Make your writing clear

Say what you mean

For example, the sentence

Parental involvement caused lower levels of peer problems.

suggests direct causation. However, if this was an observational study, the two outcomes were probably observed to occur together, but direct causation was not demonstrated. If so, the sentence should be phrased as

Parental involvement was associated with lower levels of peer problems [1].

Clarify pronouns

Pronouns are words that are used to refer to other nouns. Examples include it, which, this, that and they. The noun that a pronoun refers to should be made clear to readers.

For example, in the sentence

X binds Y, which enables Z.

it is unclear what the word which refers to: X, Y, or the binding of X and Y.

To clarify, the sentence could be rephrased as

X binds Y, and this binding enables Z.

or more directly

Binding of X to Y enables Z.

Make your writing specific

Use specific descriptions

For example, in the sentence

Anger-related language on social media sharply improved over time.

Whether the improvement means an increase or a decrease is unclear. This sentence would be better rephrased as

Anger-related language on social media sharply decreased over time [2].

Provide quantification

Avoid vague descriptions when describing quantities. For example, change

Nearly half a dozen studies have been performed.

to the more specific

Five studies have been performed.

Also quantify vague descriptions when possible. For example,

There is a growing demand for cybersecurity professionals.

could be changed to the more specific

It is predicted there will be 3.5 million unfilled cybersecurity jobs globally by 2021 [3].

Use simple language

Remove unnecessary words

Rephrase wordy sentences to convey your meaning more directly. For example,

Our results showed that the danmaku design had a significant promotion effect on user participation.

could be rephrased as

The danmaku design significantly promoted user participation [4].

In another example,

Due to the fact that we observed that all the variance was contained in the first coordinate, it is tempting to speculate that the space may be approximated well using one dimension.

could be rephrased as

Because almost all the variance was contained in the first coordinate, the space may be approximated well using only one dimension [5].

Avoid redundancy

The same information should not be repeated in adjacent sentences. For example, in

The time it took to find Waldo increased as the difficulty of the picture search increased. Therefore, the time taken to find Waldo is directly related to the difficulty of the picture search [6].

the second sentence could be deleted, because both sentences describe the same direct relationship.

Use sentence structure to guide readers

Place the topic near the beginning of a sentence, and consider making the topic the subject of the verb.

Structuring sentences in this way helps clarify the main topic of the sentence to readers.

For example,

The routing approach decreased traffic volume.

suggests to readers that the sentence is about the routing approach.

In contrast,

Traffic volume was decreased by the routing approach [7].

suggests to readers that the sentence is about the traffic volume.

Refer to concepts that you have already introduced at the beginning of a sentence, and place new information at the end of a sentence.

Structuring sentences this way helps guide readers through the information presented across sentences. Referring to something that has already been mentioned at the beginning of a sentence helps readers understand how that sentence relates to the sentence before it. Placing the new information at the end of a sentence (which is sometimes called the “stress position”) emphasizes that information to readers.

For example, in the following sentence

On Earth, there are abundant rare-earth elements, such as neodymium, scandium, and dysprosium. Only very small concentrations of rare-earth elements have been found and they are considered rare.

the first sentence does not clearly specify that the topic is rare-earth elements, because rare-earth elements is not introduced until the middle of the sentence. The second sentence starts by stating only small concentrations of the elements have been found, which seems to contradict the first sentence. Therefore, readers may find the information difficult to follow.

The same information could be rephrased more clearly as

Rare-earth elements, such as neodymium, scandium, and dysprosium, are abundant on Earth. They are considered rare, however, because they appear in very small concentrations [8].

Here, the first sentence establishes that the topic is rare-earth elements. The second sentence connects to the first sentence using they, uses however to signal a contrast to readers, and then provides new information describing why the elements are considered rare. The story is easier for readers to follow.

Phrase comparisons clearly

Comparative words

If you use comparative words (such as words ending in -er, more and less), be sure to specify what is being compared.

For example, the sentence

Convolutional neural networks have higher energy efficiency.

is unclear because the word higher suggests a comparison, but the sentence does not specify the comparison. The sentence should be rewritten to state both of the things being compared

Convolutional neural networks have higher energy efficiency than GPUs.

or the comparative word (higher) could be rephrased as

Convolutional neural networks have a high energy efficiency [9].

Increases and decreases

Increase and decrease should be used to compare changes in values within the same group over time but not to compare differences in the values across different groups.

For example,

The baby’s height was decreased compared with the man’s height.

is phrased awkwardly and may incorrectly suggest that the baby and the man started out with the same height, and then the baby’s height decreased. Instead, what is actually meant is:

The baby’s height was less than the man’s height.

or, if phrased more directly,

The baby was shorter than the man.

This same principle applies to describing different groups in a scientific experiment. For example,

Trust in the virtual robot was increased in the experiment group compared to the control group.

should be rephrased as

Trust in the virtual robot was higher in the experiment group than the control group [10].

But increased can be used if a group is being compared to itself, e.g.,

Trust in the virtual robot increased over time in the control group.

Times, fold or percentage changes

Descriptions of changes in terms of times, fold or percentage can often be unclear and should be phrased carefully.

Stating that a value changed to a certain value is different from saying that the value changed by a certain value. The word to should be used if the final value is being described, and the word by should be used if the degree of change is being described.

For example, if X increased from 100 to 300, the following descriptions could be used:

X increased to 300.

or

X increased by 200.

or

X increased to three times its initial amount.

or

X increased threefold.

or

X increased by 300%.

References

1. Ghosh AK, Badillo-Urquiola K, Wisniewski P. Examining the effects of parenting styles on offline and online adolescent peer problems. In Proceedings of the 2018 ACM Conference on Supporting Groupwork 2018 Jan 7 (pp. 150-153).
2. Baucum M, Cui J, John RS. Temporal and geospatial gradients of fear and anger in social media responses to terrorism. ACM Transactions on Social Computing. 2020 Jan 9;2(4):1-6.
3. Tang C. Community College Corner Cyber2yr2020: ACM guidelines for associate-degree cybersecurity programs. ACM Inroads. 2020 May 12;11(2):8-11.
4. Wu Q, Sang Y, Huang Y. Danmaku: A new paradigm of social interaction via online videos. ACM Transactions on Social Computing. 2019 Jun 20;2(2):1-24.
5. Piovarči M, Levin DI, Rebello J, Chen D, Ďurikovič R, Pfister H, Matusik W, Didyk P. An interaction-aware, perceptual model for non-linear elastic objects. ACM Transactions on Graphics (TOG). 2016 Jul 11;35(4):1-3.
6. Andrews D, Hill E, Lewis H, Martin M. Analysis of words per minute correlated to fixation length while searching. Proceedings of CPSC4120. ACM, New York, NY, USA. 2017.
7. Johnson I, Henderson J, Perry C, Schöning J, Hecht B. Beautiful… but at what cost? An examination of externalities in geographic vehicle routing. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies. 2017 Jun 30;1(2):1-21.
8. Kirkpatrick K. Electronics need rare earths. Communications of the ACM. 2019;62(3):17–18. doi: 10.1145/3303847.
9. Li Y, Liu Z, Xu K, Yu H, Ren F. A GPU-outperforming FPGA accelerator architecture for binary convolutional neural networks. ACM Journal on Emerging Technologies in Computing Systems (JETC). 2018 Jul 25;14(2):1-6.
10. Lohani M, Stokes CK, Oden KB, Frazier SJ, Landers KJ, Craven PL, Lawton DV, McCoy M, Macannuco DJ. The impact of non-technical skills on trust and stress. In Proceedings of the Companion of the 2017 ACM/IEEE International Conference on Human-Robot Interaction 2017 Mar 6 (pp. 191-192).