Trust in news media in the United States has dropped to a new low, at only 31% of those polled by Gallup attributing a fair or great deal of trust in the fairness, fullness, and accuracy of media reports.[1]Worse, another recent Gallup poll suggests that half of respondents in the United States believe that the news media is deliberately misleading them.[2]The gulf between inaccuracy and intentional deception is material.
Canadian trust in the news has also trended downward and is at its lowest measured value – 39% – in a similar poll by the Reuters Institute.[3]There are many potential causes for this loss of trust, which we will term the trust deficit.
I examine a small, seemingly innocuous, element of trust in media reports: whether journalists should use the first-person point of view “I” or “we” in their articles under certain circumstances. Readers may find this question somewhat anticlimactic, but I will demonstrate that the handling of “I” and “we” can be of material importance in both media articles and academic journals.
Proper use of first person can be a cue for readers to understand the boundary between objective facts and subjective interpretation. The disclosure of this sometimes-hidden border could help to repair trust in the news media.
The issue “I” and “we”
Media articles differ from academic journal papers in several ways. Academic papers are aimed at specialized audiences of subject matter experts, whereas news media reports have a broad, general audience.[4]Academic (journal) articles are typically written in more formal and more technical language than reports produced by news media journalists. Despite these important differences, academic papers and news media articles are similar enough that we can meaningfully extend the discussion on the use of first person “I and “we” from academic papers to some types of media articles. After all, the general expectation is that both types of report are non-fiction and should be presented objectively.
The problem with dropping trust in news media is directly related to perceived impartiality, though the belief by some that the news media is intentionally deceptive goes beyond skepticism of objectivity to distrust of intent. I will restrict this discussion to objectivity.
Some writers and editors have been uncomfortable with using the first-person point of view in the past because it can undermine the perception of media objectivity. The use of “I” and “we” can bring in the subjective, and while this is sometimes appropriate for context or for immersive reasons, their use has historically been seen by editors as jeopardizing impartiality.[5]It brings the author into the story. It creates narrative, and it is perhaps a perception of too much narrative and not enough facts that has helped create the trust deficit. I will show that using first-person pronouns does not always make the article subjective; it can signal something else.
Style is a reason, just not the one connected to trust
The use of first-person pronouns has been widely discussed for journal papers, and much of that discussion has centered around the stylistic value of “I” and “we.” The use of these pronouns allows the stylistic stronger, and preferred, active voice.[6] [7]For example, consider these two sentences, addressing a conclusion:
“It is concluded that the O-ring seals have good integrity because of evidence a, b, c, and d.”
or
“We conclude that the O-ring seals have good integrity because of evidence a, b, c, and d.”
Both sentences can be used, but the second sentence can be considered stylistically superior because it is more forceful, clear and concise.[8]It is also appropriate in a conclusion to name the agent of the action, or the actor.[9]This leads to an even more important reason to use first person pronouns, as I illustrate in the next section.
The most important reason to use first person
There is a far more important reason than style to use first-person pronouns in academic journals or new media articles – their use imparts critical information. The use of “I” and “we” signals to the reader when the author has made a choice or used judgement.[10]
I illustrate this with another example following BIG Media’s coverage on excess deaths during COVID such as in Laurie Weston’s article Analysis of excess deaths in 2020 reveals surprising deviations. Laurie’s evaluation relied on reclassification of causes, which depended on her judgement. She also had to make a choice regarding the study period to estimate excess deaths. Two versions of a passage describing Laurie’s choices of study period follow. The second passage was the one used in her article.
“A four-year reference period for Statistics Canada was used. However, since this comparison considers full-year comparisons, seasonal adjustments are not necessary, and there were insignificant differences in Alberta deaths due to decreased air pollution or traffic between 2020 and typical years.”
or
“Trusting that Statistics Canada has considered the relevant factors and has sound reasoning for choosing a four-year reference period, I used the same. However, since I am looking at full-year comparisons, seasonal adjustments are not necessary, and there were insignificant differences in Alberta deaths due to decreased air pollution or traffic between 2020 and typical years.”
Both these paragraphs could serve the article, but the second one is better than the first. While the first paragraph is shorter, it lacks force. More importantly, the second paragraph is more transparent in describing the choices Lauire made in her data handling. Using first-person pronouns to signal a choice, as in the example above, does not automatically make the article an opinion piece or subjective. All new knowledge is inductive, meaning that its veracity is subject to strength of evidence but not absolute certainty.[11]Data choices must be made – even in the driest, most objective, technical studies – what is important is that readers know that a choice has been made, and why. It is easier to trust journalists when they are clear and open about their choices in managing data, as Laurie has done.
Another example from Overcoming stereotype threat and self-fulfilling prophecy, discussing stereotype threat, explicitly acknowledges the author’s judgement.
“We interpreted this thinking as counterproductive and morbid and argue that these tenets of CRT would increase threat levels for individuals of all groups.”
In this case, the author – me, in this case – is not only identifying himself but unequivocally telling the reader that he has made an interpretation. The interpretation is based on data and logic described in the article, but it is nevertheless an interpretation. The boundary between data and judgement is an important one, and the reader should know when it has been crossed. All articles that exceed the strict and literal bounds of the data should in some way disclose when the author’s judgement is being used. This is as true for academic journals as it is for news journalism.
If only “I” and “we” were used more
A less visible problem occurs when the journalist uses judgement without informing the reader – particularly regarding what data to include or exclude in analysis. These choices – of what data to include and what data to exclude – are crucial in providing a fair analysis. When researchers or journalists only choose to include or represent data and ideas that agree with their thesis, they create what is known as confirmation bias.[12]Slanted, one-sided, and even intentionally deceptive articles can be written using confirmation bias as a tool. Such use may create powerful, seemingly unequivocal arguments. This is often done without the readers’ knowledge and could be another cause for the trust deficit in journalism.
Confirmation bias begs the problem of writers identifying their choices. I will use an element from a recent article by David Haskell of the Aristotle Foundation as an example. The article discusses diversity, equity, and inclusion (DEI) training. It is a strong article, well researched, clearly written, and forceful, but when it discusses DEI research carried out by Legault et al., it only covers one side of her team’s evidence.[13]An excerpt follows.
“For example, in a laboratory setting, a University of Toronto research team led by Lisa Legault (now at Clarkson University) determined that race-focused DEI campaigns that exert strong pressure on people to be non-prejudiced backfired, yielding heightened levels of bigotry.”[14]
Haskell’s quote accurately represents certain elements of Legault et al.’s work. What is not mentioned in Haskell’s article is that Legault’s team also found that some less aggressive and more positive methods of DEI training were effective in reducing stereotypes.[15] I discussed both elements of Legault’s work in The Robbers Cave experiment’s important lesson for DEI. I argue that by excluding the positive evidence for DEI training from his description of Legault’s work, Haskell introduced an element of bias in what was otherwise an outstanding article. But most readers would not know this had even occurred – they would have to go to the source material of Legault et al., to find out.
Interpretation
It is permissible for writers to identify themselves in the first person, by pronouns such as “I” or “we” in news media or academic papers. This should be done sparingly, but sometimes this action goes from being merely permissible to highly appropriate. While most non-fiction articles aim to be objective, and the author generally should not be a big part of the story, the correct use of these pronouns better informs the reader. Even the most technical subjects in academia commonly require judgement or interpretation from the author, and that judgment should be made explicit.
The exercise of choice and judgment, and the subsequent use of first-person pronouns, do not automatically create excess subjectivity or relegate an article to being an opinion piece. Good scientists make choices and can present their objective conclusions as their conclusions based on the data and reason within their work.
By extension, this argument applies to authors’ choices of what data to use or not use in their analysis, which is a discussion readers should be aware of in evaluating the objectivity of any article.
This article is not an attack on the importance of objectivity in news media. The writer should generally not play a starring role in the copy, and the use of first-person pronouns should be limited. However, writers almost always make choices regarding data or use some element of judgment in their work. Being open about the part we play in our news media articles may even help I, we, and them to trust each other.
References
[1] Brenan, M., October 14, 2024, Americans’ Trust in Media Remains at Trend Low, Gallup
[2] Bauder, D., February 15, 2023, Trust in media is so low that half of Americans now believe that news organizations deliberately mislead them, Fortune
[3] Newman, N., R. Fletcher, C. Robertson, A. Arguedas, and R. Nielsen, June 17, 2024, Reuters Institute Digital News Report 2024
[4] Muthondeki, S., September 1, 2024, Difference between a research paper and report writing, LinkedIn Pulse
[5] Conover, T., October 19, 2021, The First Person in Journalism Must Be Earned, The Postscript
[6] Editorial staff, November 8, 2016, Scientific language is becoming more informal, Nature 539, 140
[7] Wordvice, November 20, 2022, Can You Use First-Person Pronouns (I/we) in a research paper?
[8] Strunk, W., E.B. White, 2000, The Elements of Style, Fourth Edition, Pearson Education, ISBN 0-205-30902-X
[9] Schults, D., February 23, 2011, Are first-person pronouns acceptable in scientific writing? Eloquent Science
[10] Claerbout, J., 1991, A scrutiny of the introduction, TLE, 10(1), 39-41
[11] Hunt, L., 2013, The importance of making conclusions and frameworks in reasoning: CSEG Recorder, 38, 7, 56-60
[12] Mercier, H., & Sperber, D., 2011, Why do humans reason? Arguments for an argumentative theory: Behavioral and Brain Sciences, 34, 57-111
[13] Haskell, D., February 12, 2024, What DEI research concludes about diversity training: It is divisive, counter-productive, and unnecessary, Aristotle Foundation for Public Policy
[14] Haskell, D., February 12, 2024, What DEI research concludes about diversity training: It is divisive, counter-productive, and unnecessary, Aristotle Foundation for Public Policy
[15] Legault, L., J. Gustell, and M. Inzlicht, 2011, Ironic effects of antiprejucidice messages: how motivational interventions can reduce (but also increase prejudice), Psychological Science, 22, 12
