthetically (optional):
"Hours in sunlight... (ANOVA p=0.0025; Table 1)."
Now, describe the pattern, and tell the reader where this information is
presented:
"On average, plants in the 'long exposure' treatment grew significantly
more than the plants in the 'medium exposure' or 'short exposure' treatment (Bonferroni t-test, p= 0.05; Figure 1)."
Figure 1 might be a bar chart comparing growth of plants in the three light
exposure treatments. The appropriate mean comparison test documents that the
means are significantly different at the p=0.05 level.
Next, present a conclusion statement for the paragraph.
"Evidently, plant growth was stimulated by increased exposure to
sunlight".
So, here is your first paragraph: "Hours in sunlight had a significant
effect on plant growth (ANOVA, p = 0.0025, Table 1). On average, plant in the
'long exposure' treatment grew significantly more than the plants in the
'medium exposure' or 'short exposure' treatment; (Bonferroni
t-test, p= 0.05; Figure 1). Evidently, plant growth was stimulated by increased
exposure to sunlight."
Important
Don'ts!!
1. Don't begin a statement with:
Statistical tests don't show
anything. They just crunch numbers in a particular way. It is up to the experimenter - YOU - to
interpret the result of a statistical
test. For instance, a Chi-square test may indicate that a particular set of
data would only occur by chance 5% of the time (p = 0.05). This information has
no intrinsic meaning; some experimenters may interpret this pattern as
significantly deviant from random chance while others, using a more stringent
criterion (p = 0.01) may not. Make a
declarative statement and refer to the statistical analysis that supports this interpretation.
2. Don't present a long list of
significant results without interpretation:
"Hours in sunlight significantly affected growth (Table 1). Soil moisture
significantly affected plant growth (Table 2). Soil nitrogen also had a significant
effect on plant growth (Table 3)."
You should develop each point completely before moving on to another point.
After you say that an independent variable has an effect on a dependent
variable, describe the effect; how did the levels of the independent variable
differ? What does this mean? Show the reader the significance of the result.
3.
Number your tables and figures after you write your results section!
The first table that is referred to in the results section is, by definition, 'Table 1'. Likewise, the first figure referred to, even if it is after table 1, is called 'Figure 1'. You may not present the results in the sequence in which they were analyzed. Recognize that the logical development of your results may demand a different sequence, and the table and figure numbering should complement this new order.
The results section is where you
'present your case'. The logical flow is critical; you must convince your
reader that your argument is sound. If the readers are confused by your
results, or do not follow your interpretation, they will not believe you. They
will not accept that your conclusions are correct and important, and they will
not recognize the relevance of your experiment.
F. Discussion. The discussion is where you explain your results and interpret them in light of other work in the field. Usually, the discussion takes the shape of an inverted funnel. Start by presenting the essential conclusions of your specific study. (This leads directly from your results section, and provides a natural transition.) Then, apply your conclusions to the body of background information you relayed in your introduction. You may broaden your focus as you proceed. Remember the background information you presented in your introduction? That was the information you felt was relevant to your experiment. Now, discuss how your new findings relate to this background information. Are the major hypotheses in the field support by your research, or contradicted? The discussion and the introduction should reflect one another to some degree, with the discussion bringing your paper “full-circle”, integrating your results with the literature you described in the introduction.
The discussion also may include suggestions for future research, or disclaimers and explanations of methodological errors made during the course of the experiment. These are not REQUIRED elements, however. Many of you wrote lab reports for classes, where these were required elements of the report. Here, they may be appropriate but may not be necessary. In any case, if they are included, they must be well integrated. They should not just sit there at the end of the discussion. Rather, they should be integrated into the body of the discussion, so that your discussion can end on a positive note - like the major conclusion of your study, or the new question that it raises. Don’t end on a negative about a shortcoming of your experiment.
G. Acknowledgments. Thank the people who helped you research the question, design or conduct the experiment, and review drafts. Also acknowledge any funding support, and the source (check a few acknowledgement sections for examples).
H. Literature Cited. This section contains bibliographical information on the references that were cited in the body of the paper. It is not a bibliography; only list the references that were actually cited in the body of the text. YOU SHOULD USE THE COUNCIL OF SCIENCE EDITORS “NAME/YEAR” CITATION FORMAT. If you are submitting the manuscript for publication in a journal that uses a different format, and you wish to prepare your paper in that format, include the citation format instructions from the journal in your submission to the committee.
A. Citing an article by a single author:
Research papers in the sciences use a simple format for alluding to work done
by previous investigators. When you present information that you found in a
published document, you cite the author and year of publication
parenthetically, immediately after the information. For instance, suppose I was
writing a paper on the effects of resource patchiness on community structure,
and I read an article by J. Weins titled
"Population responses to patchy environments" published in Annual
Reviews of Ecology and Systematics in 1976. In this article, Weins states that foraging patch scale is determined by the
perception of the organism searching for the resource, and is not an inherent
quality of the resource. In my introduction, I might write:
"However, the relevant spatial scales of predictability and ephemerality
are defined by the perceptual and dispersal capabilities of the foraging
organism (Weins 1976)."
(Note: the use of present tense implies a fact which must be supported by a
citation or your data.) Cite only the author of the immediate information. If
you are citing a chapter authored by Burgdorff in a
text edited by Crane, cite: (Burgdorff 1983).
B. Citing a direct quotation:
You must cite all information that was published elsewhere and is not original
to your paper. Preferably, you paraphrase the information and present the
citation at the end of the sentence (as above). Sometimes, however, the
phrasing of the original information is particularly eloquent. Or, sometimes
you want to stress the authority of the source. In these cases, you want to
quote the information exactly. You must enclose the quoted material in
quotation marks:
As Price (1984) stated, "it is noteworthy that so many of the hypotheses
involve resources as the basis for understanding community organization, and
that competition is not invoked as a major organizing influence"(p. 476).
Use quotations sparingly! Sequences of direct quotations are difficult to read
because the style keeps changing. It also suggests that you don't understand
the topic well enough to interpret the information in your own words. This is especially
true of conceptual material; when you quote something that is not particularly
eloquent or authoritative, it suggests that you could not understand it well
enough to paraphrase.
C. Citing a series of articles at once:
Often, there are several citations that relate to a particular statement;
simply list these in chronological order, separated by a semi-colon:
"These resources often support diverse insect communities (Elton 1966;
Heed 1968; Beaver 1977; Schoenly and Reed 1987), yet
they are packaged into discrete units that are typically perceived as patchy
and unpredictable (Lacy 1984)."
Also, notice that the citations only accompany the clause that applies to them.
Lacy (1984) did not suggest that these resources support unusually diverse
communities so he is not cited after that clause. Obviously, citations can
become cumbersome and can influence the structure of your sentence. If a long
list of references comes between two clauses (as above), you might consider
breaking the sentence in half. However, lots of short, single clause sentences
are monotonous to read because they don't flow. Reading and rereading your
drafts will help you recognize rough spots. Obviously, this can't be done
overnight. You must give yourself ample time to write and rewrite your paper. A
rough job will be noticed.
D. Citing several citations by the same author:
If you have several citations and some are by the same author, group the
citations by author, separating authors by semi-colons:
(Jaenike 1978a, 1978b, 1986; Lacy 1979).
Notice that Jaenike's complete list goes first, even though his citations are chronologically split by Lacy's article. Also, if you refer to two citations by the same author in the same year, refer to the first citation cited as 'a' and the second one 'b'. Do not use the 'a' and 'b' designations that other authors used; those were dependent upon their order of use.
E. Citing multi-authored works:
If the citation has only two authors, present both their surnames followed by
the publication date: (Schoenly and Reed 1987). If
there are more than two authors, cite the first author's name followed by the
words 'et al.' and the year. The book Insects on Plants by D. R. Strong,
J. H. Lawton and R. Southwood is cited in the text
as: (Strong et al. 1976).
F. Citing unauthored pamphlets, etc.:
Some government and corporate publications are unauthored.
Cite these as 'anonymous', followed by the date of publication: (Anon. 1952).
G. Citing sources for equipment:
If you are using an unusual piece of equipment or material from an exotic
source, you can cite the source directly so that others trying to replicate
your experiment can get the same material:
"I counted the drosophilids and dusted the flies
collected on each plot with a different color of micro-ionized fluorescent dust
(USR Optonix, Inc., Hackettstown, NJ)."
H. Citing unpublished material:
Suppose you want to cite a manuscript that has not been published. You would
cite (Author, unpublished).
I. Citing personal communications:
Suppose you want to cite an interpretation that someone else made regarding
your data. You would cite (Author, personal communication). This situation may
arise regarding a professor's lecture notes or a chat you had about your
research. Be sure to get permission before citing the information.
The complete citations of published work are presented in alphabetical order (by
surname of first authors) in the Literature Cited section. Unpublished
manuscripts (unless they are in press) and personal communications are not
listed. The Literature Cited section follows your acknowledgments, and always
begins on a new page. Consult this website for style: The Writer’s
Handbook, University of Wisconsin – Madison.
You should choose your tables and figures carefully; they will form the backbone of your results section and should present your results in a way that clearly describes the patterns in your data. Don't include figures and tables that are extraneous to your report. Every table and figure must be referred to somewhere in your paper. Also, only use tables and figures to summarize the patterns in large sets of data; do not include tables of raw data. If you are only comparing two responses, a descriptive sentence in the results section might be sufficient. Don't be redundant! Use either a table or a figure to summarize a particular pattern, do not use both. Tables and figures should be appended to the back of the report, after the Literature Cited section. Each table and figure should be presented on a different page. Table legends appear at the top of the table; figure legends are commonly presented on a separate page that precedes the figures. Also, do not use multiple colors on tables or figures; shades of black and white, with hatching or stippling, is best; some readers are colorblind and may not perceive the differences in colored bars. In addition, most journals will charge extra for color figures – so only use color when necessary, like in photographs of fluorescing tags.
A. Legends
Each table and figure must have a descriptive legend. The legend should be complete; the table/figure should be comprehensible without reference to the paper.
As such, you need to include:
Examples:
Table 2. ANOVA measuring the effects of patch, pollination treatment (pollen added vs. natural) and raceme density (high vs. low) on two measures of fruit set in seven patches of Maianthemum canadense in central New Jersey: 1) the percentage of flowering ramets that bore fruit; and 2) the mean number of fruits/infructescence. Percentage data were transformed with an arcsin square-root transformation before analysis (Sokal and Rohlf 1981).
Table 1. The effect of replicate and plot density on mycophagous flies captured on experimental plots. ANOVA results for species richness and number of individuals (both log transformed; **** = p < 0.0001). Density means are significantly different; Bonferroni t-test, p < 0.05).
Figure 1. Fruit set in pollen added (solid bars) and naturally pollinated control quadrats (open bars) in seven patches of Maianthemum canadense in central New Jersey. The results of one-way ANOVA's are also presented (* = p < 0.05; ** = p < 0.01; *** = p < 0.001).
B. Figure Presentation
The type of variable that you measure may dictate the appropriate type of figure, especially if you are presenting a frequency distribution. For discontinuous data (where only certain values were possible, such as counts, states, attributes), a bar chart is appropriate. In this chart, the bars do not touch. For continuous data (measurements), histograms are appropriate (where the bars touch). If you are presenting means, margins of error (standard deviation, standard error, or confidence intervals) should be presented around each mean.
After you have done a thorough literature review, analyzed your data and drawn your figures, you can start to write.
Start with your methods; this is the easiest section to write and it will get you into writing and thinking about the project. You will feel good that you have started the paper and have not left the whole thing for the last minute.
After you analyze your data, make your figures, and draw conclusions, then write your results section. Again, have everything straight in your mind before you start to write! It may take a couple tries to get your points in a sequence that flows.
Now that you know what you found, you can write the introduction or discussion. By delaying this step until after your results section is done, you know which results were important and you can stress the relevant background information accordingly.
Some people like to write the discussion immediately after they write the results; this keeps their results fresh in their mind as they develop links with the background information. However, it is probably a good idea to have an outline of the introduction at this stage. This forces you to integrate the background material and really digest it, so the arguments you make in the discussion are logical and relate to the points you raised in the introduction. Remember, in the introduction you presented the background information that justified your question; in the discussion, you need to show how your results fit into that body of work you introduced as relevent.
Well, the body of the text is finished! You might want to do the Literature Cited section next. Although there is no particular logic to this, when literature cited sections are left until the end they tend to be done in a sloppy fashion with typo's and omissions. It is very important that the Literature Cited section be correct; it must contain all the citations and each must be presented in the appropriate format. This is busy work, but it must be done correctly so the reader can find your sources. If it is left until the end, you may make lots of typo's as you rush to finish. Notice on the GRADING RUBRIC that correct citation format counts quite a bit!
Tackle the abstract and the title next. These take time (especially the abstract) and should not be rushed.
Finally, after all the work is done, acknowledge the help that others have provided.
One final comment about papers. Drafts are a necessary step! Poor writing detracts from the quality of the presentation, and reflects upon the quality of the experiment. If time is pressing, get your methods and an outline to your introduction written before your results are produced. After you write the results section, write an introduction. When you are all done, set it aside for at least one day. Re-read your paper, sentence by sentence, as if you were reading it for the first time. Read it out loud; then you will really know how it sounds. Be hard on yourself; any improvements that you make at this stage will directly enhance the quality of your paper and your grade.
Formatting: Use double spacing (or 1.5), 12 pt. new times roman, and one inch margins. Page numbering in upper right is preferred.
VI. Commonly Misused Words
The following list of commonly misused words was initially prepared by the Iowa Experiment Station Publications at Iowa State University, and was modified by the editors of the Journal of Mammalogy.