A Look into Cambridge Analytica — First Week of the Crisis

George Kary
Apr 12, 2018 · 5 min read

We partnered up with Christos Iereidis of Ethnos Newspaper, to explore the international media landscape on the topic of Cambridge Analytica.

Our research team here at FightHoax selected and analysed 32 articles on the topic of Cambridge Analytica, published between March 18 and March 24 of 2018, during the event’s first week of media coverage.

As of the study’s publication date (Friday, March 30, 2018), these articles have generated a total number of 45 million impressions worldwide.

Ethnos Newspaper — Article by Christos Iereidis together with FightHoax

FightHoax’s technology focuses on creating a complete and objective informational view of the analysed article or articles, in a manner much like what you would do as a person to verify the content you are reading.

You can download the original case study — in Greek, by clicking here.

We break down 6 important aspects of our analysis.

Based on the following chart, a crucial part for the event’s coverage by the international media was Zuckerberg’s — Facebook’s CEO — announcement and press release on the social network on the 21st of March, which resulted in a near triple increase in digital impressions, as the media had reacted on a global scale by the following day, March 22nd.

Exponential Growth observed after Zuckerberg’s Facebook post — FightHoax

An event’s social impact directly affects its level of media coverage, as public interest is relative to which stories either take immediate effect, reach the front page or are followed up consecutively.

As expected, Facebook was the biggest source of social impressions.

Fact checking relies heavily on the human element of content creation, which leads our algorithm to explore the digital footprint of both the source / publication of the article and the author.

6 out 10 authors had a digital footprint — FightHoax

A digital footprint consists of information relative to the publication and the author, which also adheres to the article and subject we’re currently reading. In general terms, an author who has written in the past about the same or relative subject is considered a positive result in our search.

“Authors with no digital footprint create gaps in the psychological spectrum of our analysis”

Valandis Zotos, our Head of Product, highlighted on Ethnos

We need to take into consideration that the topic of Cambridge Analytica involved a higher level of technical jargon, was closely attributed to data breach and that the event’s protagonists were of global recognition.

During its first week, media coverage was solid and straightforward in terms of both coverage and publication history.

By extension of the event’s severity, effect on a global scale and association with data breach, a topic that is rather sensitive with public opinion, it is with definitive normality that factual coverage was observed in high levels and steady streams from the beginning to the end of the first week.

8 out of 10 articles were solid with facts and information — FightHoax

We may be reaching a line of over-explanation here, but it’s important to emphasise that the application of the political axis on the source’s publication applies strictly on its perception of events and topics.

FightHoax’s algorithm attempts to analyse all the different aspects of a publication website’s track record and history.

Interestingly enough, we observed a higher polarisation on the subject of Cambridge Analytica by publications prone to a “left” perception bias.

More than half of the publications were observed with a “Centre-Left” perception bias — FightHoax

We could create some initial observations here about the effect of the polarisation of media coverage noted in our research and analysis, however, their implications could affect the observations themselves.

It is noteworthy to outline that both our sentiment analysis and the subject negotiated in the event of Cambridge Analytica — of data breach — show that public opinion had a direct impact on media coverage.

Although lesser content length and poorer language quality do not necessarily highlight that a piece of news content is false or harmful, we observed in our analysis of the 32 articles that the majority of content explored the subject in greater detail, with careful writing and use of objective expressions.

13 out of 19 articles covered the event in satisfactory conditions — FightHoax

The emotional chart depicted below verifies our initial observations about the topic’s overall perception bias and the impact of public opinion.

Negative emotions attributed to data breach — FightHoax

Personal data use and Cambridge Analytica’s event of exponential data breach through Facebook were received in general retrospect with quite the severe negative disposition towards the subject itself.

A strong emotional pulse could pose a danger, as it helps increase false and potentially damaging content.

Coming to An Understanding

The first week of media coverage of the event of Cambridge Analytica unfolded in a steady stream of solid content and news. Albeit its worthwhile polarisation, news circulated offered a clean and mostly unbiased information view on what happened, how it happened and its implications.

We should, however, bear in mind that, due to the subject’s sensitivity in comparison to the public’s interest, we might observe an increase in harmful and false content, as the event steadily drops from the forefront of media coverage and analysis in the political spectrum.

Explore the detailed source that the FightHoax technology analysed by clicking the following link: FightHoax Cambridge Analytica Case Study

This analysis was co-authored by George Kary and Valentinos Tzekas from the team at FightHoax. For any inquiries, please e-mail us at info@fighthoax.com.

FightHoax

A technology company specialising in contextual data for…