Writing a review – day 4: referring to yourself and more useful phrases

Writing a review – day 4: referring to yourself and more useful phrases

Free Preview
Writing style – how to refer to yourself, the author, in your own work

There are different conventions in different fields of science for referring to yourself in your written work, so you should look through some other papers in the field and then decide which is going to be best for your paper. However, the different options are listed below.

Use personal pronouns – ‘I’ and ‘we’

The most direct and personal approach is to refer to yourself in the paper as ‘I’ if you are the sole author.

I used the [xxx] method for calculating…

My opinion is that…

I would like to….

Groups of authors can use the term ‘we’. Sometimes single authors could use the term ‘we’ to refer to themselves and the readers, but it could be confusing as to who the ‘we’ is referring to, so it is best avoided.

We believe that…

We followed [xxx] method….

Our opinion is that….

Reviews are written by one person, so ‘I’ would be the correct pronoun for writing a review and ‘we’ for when you are writing a full paper as a group.

Impersonal pronoun – ‘one’

Some authors use ‘one’ as an impersonal pronoun, however, it can appear as old fashioned.

One could consider this to be…

One may believe that…

Follow [xxxx] method one would assume…

Impersonal approach – the authors

Some authors of research papers refer to themselves as ‘the authors’ or ‘the author’ if singular.

The authors did [xxx]…

The authors believe….

This is fine in most cases in a research paper, but it should not be used when referring to other authors work otherwise it will become confusing – therefore it cannot be used when writing a review.

Passive/impersonal approach without referring to yourself/yourselves

It is not clear why [xxx] method was used…

This paper focuses on…

It has been observed that…

These results lead to the conclusion…

The revisions have much improved the paper…

As an author or reviewer you have options about your approach, but the convention in your field may have already been set. Whichever approach you take, it is important to make it clear who you are referring to when you use certain pronouns or phrases, so try to avoid ambiguity.

This Journal will address how to refer to the reader in a later section about writing a paper.

Now let’s look at another example review (below) and consider the approach this reviewer took to descrbing themselves and then we can conside some more useful phrases.

Reviewer #1:

The authors address a highly relevant topic by conducting a scoping review on estimation and use of uncertainty around predictions by deep learning algorithms. In the abstract and introduction, they identify mistrust as a barrier to implementing deep learning in the healthcare setting, which I completely agree with.

Major comments:

1) They set the goal to propose a conceptual framework to overcome this barrier, however this section in the Results rather reads like a summary of definitions and methods. I find the introduction of these highly relevant, but I think the concepts should be explained even more thoroughly (with examples), because I imagine that many in the target audience are not aware of these methods. With regard to this, I don’t find Figure 1 very intuitive.

2) The explanation of epistemic and aleatoric uncertainty is somewhat confusing. In the abstract, the authors write “Overall, while uncertainty estimation accurately quantified aleatoric uncertainty”, based on the Results section it seems like “Aleatoric uncertainty is difficult to address directly”. Also, the authors claim that epistemic uncertainty can be reduced “by adding more training data”, then they write that aleatoric uncertainty can be minimized by “broader data collection”. It is not necessarily wrong logic, could be true for both, it just reads a bit confusing in contrasting the two uncertainties and highlighting the same solution.

3) I completely agree with the authors stating that comparing results is difficult due to different performance metrics across studies, and therefore I would tone down this part in the Results section and suggest to move sentences like “tended to have smaller sample sizes” etc. to the Discussion.

4) The methods are introduced to tackle mistrust, improvement of predictive performance is really highlighted in the Introduction, however in the Results, this aspect gets quite a bit of attention. Is there a difference between strong predictive performance of a model and strong performance in estimating uncertainty? It would be relevant to explain it to readers. Is improved performance e.g. with the MC method due to an ensemble effect? Some readers new to deep learning might understand this better with an example, analogy (decision tree – random forest, if this fits the purpose).

5) In the title, the authors define the healthcare domain, therefore I suggest the exclusion of articles on oil prices, fruit images etc. Steinbrener (25), Akbari (39), Wang (41). They could be mentioned in the Discussion, but not as included articles in the scoping review.

Minor comments:

6) I suggest the use of the PRISMA-ScR Flow Chart in its original format and the inclusion of this in the main body of the article.

7) Reporting of frequencies and percentages could be slightly improved if written out more precisely, e.g. 11 out of 12 (92%) in the first sentence in Medical Imaging section that refers to another number (12) in that sentence as the 100% and not all 25 articles.

This review is published in PLoS Digital Health https://journals.plos.org/digitalhealth/article/peerReview?id=10.1371/journal.pdig.0000085

This reviewer refers to herself/himself in this article as I and to the authors of the article as the authors (and sometimes afterwards they). This is done consistently throughout the article which helps with the clarity of the review and makes it easy to understand what the reviewer is describing.

Overall this is a helpful review making detailed points clearly – occasionally there is some poor English which makes one or two sentences a little difficult to understand.

More useful phrases for reviews

The explanation of [xxxxx] is somewhat confusing: this phrase highlights an area in the paper that is not easy to understand. The reviewer later writes that the same item ‘it just reads a bit confusing – this could be more simply put as ‘it is confusing’.

The reviewer also uses the phrase ‘wrong logic’ in this paragraph. In English the word ‘wrong’ is not normally put before logic. Instead you could use:

‘This is not necessarily false logic’; or ‘this is not necessarily flawed logic’.

Not necessarily means it might be or it might not be.

It would be relevant to…. This phrase highlights that something important should be addressed.

Is improved performance…. Due to: This is asking the authors to address a specific question related to their results.

Therefore I suggest…. or I suggest the use of: these two phrases politely ask the authors to make some changes to the article, in one case including something and in another case excluding something.

They could be mentioned in….: this phrase is used to suggest that the item to be excluded could be added somewhere else.

Reporting of frequencies and percentages could be slightly improved if…: This phrase allows the reviewer to specify how they would prefer to see the results reported.

Further study for this week

If you have time for further study this week try to write a review for a piece of research you are familiar with. Use the advice above and the advice in the other Journal entries for this week.

Take the day’s short quiz below.

Lesson tags: English for scientists, referring to yourself, Useful phrases for reviews, Writing style
Back to: English for Scientists 2022