Scientific writing 3.0: A reader and writer's guide - Jean-Luc Lebrun, Justin Lebrun 2021

Examples and counterexamples
Conclusions: the smile of your paper
Paper structure and purpose

Examples

In the following example, the author repeats a main aspect of his contribution already announced in the discussion section. It is an encouragement for others to use his method.

Our method has been used to determine the best terminal group for one specific metal-molecule junction. In addition, we have demonstrated that, in principle, it is applicable to other metal-molecule couplings.

It is not always necessary to have conclusive results to conclude. Sometimes, the hypothesis presented in the introduction can only be partly validated. The choice of words to say so is yours, but you must admit that the phrasing is quite critical here. Which of these sentences is better?

In conclusion, our modified gradient vector flow failed to demonstrate that…

In conclusion, our modified gradient vector flow has not been able to demonstrate that…

Or

In conclusion, our modified gradient vector flow has not yet provided definitive evidence for or against …

The last sentence is much better, isn’t it? The word “yet” suggests that this situation may not last. Far for being despondent, the scientist is hopeful. “Yet” creates the expectation of the good news that comes later in the paragraph. To convince the reader, the author shares his conviction through the use of the present tense (in bold in the example).

In conclusion, our modified gradient vector flow model has not yet provided definitive evidence for or against the use of active contour models in 3D brain image segmentation. However, it confirms that polar coordinates, as suggested by Smith et al [4], are better than Cartesian coordinates to represent regions with gaps and thin concave boundaries. In addition, we have now removed the need for a priori information on the region being modeled without affecting model performance.

The findings are inconclusive, but they reveal that (1) an undesirable constraint has been removed; and (2) for a particularly complex type of contour, another coordinate representation scheme is confirmed to be more efficient. Even partial achievements are important to the scientific community when they validate or invalidate other people’s theories and observations, and when they establish the benefit of a method against other methods for a particular type of experiment. Science explores, step-by-step, a labyrinth of many dimensions. Marking a dead end before turning back is necessary, especially when much energy has been spent exploring that path.

If the findings are conclusive enough, why wait until all the possible paths have been explored before submitting a paper. Mention what you intend to do next to discourage potential competitors, or to encourage others to collaborate with you.

The 25% improvement in re-ranking the top 10 documents by using words adjacent to the query keywords found in the top five documents demonstrates the validity of our assumption. We anticipate that the high frequency but non-query keywords found in the top five documents may also improve the re-ranking and plan to include such keywords in future research.

In the previous example, the writer stated precisely his future research plan to establish the anteriority of his idea and protect future research.

Presenting in conclusions any limitation that has put a lower ceiling on your high hopes is a perilous exercise. But the ceiling is not permanent — at least, that is what you want to convey to the reader. You know that tackling these limitations is definitely worthy of future research. Relaxing one of your strong assumptions or finding a way to bypass a limitation may enable others to solve their problems. Taking the time to state assumptions and limitations is not only good scientific practice, it is also a way to promote science and your name in science. But how does one constructively present these opportunities in a conclusion? The next example adapted from an IEEE paper starts with a sentence that sends a chill down one’s spine.

Finally, we summarize the limitations of our optimizing algorithm and offer our future research plan.

There are so many limitations that the authors find it necessary to summarize them. Two of the authors of the paper were senior fellows, and I suspected they knew how to remain positive in the face of adversity. Indeed, they knew. Here is the first item on their list:

• Parameter tweaking. As discussed in section 4.2, the value of alpha is obtained without difficulty, but a satisfactory gamma value is obtained only after experimenting on the data set. We have given the reader pointers to speed up the determination of gamma in this paper. We plan to investigate a heuristic method that allows direct determination of all parameters. In this respect, we believe that Boltzmann simulated annealing will be an effective method.

• …

The parameter tweaking limitation is minimized in two ways: (1) by emphasizing that a method has been given to speed up the labor-intensive part of the algorithm, and (2) by showing confidence that a solution is at hand to bypass this limitation.

Counterexamples

When it comes to conclusions, be conservative and exercise restraint. Do not destroy your good work with sentences like these:

In the future, we would like to validate the clustering results not only from the promoter binding site analysis, but also incorporate more information such as the protein-protein interactions, pathway integration, etc, in order to have more convincing and accurate results.

As a reader, how did you view the achievements? Did you feel that the author was pleased with his contribution?

Here is a familiar sentence written in a humble, self-effacing way, too low key to encourage the reader.

Our method has been used to determine the best terminal group for one metal-molecule junction only, although in principle, it can be applied to other couplings.

Would you trust the conclusions of a paper that ends with the following sentence?

In the future, we intend to experiment our approach using larger data sets.

Does it mean that the current method relies on data sets which the authors think too small?

The next sentence seems fine… if only the writer had not used ’we believe.’

“Although these protocols will continue to change, we believe they represent a reliable starting point for those beginning biochip experimentation.”

The positive contribution is placed in the main clause at the end of the sentence, but some readers perceived this sentence as slightly negative. Read this sentence again and skip ’we believe.’ You may find the protocols more appealing. The facts appear to speak for themselves, without the need for beliefs to influence the decision of the reader.

In the next sentence, both main and subordinate clauses contain positive facts. Since the main clause contains information about the future, the future should appear appealing. But this is not quite the case:

Although the model is capable of handling important contagious diseases, new rules for more complex vectors of contagion are under construction.

Both subordinate and main clause establish positive facts, yet the overall perception is not always positive. Why? The readers are confused. Ordinarily, if the although clause contains a positive argument, readers expect the main clause to negate or neutralize the value of that argument. In this case, the main clause also contains a positive argument. As a result, the overall impression is mixed.