Transgender Perceptions in the Media, a Preliminary Study
Sarah J. Lewis (Volunteer Researcher) firstname.lastname@example.org
To date there has been very little research conducted into how the media reports on transgender issues and individuals. This preliminary study aims to provide a baseline for future research into this area by focusing on the reporting of a single news organization, the British Broadcasting Corporation (BBC).
This research contributes the following:
- A new annotated dataset of news articles containing the terms "transgender" and "transsexual".
- An analysis of this dataset inspecting frequency and type of mentions.
- An analysis of this dataset to determine the frequency of transphobic articles.
- Recommendation for future research.
We looked at 183 news reports posted on the website https://news.bbc.co.uk between 2006 and 2014. To gather the list of articles we used the BBC's search functionality to search for articles containing the terms "transgender" and "transsexual". We aggregated the information into a single comma-delimited file detailing:
- The year the article was published
- Whether the article referred to transgender, transsexual or both
- The url of the article
- The type of article - this column should be considered freeform
- Whether the mention of gender identity was relevant, not relevant or token i.e. part of the expansion of "LGBT"
- To this end, the classification of relevant/not relevant could perhaps be controversial, as a guide we devised the following set of questions, a yes answer to any would classify the article as relevant:
- Is the article reporting on a court case / work tribunal where gender identity is a core question ?
- Is the article reporting on an individual who has come out as transgender?
- Is the article reporting on an announcement of gay/transgender rights where one of the subjects is transgender?
- Whether the article contained transphobic language.
- This field is set to 'y' if the article misgenders a subject (either in text or through visual aids such as photographs) or uses dehumanizing language (e.g. "they are a transgender."). We have also marked an article as containing transphobic language if it mentions the subjects former name unnecessarily.
- These parameters were based loosely on the GLAAD Media References Guide for Transgender Issues .
- Notes - freeform author notes regarding the article, concerning future analysis or observations e.g. If the article mentions an individuals name prior to transition, or that the article is in reference to the Indian elections.
After the data was collected, we conducted the following analysis:
- How many articles were published by the BBC concerning transgender individual in each year between 2006 and 2013
- How many articles, per year, used the term "transsexual" versus "transgender"
- How many articles, per year, were relevant, not relevant and token references.
Number of Articles Published per Year
Perhaps unsurprisingly, the number of articles featuring the terms "transgender" and "transsexual" increased between the years 2006 and 2014.
However, the growth as seen by the graph above is anything but linear. 2014 saw a huge increase in the number of articles concerning transgender subjects compared to previous years. The exact cause of this growth is left as an exercise for future analysis.
Transsexual v.s. Transgender
Transgender as a term for describing persons with gender dysphoria has increased in popularity dramatically between 2013 and 2013. Transsexual as a term has stayed fairly linear. It is unclear whether the BBC have a policy on which terms to use. In some articles the two are used interchangeably.
Token References v.s. Relevant Feature Articles
We found that the majority (53%) of articles published by the BBC were relevant. That is they feature transgender subjects or issues.
A large minority (32%) contained only token references e.g. an expansion of the LGBT acronym with no further follow up of specific transgender issues. Interestingly in the final year of study, 2014, the number of token references v.s. relevant feature articles was exactly 50:50 with 36 articles each.
8% of articles, mostly published before 2010, contained irrelevant references to a subjects gender identity e.g. one of the articles featured news of the robbery trial - the defendants gender identity was unrelated to the trial proceedings but was still mentioned by the article.
We found that 59 of the articles (~29% of the total, 43% of the non-token articles) contained transphobic language, presented personal medical information or otherwise dehumanised their subjects. To be specific:
- 12 articles misgendered their subject.
- 31 articles mentioned their subjects former name (e.g. "born as..."), without relevance to the feature topic.
- 4 articles contained a picture of a transgender murder victim in their previous gender identity.
- 1 article referred to transgender people as "transvestites" while another used "ladyboys" - while the latter was in the context of interviewing a doctor in Thailand, the author made no attempts in the article to clarify terms.
Although not included in the definition of transphobic. We also found 18 articles containing the term "sex-change" - which carries many negative and invalid connotations.
Interestingly we found a lower prevalence of transphobia in 2013 and 2014 than in previous years. This is partially explained by the rise of usage of terms such as LGBT, which increase the number of articles in which terms such as "transgender" appear, although removing token articles from the results still yields a lower total percentage of articles featuring transphobic articles from the previous years (~38%).
The relationship between LGBT and Transgender - Just over a third of references to transgenderism published by the BBC were token references, mostly in the expansion of the LGBT acronym. In 2014, this increased to 50%, perhaps inflated by the number of reports on local pride parades (necessitating the use of the LGBT acronym).
It is impossible to say whether, as stated colloquially, the 'T' is silent in LGBT. The fact that transgender stories have increased over 300% in 2014 would seem to undermine this concept.
However, this study did not look at articles containing the terms "gay", "lesbian" or "bisexual". Perhaps the number or articles covering these, and similar issues, is substantially more than the number of transgender articles found.
Transphobic Articles - The study found that a very large number of articles, 29% of our total sample and 43% of the non-token sample, contained transphobic content. It is difficult to tell whether these articles are founded in a lack of journalist education on transgender issues or institutionalized transphobia. We would suggest that the BBC publish guidelines for reporting on transgender subjects, taking into consideration the GLAAD Media Guide.
Transgender v.s. Transsexual - It is clear that /'transgender" is becoming a more popular term, especially in the expansion of LGBT. Despite this the BBC has, confusingly, continued to publish articles containing the term transsexual at the same rate. We could find no consistency in how each term was used. We would suggest that the BBC clarify the terms so that reporters can use them consistently.
Limitations of the Study - The sample size of this study is very small, analyzing the articles of a single institution, without a baseline means that no firm conclusions can be drawn about the state of media in general.
Further not all the information published by the institution was analyzed e.g. The BBC has published many stories about transgender celebrities under the "entertainment" section of their website, which was not considered during this review. Further research should focus on multiple organizations and across multiple media formats (television, printed newspaper and online media, for example).
In addition to the above, the BBC do not publish a list of all of their transgender content, this means it is likely some articles were missed during collection.
Duplicate Articles - The BBC also, on occasion, publish near-duplicate articles under different URLs which show up in search. To prevent bias, we included all articles found, even if they appeared to be duplicates. This could have skewed some of the numbers.
Classifying Transphobic Content - The lack of a formal specification, especially in classifying transphobic content, and the reliance on manual effort to categorize the data could have lead to biases. Future research should attempt to outline the exact criteria for an article to be categorized as containing transphobia content, perhaps even utilizing machine leaning techniques to automatically classify data in an attempt to minimize bias.
Overall, the study shows that the number of articles covering transgender issues or simply using the term transgender has been consistently increasing over the past 8 years. We expect the number to continue increasing in future years.
Despite the limitations of this study we have found striking differences in how the BBC has reported on transgender subjects and issues between the years 2006 and 2014, both in terms of content and language.
We have found evidence of transphobia in a large proportion of BBC articles. We strongly suspect such transphobia is common across all forms of media, although further research is required. However, we have also found that the levels of transphobia have dropped in recent years, although there is not enough evidence to suggest a downward trend.
Finally, we have presented a dataset, free for the community to access, to aid future researchers in this and other areas.
This article was published on