Spelling Errors to Watch For

A trail through the woods
A “trail,” not a “trial.”

A spell checker is an excellent tool, but it is not foolproof. For instance, if a misspelled word is also an English word, the spell checker will not catch it. I often see the following errors in my edits:

  • “form” instead of “from”
  • “expect” instead of “except”
  • “leaning” instead of “learning”
  • “trail” instead of “trial”
  • “filed” instead of “field”

If any of these words pop up in your paper, it is worth having a quick check for their common misspellings.

Three Tips for a Shorter Paper Title

four cute and fluffy kittensConsider the title “A novel method for kitten clustering based on fluffiness.” To ensure it is as short and concise as possible without losing information, I suggest looking at the following three issues.

  • Avoid starting the title with an article.
    • Starting the title with an article can affect alphabetic indexing. For instance, if many titles begin with “A,” there will be many titles listed at the beginning of the index. To avoid this, some indexing systems list titles with the article at the end: “Novel kitten clustering based on fluffiness, A.” However, this is a complicated solution. It is best to avoid articles altogether, and indeed, some journals specifically forbid it.
    • New title: “Novel method for kitten clustering based on fluffiness.
  • Do not use the term “novel” or “new.”
    • It is often said that “publication implies novelty.” This means that all work  that has not been rejected at the peer-review stage is novel (or new), as that is a requirement for publication. Hence, it is unnecessary to specify that the work is novel.  Of course, novelty is not necessarily true for articles published in “sound science” journals, but these are an exception.
    • New title: “Method for kitten clustering based on fluffiness.
  • Do not use terms such as “method,” “algorithm,” “technique,” “framework,” or “process” unless necessary.
    • At least in computer science, almost ALL papers are about some sort of method, algorithm, technique, framework, or process. Hence, it isn’t necessary to specify this either. In a few cases, these types of terms cannot be avoided. For instance, the term “framework” might be needed, as a framework is broader in scope than a method, and this fact might be important.  However, if framework is just a synonym for method, it should be eliminated.
    • New title: “Kitten clustering based on fluffiness” or even “Fluffiness-based kitten clustering.

Using these tips, we went from the nine-word title “A novel method for kitten clustering based on fluffiness” to the three-word title “Fluffiness-based kitten clustering” without losing any critical information about the content of the paper.

Word Count Reduction: An Example

256px-Elektro_and_Sparko_(263022490)Today, I have another example that shows how you can make your writing more concise and understandable.

Original: (40 words)

In our robot mobility system, we provide a simulation comparison which is shown in Figure 12 that uses random heights for the robots for the purpose of comparing the height as one of the important factors in our system.

Revised: (15 words)

In this example, we use random heights for the robots, as shown in Figure 12.

The sentence no longer states that height is an important factor in the system, but is it necessary to say this? After all, would a careful author devote an entire figure to a factor that wasn’t important? The exception here would be if the wider research community thought height wasn’t important and the main purpose of the figure was to disprove this.

Lists, Damned Lists, and How to Punctuate Them

“There are three kinds of lies: lies, damned lies, and statistics.” – Mark Twain (among others)14784640622_610f5b567e_o

Lists, especially enumerated lists, are a great way to ensure your ideas are properly emphasized. However, it can be hard to remember how to format these lists correctly. A part of the problem is that there are many forms these lists can take.

Hence, in this post, I run through these various list formats (according to the Chicago Manual of Style) in order from least to most emphasis on the list items.

Simple list of items in a sentence:

There are lies, damned lies, and statistics.

There are three kinds of lies: lies, damned lies, and statistics.

There is a comma after each item, and an “and” before the final item. The first list is introduced by a sentence fragment followed by a comma, and the second is introduced by a complete sentence followed by a colon.

Enumerated list of items in a sentence (a run-in list):

There are 1) lies, 2) damned lies, and 3) statistics.

There are three kinds of lies: 1) lies, 2) damned lies, and 3) statistics.

Note that despite the addition of the numbers, the punctuation remains the same.

Enumerated list of items set into the text:

There are:

  1. lies;
  2. damned lies;
  3. statistics.

There are three kinds of lies:

  1. Lies
  2. Damned lies
  3. Statistics

In the first list, where the list is not introduced by a complete sentence, the items are punctuated by semicolons or a period (for the final item). This style is not recommended by the Chicago Manual of Style. The preferred style is to introduce the list with a complete sentence followed by a semi-colon, as in the second example. In this style, the items are capitalized and there is no punctuation, not even for the final item.

Enumerated list of complete sentence items set into the text:

There are three kinds of lies:

  1. There are lies, which are bad.
  2. There are damned lies, which are worse.
  3. There are statistics, which are completely without any redeeming qualities whatsoever.

In this style, the list is introduced by a complete sentence followed by a colon, and each list items is capitalized and punctuated as a complete sentence.

To sum up, you have several choices when writing a list of items. All are correct, and your choice of format will depend on 1) the complexity of each item and 2) the amount of emphasis you wish to place on the items in the list.

There are lies, damned lies, and statistics.

There are 1) lies, 2) damned lies, and 3) statistics.

There are:

  1. lies;
  2. damned lies;
  3. statistics.

There are three kinds of lies: lies, damned lies, and statistics.

There are three kinds of lies: 1) lies, 2) damned lies, and 3) statistics.

There are three kinds of lies:

  1. Lies
  2. Damned lies
  3. Statistics

There are three kinds of lies:

  1. There are lies, which are bad.
  2. There are damned lies, which are worse.
  3. There are statistics, which are completely without any redeeming qualities whatsoever.

Concise Wording in Computer Science Research III

Concise language in computer science is a Good Thing. (See my earlier posts here and here for reasons why.) There are many ways to achieve this. One way is to avoid referring to things that you note that you have “mentioned.” It is better just to refer to the things themselves.

For instance:

“We use a torus interconnect topology to address the issues mentioned above.”

can usually be edited to read

“We use a torus interconnect topology to address these issues.”

Similarly:

“We use a torus interconnect topology to address the issues mentioned below.”

can usually be edited to read

“We use a torus interconnect topology to address the following issues.”

The word count saved is not spectacular, but the sentence is more direct, and hence the edits are worthwhile.

Concise Wording in Computer Science Research II

14781842571_2702a8a489_z

Even if you are an experienced writer, writing an academic paper has unique challenges. The ideas are often complex and highly abstract. In addition, competition for spaces in the top venues is tough, and peer reviewers are distracted and pressed for time. It is not enough to have completed a significant study; if the paper is not readable, few will make the effort to understand the ideas it contains, and it will be rejected.

Concise language has two advantages. First, it conveys those complex ideas as clearly as possible. Second, it reduces your word count. Many journal and conference papers have strict page limits; if you can cut out unnecessary words, there will be more room for your research.

For instance:

It is important to note that each pig has access to water.

Note that each pig has access to water.

Each pig has access to water.

It may seem harsh to get rid of something like “it is important to note,” but every sentence you write in a paper already should be important to note. If it isn’t, it should be deleted.

It is clear from Figure 2 that the second method has a higher performance than the first method.

Clearly, Figure 2 shows that the second method has a higher performance than the first method.

Figure 2 shows that the second method has a higher performance than the first method.

If it really is clear, shouldn’t be necessary to point out that it is clear.

Figure 1 shows the image before automatic segmentation, and Figure 2 shows the image after automatic segmentation.

Figures 1 and 2 show the image before and after automatic segmentation, respectively.

The use of “respectively” has been derided in some circles, but it is still accepted in scientific writing and there are cases where it significantly shortens the text.

More simple examples of how reduce words follow (you can find more examples here):

The largest extreme value point of the distance was 5 m.

The longest distance was 5 m.

 ***

The method was improved through the use of a Hough transform.

The method was improved with a Hough transform.

  ***

All slopes of the regression were positive.

All regression slopes were positive.

  ***

The purpose of Barnum’s approach is to maximize the cohesion in the model.

Barnum‘s approach maximizes the cohesion in the model.

Barnum maximizes the cohesion in the model.

Barnum maximizes the model cohesion.

 ***

Figure 2 shows that the second method has a higher performance than the first method.

Figure 2 shows that the second method performs better than the first.

 ***

The red legos, the green legos, and the blue legos were on the floor, but the other legos were on the table.

The red legos, green legos, and blue legos were on the floor, but the other legos were on the table.

The red, green, and blue legos were on the floor, but the other legos were on the table.

The red, green, and blue legos were on the floor, but the others were on the table.

 ***

Concise Wording in Computer Science Research I

14598557488_5a2646eb91_bEven if you are an experienced writer, writing an academic paper has unique challenges. The ideas are often complex and highly abstract. In addition, competition for spaces in the top venues is tough, and peer reviewers are distracted and pressed for time. It is not enough to have completed a significant study; if the paper is not readable, few will make the effort to understand the ideas it contains and it will be rejected.

Concise language has two advantages. First, it conveys those complex ideas as clearly as possible. Second, it reduces your word count. Many journal and conference papers have strict page limits; if you can cut out unnecessary words, there will be more room for your research.

A few simple examples of how reduce words follow:

“And futhermore, …” should be “Futhermore, …”

“And furthermore” is a familiar phrase in spoken English, and as a result, it is hard to spot this error because it sounds intuitively correct. Nevertheless, in academic writing, it is redundant.

“And” signals that there is an additional idea or argument coming up in the next sentence. “Furthermore” signals exactly the same thing. Only one of these words is necessary. Beginning a sentence with “and” is informal and should be avoided in academic writing.  Hence, the word to use is “furthermore” on its own.  If you wish to save four additional characters, you could shorten it even more to “further.” Alternatively, you could use “in addition” or “additionally.”

“The GA algorithm is well-known to be…” should be “The GA is well-known to be…”

Here, I assume that GA means “genetic algorithm” or perhaps “greedy algorithm.” If you expand the acronym, you then get the phrase, “The genetic algorithm algorithm is well-known to be…”

Clearly, one of those “algorithms” can go.*

“The initialization process of the module was performed.” should be “The module was initialized.”

This is is an extremely understandable error. After all, it is highly likely that in your computer code, there really is an initialization process. It is a bona-fide thing, perhaps a method or a procedure, that had to be written, debugged, etc. Furthermore, it has to actively “be performed” via a method or function call.

However, in academic writing, this level of implementation detail is not necessary. Most readers understand that these common processes occur when a module is initialized and are more interested in the novel aspects of your code. Hence, it is better to focus the text on the action that was performed.

In the same way, every time the words “process” or “was performed” occur in your text, it is likely that a similar edit is possible:

“The filtering process on the image was performed.” should be “The image was filtered.”

“The evaluation process of the method was performed.” should be “The method was evaluated.”

By making the effort to clarify your text as much as possible, you present your ideas in the best possible light and increase your chances of publication success.

 

*And then there are those who define recursive acronyms on purpose. These acronyms are definitely outside the scope of this post.