You have finished your science project at last! Perhaps you submitted it to a science fair. Perhaps you have been asked to describe your project in a written report. Making your work accessible to others is a primary task for any working scientist. Often, scientists submit their work to journals. Research organizations such as NASA have their own "in-house" scientific publications. Regardless of how the work is published, authors will have their manuscript reviewed by other scientists familiar with the kind of work being described. Almost always, authors will be asked to make revisions to their manuscript before it will be accepted for publication. This "peer review" is essential to the science process. Although peer review is not always fair, it is nonetheless the primary mechanism for controlling the quality of published work. When the process works well, everyone, including the authors, benefits. The review process for peer-reviewed journals can be long! In the example shown below, the manuscript was submitted in January, but not approved for publication until November – not an unusually long delay. The increasing popularity of online publication has as one of its goals shortening the time from submission to publication without compromising the integrity of the review process.
Although each project is different and (hopefully) unique, there is a widely accepted format for describing scientific work. A scientific manuscript can be based on a science fair project, but it will require some additional work to transform even a very good science project into a written report format.
Typically, formal scientific papers are not written in the first person ("I measured...", "we analyzed..."). This more informal writing style may be OK for reports posted online, for example. If you do use the first person, be consistent throughout your paper.
Avoid using "cute" fonts and do not use more than one font for text and, perhaps, a different font for section headings. If you use color to display data, make sure that the report format will display colors. This can be a problem for printed journals, which will almost cerainly charge authors extra for printing color graphics within the body of a paper. But, it should not be a problem for work published online. During the review process, manuscripts should be saved in a standard word processing format. When the manuscript is finalized, you should save it as a PDF or (for online publishing) HTML document.
How long should your paper be? That depends on how it is going to be published or otherwise made available. Journals and other publishing venues may have page limitations. Some results can be described in only a page or so – a "research note" – while others may require many pages. Peer-reviewed professional journals may charge authors by the page, but this shouldn't be a problem for student research papers in less formal circumstances. But, before you start writing, you should first determine what the length restrictions are. If your project involves collecting a lot of data, you may need to be selective about what you present. Thinking about what to present, and how, can help you focus on the most important aspects of your research. This is an important process even without length limitations and is always a better idea than simply presenting all of your data and leaving it to the reader to figure out what is important.
Every scientific paper should include the following components:
The title should be a succinct description of the purpose of the paper. If a paper has more than one author, the author listed first should be the primary contributor to the paper. Additional authors can be listed alphabetically or in order of the importance of their contribution. For group projects, there may be disagreements about who should be the lead author. It is a good idea to settle this question as soon as possible, preferably before the paper is actually written! The affiliation (school or organization) of each author should be listed. Sometimes, contact information is included for the "corresponding author" – the individual to whom questions about the paper should be directed. Usually, this is the lead author.
The abstract appears at the beginning of a paper. It is a brief one- or two-paragraph summary of
your work. It should contain a statement of the research problem or question, a summary of how
you addressed the problem or answered the question, and a brief statement of important results.
Often, a paper's abstract is published separately. It should tell a prospective reader
what to expect if he/she reads the entire paper. Usually, references are not included in abstracts.
The introduction should describe and explain the research question or problem in detail. References to past work are included here. You should explain the relevance of other authors' work, or your own previous work, to what you have done for this paper. You should explain how your work is different from previous work, or expands on previous work, or even disagrees with previous work. Often, when you start a project, you will accumulate many references. Avoid discussing work which is not related to what you have done, no matter how interesting that work may be. Do not include references that turn out to be not relevant to your work – this will distract a reader from the message you are trying to convey.
Especially for an experimental paper describing data you have collected, the Methods section should answer questions like: "What did you do and why? If I want to duplicate your results, how would I do it?" You should include a description of the equipment you used, how and where you set up an experiment site, how and how often you collected data, and how you analyzed it. Whatever mathematics you use in explaining or analyzing your data should be included here. If there are references relevant to how you designed your experiment or analyzed your data, include them here.
Photos of your experiment site and instrumentation can be very helpful for someone reading your paper. Photos should be sharp and cropped to focus just on what is relevant to the project. Do not use blurrly photos. Photos of people have no place in scientific manuscripts unless they are directly related to your project. For example, it is possible that a photo of a person or something like a hand next to equipment might provide a reader with a better sense of the size of your equipment, if that information is relevant.
Describe the results of your work here. Typically, this section will include tables and figures to tabulate and display results. Tables and figures should be numbered consecutively and given descriptive captions. Graph axes and supporting text within a graph should use a consistent font – using "cute" fonts is a distraction and should be avoided. If you use a spreadsheet to create your graphs, it is almost certainly the case that you will want to edit at least the default font sizes so text will be legible if you need to reduce the overall size of the graph.
The text
should explain thoroughly the contents and significance of each table and figure. Do not leave it to the reader
to figure out what a table or graph means! This example shows total column water vapor (PW) data from NOAA compared
to data from a handheld instrument for measuring water vapor from the ground. The two data sources are distinctly
displayed and labeled. The Methods section would contain
a detailed explanation of the data sources. The Results section would focus on
interpretation of the data.
In this case, since the topic of the paper (the one whose title and abstract page is shown above)
was the handheld instrument, it was appropriate to include a
photo of the instrument in use, to show its size and layout. Although color would have made these images
"prettier," it wasn't necessary, and it wasn't an economically reasonable choice for this journal publication.
Photos also have a place in the Results section if they convey useful information.
A photo can be worth a thousand words of
explanation about information contained in a graph. In the image shown below, plots of sky brightness as a function
of distance in digital camera pixels from the center of the sun (blocked with the circular disk) are supplemented by
photos of the solar aureole displayed for the clearest and haziest conditions. In this case, color is important both
for showing sky conditions and because
a lot of sometimes overlapping data are being displayed. (Both sky images were taken with the same camera using the
same manual settings.
Use this section to summarize and interpret your results. Include comparisons with other work, as appropriate. Did you have a formal hypothesis prior to starting your project? Was the hypothesis upheld or not? (Note that a lot of good science doesn't result from starting with a formal hypothesis!) If results were unexpected, can you suggest reasons? Perhaps you should have collected more data. Perhaps, in retrospect, you should have designed your experiment differently. Remember that getting unexpected results is not necessarily a "bad" thing. Never try to "cover up" unexpected results by altering your data or discarding data for no other reason than that they were unexpected.
Would additional work help you to understand your results better? Do your results suggest further ideas for future research? Use this section to ask and answer these questions.
It is your responsibility to acknowledge contributors to your work whose
participation was not significant
enough to include them as an author. It is highly unlikely that any research in today's highly connected world is carried out
successfully without contributions from others. For student research, individuals whose contributions should be acknowledged
might include a teacher who encouraged your work,
someone who provided equipment
or laboratory space,
pointed
you to helpful references, helped with data analysis, or provided suggestions about early versions of your paper. Sometimes, as in the example below,
other relevant information should be provided.
 |
 |
References should use a standard citation format, of which there are several. Here is one source. The example belows shows one standard format for referencing journal articles. Be especially careful to reference online work. It is not acceptable to use anyone's work without attribution just because it is freely available online! A persistent problem with online sources is that it may not be possible to ascertain when they were put online. This information is certainly relevant when such a source is referenced in a scientific paper. The date on which you retrieve an online source should be included in your citation. If the source itself is dated, through a "last modified" line, for example, that date should also be included in your citation.
There is some disagreement about using online encyclopedias such as Wikipedia as references in scientific papers, because the quality and reliability of online material varies so widely. Unlike peer-reviewed journal papers, sources such as Wikipedia may not be reviewed at all, or perhaps "crowd reviewed," with variable results. However, there is no doubt that these kinds of sources often provide very useful and easily accessible information about specific topics. For example, googling "wikipedia dew point" will bring up several references to Wikipedia articles about calculating dew point temperature. Before using the equations in a paper, you should cross-check sources, compare results based on equations found in other sources, and, if possible, refer to primary sources given in the Wikipedia article. (Good Wikipedia science articles will always include such references.)
The fact that many peer-reviewed journals do not make their articles, even very old ones, freely available online is a
significant problem for student research. There is no easy solution to this problem.
Colleges, universities, and research centers have contracts with content providers that gives their scientists
access to journals. Sometimes journal content finds its way to online sources such as Wikipedia, which is a good argument
for not dismissing the use of Wikipedia references in some circumstances. Sometimes, "pre-publication"
versions of journal articles can be obtained from authors. Sometimes published articles are posted online even though,
strictly speaking, they shouldn't be there – a circumstance which no doubt reflects the frustration felt by many scientists
who believe that the results of scientific work should be more freely available. The best solution for student researchers is to
find a research mentor who can help you access the information you need.
