Mary Westervelt, Director, Technical Communication Program
Published research papers observe certain norms of format and language. Some of these are imposed by the publishers (type size and font, how to cite sources, etc.). Other characteristics such as overall organization reflect not only the requirements of the publisher, but the expectations of the community: A reader expects to find a general summary in the Abstract, background and motivation in the Introduction, what was done in Materials and Methods, results in Results, and analysis of results in Discussion. The reader might read the Abstract, glance at the figures, flip to the Discussion to learn more about the results, and then decide whether or not to read the entire paper.
All writers observe the overall organization norms outlined above. In addition, a skillful writer carefully chooses words and grammatical structure to guide the reader through what may be difficult material. For example, the skillful writer keeps in mind that the reader does not necessarily see the logical connections between sentences within a section. The writer takes pains to make sure that these connections are clear.
The following analysis describes the basic sections of a research paper or lab report. For each section, we’ll examine examples from published papers and see how the authors were able to use careful language to help the reader follow the discussion.
The Abstract is
Sample analyses of published abstracts
Careful wording helps the reader follow the flow of ideas, and indicates the strength of assertions made about them. These words have been underline in the samples.
Example One (Paper One)
C.K. Stein, T. W. Glover, J. L. Palmer, and B. L. Glisson, “ Direct correlation between FRA3B expression and cigarette smoking”, Genes, Chromosomes and Cancer 34:333-340, 2002. Used with permission of the first author.
Introductory background information: broad picture |
Cytogenetic deletions and/or loss of heterozygosity (LOH) of the short arm of chromosome 3, often with a break at 3p14, are well documented in lung tumors. |
Intro to the hypothesis under investigation |
The coincidence of a chromosomal fragile site, FRA3B, at a common chromosomal breakpoint in lung cancer has suggested that fragility at this site may predispose to breakage that could contribute to multistep carcinogenesis. This idea is supported by the more recent finding that FRA3B maps within the FHIT (fragile histadine triad) gene, and that aberrant transcripts and genomic deletions of FHIT/FRA3B occur in a variety of tumors including lung tumors. |
This study: methods |
To determine whether some individuals have increased fragility of FRA3B that might increase the risk for breakage or deletion in 3p14.2, fragile site expression was examined in smokers, nonsmokers, and small cell lung cancer (SCLC) patients. |
Results |
The data clearly show that active smokers exhibit a significantly higher frequency of fragile site expression, including FRA3B, compared to that of nonsmokers and patients diagnosed with SCLC who have stopped smoking. |
Discussion |
These results suggest that active tobacco exposure increases chromosome fragile site expression, and that this fragility is transient and reversible. The data support the hypothesis that exposure to tobacco carcinogens increases the potential for chromosome breakage at fragile sites. |
Abstract Example Two (Paper Two)
Lima, R. Wada, S., Takeda, M., Tsubota, K., Yamaguchi, T. (2007, January 16). In vitro confocal micro-PIV measurements of blood flow in a square microchannel: The effect of the haematocrit on instantaneous velocity profiles. Journal of Biomechanics, 40, 2752 – 2757
Broad picture |
(Missing in this shortened abstract. Note the extra-detailed title, however.) |
Hypothesis under investigation |
|
This study: methods |
A confocal microparticle image velocimetry (micro-PIV) system was used to obtain detailed information on the velocity profiles for the flow of pure water (PW) and in vitro blood (haematocrit up to 17%) in a 100-µm-square microchannel. |
Results |
The averaged ensemble velocity profiles were found to be markedly parabolic for all the working fluids studied. When comparing the instantaneous velocity profiles of the three fluids, our results indicated that the profile shape depended on the haematocrit. |
Discussion |
Our confocal micro-PIV measurements demonstrate that the root mean square (RMS) values increase with the haematocrit implying that it is important to consider the information provided by the instantaneous velocity fields, even at low Re. The present study also examines the potential effect of the RBCs on the accuracy of the instantaneous velocity measurements. |
Introduction
The Introduction contains
Example from Paper Two (in this and later examples, material in square brackets [ ] summarizes material cut from the original):
The velocity profiles of blood flow in vivo and in vitro have been measured using several techniques, including double-slit photometry (Gaehtgens et al., 1979/ Baker and Wayland, 1974), video microscopy and image analysis (Bugliarello and Hayden, 1963; Tangelder et al., 1986; Parthasrathi et al., 1999), laser-Doppler anemometry [more sources], and particle-measuring methods [more sources]. Nevertheless, no general consensus yet exists concerning the actual velocity profile in microvessels. While some studies have reported parabolic profiles [more references], others have suggested blunt profiles [more references]…. |
General topic of the report Verb: not past; passive. The authors are giving summary background information, but not discussing specific (past) studies.
Citations occur in the text.
The motivation for this study is introduced here.
|
Due to its outstanding spatial filtering technique and multiple point light illumination system, confocal microparticle image velocimetry (micro-PIV) has become accepted as a reliable method….Very recently, we demonstrated the ability of confocal micro-PIV to measure both homogeneous and nonhomogenous fluids (Lima et al., 2006a). |
New topic = new paragraph
Here, the authors discuss a specific result from a study performed by them. Switch to past time is introduced with ‘very recently’. Subject = we. |
The present study compared the instantaneous and ensemble velocities profiles of pure water and blood flow in vitro. The velocity profiles of both pure water and in vitro blood with two different haematocrits (9 and 17% Hct) were acquired in the centre plane of a 100-µm square microchannel. |
Or: “This study compares the instantaneous….” Also needed: some link between this paragraph and the preceding one. What was the motivation for the study? What did they expect to find? Adding that information would make past tense more appropriate. The last sentence is better moved to the Materials and Methods section. |
Possible rewrite of the last paragraph of the Introduction:
…Very recently, we demonstrated the ability of confocal micro-PIV to measure both homogeneous and nonhomogenous fluids (Lima et al., 2006a).
In order to measure the actual velocity profile in microvessels using confocal micro-PIV, we compared the instantaneous and ensemble velocity profiles of pure water and blood flow in vitro.
Materials and methods (Experimental methods)
In a published paper, this section outlines the steps taken by the authors/investigators. Past tense is the norm. Passive is common; however, use of active voice and we/our is appropriate when it clarifies the process (example from the Materials and Methods section of Paper Two):
In this study, we used a 100-mm-square borosilicate glass microchannel
fabricated by Vitrocom (Mountain Lakes, NJ, USA), which was mounted
on a glass slide immersed in glycerol that had the same refractive index. A
square microchannel was selected to minimise possible refraction of the
laser beam at the walls of the microchannel.
In the following example (from the same paper), the authors explain the motivation for their next step :
In order to obtain adequate quality images for processing with the PIV
software (PivView; PivTec GmbH, Göttingen, Germany), we captured
images with a resolution of 640X480 pixels and 12-bit greyscale, at a rate
of 200 frames/s, with an exposure time of 4995 ms, and a time interval (Δt)
of 5 and 10 ms between two images. All the PIV measurements were
performed for a period of approximately 0.5 s in order to obtain both
instantaneous and ensemble averaged velocities.
More about passive voice. Passive voice is common in technical writing. Some writers mistakenly think that they must never say ‘we’ or ‘I’ in a formal paper, and that the only other choice is to use passive voice. This is not true, and use of passive often creates cloudy or flabby sentences. Let’s try putting one of the active (‘we’ = agent) sentences from the previous passage into passive:
In this study, a 100-mm-square borosilicate glass microchannel fabricated by Vitrocom…, which was mounted on a glass slide immersed in glycerol that had the same refractive index, was used.
The resulting sentence is nearly impossible to follow! The ‘we’ sentence is much clearer. However, it is quite possible to reword the original sentence to leave out ‘we’, while still maintaining active voice:
This study utilized a 100-mm-square borosilicate glass microchannel fabricated by Vitrocom…, which was mounted on a glass slide immersed in glycerol that had the same refractive index.
Which is clearer: the original, or this revision? Which is more suitable for technical writing? Opinions may differ; but surely all will agree that either the first or the last version is preferable to the passive version.
A final note: All three versions of this sentence utilize passive in modifying phrases (fabricated by Vitrocom / which was mounted / a glass slide immersed in glycerol). Passive voice is very useful for allowing the smooth combining of phrases into one sentence.
Results/Discussion
Whether this is one section or two, the purpose is to clearly present the results of the experiment or study and to draw appropriate conclusions from the results. If a paper presents results for several experiments, probably the results and discussion of each experiment will be presented together. Such a paper will probably have a separate Conclusions section to provide a summary of the entire paper.
In any paper, the report of the results includes relevant graphics (charts, graphs, and tables) with salient points noted in the text. Any graphics presented relate to the stated purpose of the report.
The Discussion section, with the Conclusions section if that is separate, answers these questions:
The Discussion section is the most likely part of the paper to lose clarity if passive voice is overused. Inexperienced writers frequently produce awkward or incomprehensible sentences by trying to avoid saying ‘we’ in the Discussion. Let’s take a clear sentence (from Paper Two) which uses ‘we’, and try removing ‘we’ from it. First, the original:
Comparing the ensemble velocity profiles of all fluids (see Fig. 1), we observed small deviations (<5%) for both Δt = 5 and 10 ms, especially in the central region of the velocity profile.
Eliminating ‘we’ produces this:
Comparing the ensemble velocity profiles of all fluids (See Fig.1), small deviations…were observed, especially in the central region of the velocity profile.
The problem with the new sentence is that the reader expects the subject of the introductory phrase (implied ‘we’) to be the subject of the sentence. However, the subject of the sentence turns out to be ‘small deviations.’ The reader has to think, not just about the technical content of the passage, but about the wording and grammar. Good writers avoid letting the wording and grammar interfere with the meaning and flow of ideas.
A more careful writer might produce this sentence:
A comparison of the ensemble velocity profiles of all fluids (See Fig. 1) revealed small deviations…, especially in the central region of the velocity profile.
Which is clearer: the original, or this version? The choice here is more between two styles, and less between two degrees of clarity.
Below is an analysis of Paper Two, a paper which has a combined Results/Discussion section followed by a Conclusions section.
3. Results and discussion 3.1. Ensemble velocity profiles Most previous studies have determined the velocity profiles of flowing blood by measuring the time-averaged velocity field. Fig. 1(a) and (b) shows the averaged velocity of 100 (Δt = 5 ms) and 50 (Δt =10 ms) ensemble PIV images, respectively. These images were recorded at the same time period of approximately 0.5 s. Fig. 1 also compares the PIV measurements with an analytical solution for steady flow through a long, straight, rigid square microchannel (see Lima et al., 2006a for more details). |
Heading and Subheading guide the reader by dividing the results/discussion section into subsections.
Justification for use of the technique employed in this study (previous studies)
Discussion of results utilizes reference to graphics: present tense
Information about how data were recorded: past tense. (Also passive)
|
Comparing the ensemble velocity profiles of all fluids (see Fig. 1), we observed small deviations (<5%) for both Δt = 5 and 10 ms, especially in the central region of the velocity profile. Using the t-test analysis we found no significant difference between the working fluids and the analytical solution at 98% confidence interval. Hence, these results imply that the ensemble-averaged velocity profiles of in vitro blood with haematocrits up to 17% flowing within a 100-mm square microchannel do not change significantly from a parabolic shape. These results agree with Baker and Wayland (1974) and Sugii et al. (2005). [Further analysis of ensemble-averaged velocity profiles for both Δt = 5 and 10 ms] These results suggest that for both cases it is possible to obtain reliable ensemble-averaged velocity profiles for all the working fluids used in this study. |
Discussion of results
Authors use ‘we’ for clarity
Description of analytical process Degree of certainty:
|
In Paper Two, the summary elements of the paper are presented in a final section called Conclusions.
4. Conclusions In this study, we determined both ensemble and instantaneous velocity profiles for in vitro blood (haematocrit up to 17%) flowing through a 100-µm-square microchannel. Although the ensemble velocity profiles were markedly parabolic, some fluctuations in the instantaneous velocity profiles were found to be closely related to the increase in the haematocrit. The present study shows clearly that the RMS values increase with the haematocrit implying that the presence of RBCs within the plasma flow strongly influences the measurements of the instantaneous velocity fields. The possible reasons for the RMS increase are the motion and interaction of RBCs and the light scattered and absorbed from the RBCs. This latter cause seems to be more predominant at Hct = 17%. As a result, for 17% Hct improvements on the signal-to-noise ratio are required to further enhance the measurement performance of the instantaneous velocities. |
Heading guides the reader.
Summary, related to point of study
Interpretation of results
Strength of assertion: This present study shows clearly that…
Strength of interpretation: implying that…
More discussion, including limitations of this study and need for further study.
|
For comparison, here is the Discussion section of Paper One. This paper has a separate Results section and no separate Conclusion.
Summary of the motivation and hypothesis of the study |
The preponderance of lung tumors with LOH or cytogenetic deletions and rearrangements of the short arm of chromosome 3 has led to the speculation that expression of the most common fragile site, FRA3B, might be associated with some of these events. This is supported by [certain findings.] To consider this from a different perspective, our study examined patterns of fragile site expression and demonstrated that active cigarette smokers have increased expression of fragile sites including FRA3B. |
Significance of results Hypothesis validated |
….The present study is the first to compare the levels of fragile site expression in nonsmokers (no cigarette smoke exposure), smokers (current exposure), and lung cancer patients who have stopped smoking (past exposure). The data show a statistically significant increase in general fragile site expression as well as FRA3B expression in smokers compared to that of nonsmoking controls and lung cancer patients (Table 1). |
Relationship of results to those of earlier studies / hypotheses |
Our findings confirm earlier studies that compared smokers with nonsmokers (Kao-Shan et al., [1987]; Ban et al., [1995]) and are consistent with the studies comparing lung cancer patients with healthy individuals (Porfirio et al., [1989]; Egeli et al., [1997]) when smoking exposure is factored in. |
Analysis of results |
However, this analysis adds a new dimension. By concurrently examining fragile site expression in smokers, nonsmokers, and nonsmoking SCLC patients, we were able to demonstrate two significant points. It is active tobacco exposure that is important for increased chromosomal fragility and, second, the increase in chromosome fragility is transient and reversible. This is established by the finding that it is the active smokers who exhibit the highest expression of chromosomal fragile sites, but former smokers (i.e., patients diagnosed with SCLC who have stopped smoking) have fragile site expression that is no higher than that of a nonsmoker. These data agree with biochemical analyses showing that chemical compounds associated with smoking are present in the peripheral circulation of heavy smokers, and that there is a reduction in the level of these substances over time after the individual stops smoking (Mooney et al., [1995]). |
Limitations of results/guidelines for interpretation |
It is important to note that, in this study, significant levels of fragile site expression were seen only in the cells that had been cultured either in low folic acid or with the addition of aphidicolin. In vitro, common fragile sites are usually not expressed unless they are induced by these or other agents that disrupt DNA replication and/or synthesis. In vivo, breakage at fragile sites appears to be affected by environmental factors that produce DNA damage (Yunis et al., [1987]). |
Conclusions to be drawn |
The current data suggest that in vivo fragile site expression may be influenced by a combination of tobacco exposure and other factors associated with tumor formation, such as abnormalities of DNA replication or repair during tumorigenesis…. |
Summary: Writing guidelines for all sections of a paper or lab report
1. Choose a logical order in which to present your information. Clearly indicate the relationship of one idea to another by careful choice of logical connectors.
2. Carefully word your discussion to indicate the degree of certainty of your results or your interpretation of them.
3. Accurately and appropriately cite sources in your text. Each citation must be linked to an item in the List of References at the end of the article (or in the Appendix to your Lab Report).
4. Use we/our sparingly. However, this does not mean you need to resort to awkward or unclear use of the passive. Choose an active-voice subject other than I/we as in these examples (Examples throughout point 4 are from Paper One, Discussion) :
Passive allows focus on processes and is appropriate in these examples:
We/our may be the most efficient way to refer to the agent:
(Possible rewording: These findings confirm earlier studies….)
5. Make sure your sentence structure is clear. Avoid leading the reader astray with misplaced modifiers, or introductory modifying phrases that don’t modify what the reader expects them to, like this one:
This sentence momentarily misleads the reader, who expects that the understood subject (we) of the modifying phrase will be the same as the subject of the sentence. However, the subject of the sentence is they. The following improved sentence keeps the focus on the images and the process, while avoiding the confusion of the first sentence: