Fact-checking Univers: to boldly go where no blogger has gone before!

With my fellow students, I was asked to do a fact-checking process for the data journalism class; and this is what we did! So, here you will see our findings!

This is about a team project that consisted of five people. First of all, a news medium (Univers online) had to be selected from which various articles were chosen to make our analysis. Therefore, each of us chose an article and the fact-checking process began. Questionable facts were noted and relevant sources were investigated to verify the reliability of facts mentioned in the articles. Furthermore, it was considered very helpful if the journalists themselves could provide answers regarding any doubts about information in the article. Therefore, we tried to contact the writers of the articles, the fact-checking findings were presented, and afterwards responses were asked so as to conclude with the final results of the fact-checking procedure. The news medium utilized for this assignment is Universonline, which is an independent news source for students of Tilburg University. In general, it is considered as a trustworthy source, providing its audience with a couple of articles, with topics related -especially- to the university, but also news concerning science, society, culture as well as sports. It is also noteworthy that most of the articles are available not only in Dutch but also in English. Three articles utilized by Universonline for the current team assignment are in English and the other two are in Dutch. Thus, the articles selected for the fact-checking process are the following:

Fact-checking process
The fact-checking process can ensure how accurate and trustworthy a story is before it is published. Fact-checking can improve the credibility of an article and reduce fact errors. After choosing five different articles from the same medium, the first step was to read the whole article carefully. The next step was to highlight and mark all the facts that we wanted to verify and double check. In order to ensure the legitimacy, authenticity and validity of the information mentioned in the articles, we started using the internet to fact-check. The internet can be a very useful tool for facts that need additional verification, because many studies and sources are available online. Therefore, we searched for any additional web sources that could confirm the details we wanted to check. We also tried to select the best sites for verifying the questioned facts. University and government sites were also checked in some cases, as they are usually updated on a regular basis so these were also powerful tools for our fact findings.

The following step was to contact the journalists of the articles directly to verify the facts. All the journalists were contacted via email and some of them via telephone calls. Concerning the emails, firstly we presented ourselves and the purpose of our project. The main goal of the emails was to find out which sources the journalists used for their articles and provide us with extra information about the facts. Moreover, we tried to conduct the whole email and the questions in a polite and professional way so that the journalists would not interpret them as critique for their work. After sending the email twice, three of us had a response from the journalist, who provided us with the sources that she used for her three articles. However, two of us tried to contact the journalists not only via emails and social media messages but also by calling the editorial office of Universonline and ask for their telephone numbers or any other way to contact them. Unfortunately, when the particular journalist finally replied, she refused to answer some short questions by telling that she was too busy; thus for both journalists who didn’t reply there were no other options available.

Some extra special cases of verifying the facts were by consulting students, who followed one or more classes of a teacher who was mentioned in an article as someone who forbids smartphone and laptop usage during lectures. Moreover, former Erasmus students were conducted to read the article about personality traits obtained by Erasmus program and to say their opinion based on their experience. Finally, people who were working in the University of Amsterdam and the press agency of Tilburg University were also conducted via telephone call and emails in order to verify the boycott to Africa.

The final stage was to confirm if the fact has been verified based on our own research and on the journalist’s response. We separate the facts in true, false, inconsistent or unknown/missing in order to make a final list and understand in general which fact belongs to which category.

Main findings
All facts in each article were checked and can be summarized as overall main findings. In total, 35 facts were found in the aforementioned articles and checked on correctness. It was interesting that the second biggest amount of facts were labeled as not entirely correct, with a total of 8 cases. In most cases the information was copied from another source without (obviously) checking the facts. This caused the facts were half correct; e.g. when another side of a story was missing or important additional information was not mentioned. For example, in the article ‘Beer not bad for you’, the journalist wrote that 2230 youth children participated in a longitudinal study. However, according to the original source, it became clear that the research indeed started with 2230 participants, but in the end, there were 1596 participants left. The conclusion written in the dissertation was based on the remaining 1596 participants.

In addition, 4 facts could be remarked as false, probably caused by no fact checking process. In the article ‘No laptops in lecture rooms in Utrecht University’, for instance, it was stated that laptops were forbidden, whereas it became clear in other sources this was just an experiment, not a rule. A similar example was found in the article ‘Amsterdam school boycotts Africa’, where was stated that the Amsterdam University of Applied Sciences boycotts Africa by forbidding internships and study-related travels to Africa. However, both UvA and HvA confirmed this was not correct; the University simply double checks whether a trip might be risky due to, for instance, the Ebola virus or hard political situations.

20141218125008

Also, 4 facts were considered as inconsistent; other sources were found and reported some contradictory information. In the article ‘Student drinks through funnel and dies’ a clear example was shown. In this article it was stated that the student died 20 minutes later in the hospital, however, other sources mentioned that he died on his arrival to the hospital. It still remained unclear what the actual truth was.

Furthermore, in 2 cases information was simply missing, indicating that a part of the information is true, but since some information was absent, the fact could not be considered as trustworthy. In the article ‘Erasmus students less likely to be unemployed’, it was shown that “the study was initiated by the European Commission (EC) and compiled by independent experts”, whereas pivotal information was missing; the anonymity of independent experts brings great difficulty to the readers. At the end, the writer of this article added that the Erasmus Impact Study was conducted by an independent consortium of experts led by several other institutions.

For 4 other facts it was still unknown whether they were correct or not. These were all cases where a quote or statement of a certain specialist was used, where it could not be verified by those experts whether this was entirely what they had said or not. However, in each article also some facts – in total 13 – were considered as true.

Overall conclusion
Although Universonline states that their information must be entirely correct, we have also found some conflicting information. In total 35 facts were checked, but only 13 can be considered as correct, 14 as not entirely correct (due to inconsistent, contradictory or missing information), 4 as unknown truths, and 4 as false. As Universonline also mentioned in their guidelines, their articles are often short which causes that information is deliberately omitted. It would be probably impossible – due to, for instance, time pressure – to check all the facts. Moreover, a story needs of course some context and, therefore, cannot be fully objective. However, an important remark can be made, since the greater part of all facts do not seem fully accurate nor reliable. In general, it can be concluded that for Universonline there is some scope for improvement concerning checking their facts.

20141218122507

Reliability in citizen journalism

The information era we’re living in has allowed journalism to be practiced by basically anyone. Professional journalists might argue about the quality but it is difficult to disagree with the occurrence of the phenomenon. People can search for newsworthy items on social media, check online for multiple verification sources, and share personal experiences for everyone to see (instead of, sometimes, waiting and hoping to get some awareness by official news stations). The last example was witnessed in events that occurred in Egypt and Libya, and now, in Syria. This video shows an example of something called “citizen journalism”, where a group of people stood up and risk their lives to report on the situation in Syria as it was too dangerous or impossible for foreign reporters to enter.

What is citizen journalism?

billboard1_citizen_journalismCourtney C. Radsch, an American journalist who has been interviewed extensively about digital activism and social media in the Middle East, defines citizen journalism as:

“An alternative and activist form of newsgathering and reporting that functions outside mainstream media institutions, often as a response to shortcoming in the professional journalistic field, that uses similar journalistic practices but is driven by different objectives and ideals and relies on alternative sources of legitimacy than traditional or mainstream journalism.”

Reporting from a certain viewpoint

A reasonable question one can ask is whether citizen journalism affects the quality of news, and if yes, in what way. News may seem credible when it is reported by someone who was an actual bystander and was able to take a picture or video footage to include in the story. However, credible news is not always reliable news. Of course this applies to basically every form of news reporting, not just in the case of citizen journalism. Media have always been used for different purposes: to unveil the truth, hide it, or alter it – whether that’s done consciously or unconsciously. The ‘problem’ (if it can be called it that) is that in cases of citizen journalism one often deals with people who are either pro Group X and people who are against it. According to The Washington Post, a report from London-based Channel 4 suggests that Syrian citizen journalists may have been determined to embellish the truth. The low quality image below is from one of the available videos which were used to report on the crimes against humanity taking place in the city of Homs. However, the person who uploaded it admitted to have altered it by adding clouds of smoke in order to increase the chances for international attention.

2014-12-07 12_26_54-Mideast_Syria_0f622.jpg (606×333)

Another example that provided insights into a specific viewpoint was seen during Occupy Wall Street when video reporters emerged to tell the story of the protest. Many of them got a a large amount of followers quickly, even though these reporters clearly mentioned that they didn’t see themselves as journalists. Check out this impressive video about Tim Pool, one of the Occupy Wall Street video reporters, who got to a point of reaching a total of 750.000 viewers during a live stream.

Reliability & real time fact-checking in citizen journalism

One of the interesting things that Tim Pool highlights in the video is how he looks towards reliability and fact-checking as a citizen journalist. He agrees that there is no such thing as objective journalism, but he does try to be unbiased and to tell and show the truth. Additionally, he points out that the two-way communication that he has with his audience is the most important tool he has where social media is the main player. When Tim is on the ground, in the middle of a demonstration, he doesn’t have producers with him or a marketing team. What he does have are his thousands of followers who share screenshots of his live streams, talk on Twitter about what they see, or make highlights from his videos. Most notably, people who are really trying to follow what is going on are fact-checking for Tim in real time as followers are checking the information of other (video) journalists to check if they report the same, turning it into a support network.

Professional journalists in a time of citizen journalism

News stations don’t have the resources to make sure that everyone can be at every interesting location. Citizen journalists as Tim Pool, however, are perfectly capable of embedding themselves within a moment, to get a sense of reality by being a part of it and sharing that experience with many others. Depending on the event, it’s starting to become a bit old fashioned to search for witnesses and make notes on their experiences. Now, anyone who is connected to the internet and has a device to make recordings with is a source and capable of sharing the content. The amount of possible witnesses sometimes grows to thousands of people and content keeps being shared in vast amounts. Therefore, experts are needed who can look over all this incoming data and find out what is accurate and what is not. This will enable us to check whether credible sources are also reliable and to get to the truth as close as possible. One example of a news station who is doing this extensively is BBC’s user-generated-content desk where reports from thousands of different sources are being analyzed. Just as in some previous posts on this blog, this again addresses the necessity for improvements and applications in journalistic data visualization techniques.


Sources used in this article:

Is fact-checking in journalism becoming history?

When it comes to fact checking, different viewpoints can be found on the internet regarding its development. On the one hand, some sources report a rise of fact checking sites, such as the launch of three new fact-checking websites in Australia that are dedicated to critically analyzing political statements of those in the need of votes. On the other hand, other sources report a decline of fact checking in the world of journalism. Could the rise of fact-checking websites be a response to the decline of fact-checking in journalism? Maybe. In this post we will look at a possible explanation of the decrease in fact-checking among journalists, about why and how often this happens and the possible consequences for news production.

An obstacle in fact-checking: you
One of the first problems in performing fact checking might actually have to do with yourself and can be explained by a term from the field of psychology: cognitive dissonance:

“Cognitive dissonance is the mental stress or discomfort experienced by an individual who holds two or more contradictory beliefs, ideas, or values at the same time, or is confronted by new information that conflicts with existing beliefs, ideas, or values.” – Leon Festinger (1957)

What if, let’s say, you’re a pro meat eater and you stumble upon research that shows long term benefits of a vegetarian diet (or the other way around)? How would this affect your writing as a journalist when you have to report on something that doesn’t match with your personal beliefs? Maybe you would search for research that states the opposite or focus on the limitations of the study. For people, it is very uncomfortable to be confronted with information that doesn’t match with the way they act and they want to get rid of this feeling as soon as possible. Since journalists are just human too, it would be naive to say that cognitive dissonance is out of the question in the world of journalism (although it should be!).

Is time really another obstacle?
The link seems easy to make between the expected speed of news reporting and the lesser amount of time journalists get for thorough fact-checking. However, media writer Tim Dunlop presents some great food for thought in one of his blog posts, where he suggests something deeper than the blame on time pressure:

“Journalists adopted certain practices as standard that actively mitigate
against them presenting the facts”.

By this he means using writing technique that include ‘he said/she said’ reporting. If someone has taken the effort to actually perform a decent fact-check, this means that the facts must be presented within a specific context by making a judgement:

“Presenting facts calls for making a judgement, for deciding that one side is
telling the truth and the other isn’t”.

And then the fear comes in, fear of being judged for things reporters don’t want to be judged for and being accused of bias. It’s much safer to present information not as either true or false.

The impact of social media on fact-checking
A report (June, 2014) by ING provides very interesting insights from journalists and PR-professionals regarding the impacts of social media on news production and dissemination. A total of 165 journalists participated and most of them were from Great Britain, USA or The Netherlands. Some questions can be asked about the meaning of the results. For example, no tests regarding significance and reliability have been conducted, but nevertheless, the results do provide interesting insights that might give some ideas about social media’s influence. Let’s look at some main findings that are related to fact-checking:

‘Publish first, correct if necessary’ seems to be the new motto and fact-checking is done less thorough. According to the report almost half of the participating journalists stated to publish articles as soon as possible and to correct it later if necessary. Eighty percent of the journalists said that they occasionally publish something without fact-checking. Additionally, PR professionals stated that they noticed getting contacted less by journalists to fact-check since the emergence of social media.

2e8619fe-061c-4369-a26d-749d084f817d_publish-soon_688

One way of how the ‘publishing rush’ affects news is that more and more articles are being copy pasted from one website to another. This usually includes references to the original writer, however, often without applying any fact-checking techniques. Sometimes, this can have devastating consequences for individuals who are involved in the story by mistake, for example in the case of a young man named Abu Bakhar Alam from Australia. An article was being written about a teenage terrorist suspect for which a journalist used a photo that he found on Facebook, because he thought the person in the photo was the actual suspect. It, however, wasn’t him. Instead, it was an innocent 18 year old who was now the face of a suspected teenage terrorist in multiple newspapers that copied the image. This caused an innocent man to become terrified of leaving the house, due to fear of people wanting to harm him or his family for something that he had nothing to do with, simply because the newspapers did not check their sources and the correctness of the image. As a news institution, the last thing you want to do is to apologize for mistakes like these: copy pasting unverified information, turning you into an unreliable news source by letting the audience down.

991386-822f7c36-44f8-11e4-8ff2-fe0ffc289a61

Other results of the report showed that 40% of journalists consider information on social media to be reliable. Consumer opinions, however, are considered more reliable than those of organizations. Furthermore, while the majority of journalists don’t see social media as a reliable source, about 50% does view social media as the most important source for information. Finally, 59% said that public opinion on social media played an important role in what the journalists publish. An interesting question to think about is related to last week’s topic: agenda-setting. The results make it seem as if the public has more and more influence on the agenda of the media instead of the other way around.

The future of fact-checking
Based on the findings and development of fact checking, some ideas can be sketched about the future of fact-checking:

*    Less fact-checking, more crowd-checking: Many journalists turn to social media for obtaining information. The news dissemination in social media occurs with great speed that has never been seen before in human history and, according to the ING study report, professionals indicate that the speed of social media creates a preference for publish first, correct if necessary”. A majority of journalists expect to see a further decline in fact-checking, whereas crowd-checking, where the public’s opinion is used and accepted as a fact, will grow in importance. We’re living in an era where everything is expected to be communicated fast, obviously with consequences for the quality of news.

*   Journalism will be driven by clicks and views: When people think about journalism, some may think about revealing complex stories, making the invisible visible and report on serious issues that are happening nearby or on the other side of the world. However, writing for an entertainment website or magazine is journalism as well, which is clear from this definition from http://www.oxforddictionaries.com: “The activity or profession of writing for newspapers or magazines or of broadcasting news on radio or television.”  This makes the newspapers, magazines and broadcasting shows no different from any other commercial company that needs to make profits to ensure their existence. Therefore, obtaining a high amount of clicks and views becomes more important than quality which could lead to more sensation driven news with a lesser focus on quality and content.

Fact-checking sites: not a cure, but a symptom of a serious problem
As mentioned in the introduction of this post, independent fact-checking websites seem to be on the rise while fact-checking in the field of journalism seems to be declining. It might be very easy to say that we should be happy for their existence, which was my primary thought. However, Tim Dunlop’s blog post made me realize I have stopped thinking too soon. It’s quite possible that these two developments are linked to each other, as a reaction to a problem. Because are the fact-checking websites really a cure? Maybe they are a symptom of media and journalists not doing their job. It’s worth thinking about why these independent websites emerge and what the concequences are on the long term. Dunlop suggests that outsourcing the fact-checking process to other sites will cause stories to be reported in the same way: as something that someone else said and report what the fact-checking site presented as facts – the ‘he said/she said’ technique:

“They [news organizations] will be almost by definition turning the facts presented on the fact-checking site into just another opinion, thus undermining its authority and leaving us precisely at the point we were at before the fact-checking sites arose from the ashes of mainstream journalism. They [fact-checking sites] are not solving the underlying problem. They are, in fact, enabling the very disease they are setting out to cure.” – Tim Dunlop


Sources used for this article:

Video games: victim of agenda-setting and framing

ambrose-bierce-technology-quotes-telephone-n-an-invention-of-the-devil-whichIf you’d buy a video game in the nineties, an average person would probably perceive you as a huge nerd. That attitude, however, has changed and gaming turned into a more socially accepted activity. It even turned into a multi-billion dollar industry which must say something about its popularity and cultural acceptance. However, just as with many new popular inventions in general, there is always a group of people who is afraid of it. We’ve seen this before with  television, rock music and the telephone. In this century, it’s hard to imagine there would be people criticising the use of a phone. This is because there is no generation left, at least in the western culture, who never used a phone before. This was the case, however, over a hundred years ago. Preachers said that the phone was the instrument of the Devil, men worried that their wives would spend too much time gossiping, while others believed that phone lines provided access to evil spirits (and telemarketeers proved them right!).

Agenda-setting
In the context of journalism, media, and politics, the question that arises is: what happens if an influential person fears, let’s say, video games, and feels the need to protect people and do something about it? The reason for choosing this example is because it’s a realistic one. Many of you have most likely read or heard something about it in the media and given it some thought yourself which causes you to have a personal opinion about violent video games. According to the agenda-setting theory, the reason why you have given this some thought, and maybe had discussions about with others, is because media are to ‘blame’.

Are agenda-setting and framing actually one thing?
According to the agenda-setting theory, media influence what we think about (agenda-setting on the first level), while framing puts a ‘frame’ around an issue which influences how we think about it. While agenda-setting and framing are considered to be two different things, they do go hand in hand. On the one hand, media choose to focus on certain topics which causes to make some issues more important than others, but on the other hand, these topics are always discussed in a certain way, from a certain point of view. Therefore, when discussing the agenda-setting theory, it is difficult leaving out the framing part. This is actually argued by some scientists: Balmas and Sheafer (2010) argued that the focus at the first level agenda-setting is shifted to media’s function of telling us “how to think about it” at the second level which considers how the agenda of attributes affect public opinion (McCombs & Evatt, 1995).

Video games and agenda-setting
For some reason, violence in video games got on the news radar. It’s a complex process with not one simple answer, but it may have been the result of different events and activities: in the previous century, many studies were already done about the effects of television on behavior, newspapers were quick to report associations between violent media and tragic events, national organizations and websites arose to provide advice and media ratings for parents (for example regarding drug use, sexual and violent content), and sometimes an influential person stood up to express concerns regarding the violence in video games which caused some countries to discuss about banning certain games. Columbine_Shooting_Security_CameraAccording to Christopher Ferguson, clinical psychologist and leading expert on video game violence, the Columbine high school shooting in 1999 was a crucial moment for the debate of violent video games and also a moment for getting more research funds since more institutions became interested. Of course there were more incidents before that were linked to video games, but at that time, this one was the biggest and scariest and got a lot of attention in the media. It was the scariest because people realized that mass killings could happen to anyone, anywhere, even in a high school that was considered to be located in a safe neighborhood. And from that moment on, every new case, where violence could be linked to a video game, was mentioned.

For example, GTA has been mentioned many times by journalists, politicians and organizations worldwide to make a statement about the negative influence it would have on people. This 19 year old student for example, who robbed and stabbed a taxi driver to death so he could afford the game, or these teens who confirmed to reenact violent actions from the game in real life. Or remember Anders Breivik? He said to have used Call of Duty and World of Warcraft to train for his assault in Norway. In turn, because such cases and the concern regarding violent games were growing, it was almost not possible to not report about new cases. Because if you don’t, some other news agency will. And before you know it, people were bombarded with headlines like these. By just looking at it, it’s quite imaginable that parents who have no personal experience whatsoever with video games, might really get concerned.

News4

What’s ‘funny’ is that this attention tends many to forget about the millions of ‘normal’ gamers out there who play violent video games too, but don’t commit any crimes nor show signs of aggressive behavior. It’s almost the same trick that lottery institutions are doing: you always see that one person or neighborhood who won something, but never hear anything about all those people who lost the money.

I bet they all drank coffee too! Therefore, coffee must turn everyone into killers!

I bet they all drank coffee too! Therefore, there must be a link between coffee and killing!
*cough* false correlations *cough*

Video games and framing
The framing strategy is often used by influential people or activists. They are often looking for arguments that support their belief or change the information in a sly way to make it fit with what they want to preach (cognitive dissonance to the max!). Often these people also use scientific research to increase their credibility. The question, however,  whether there is a causal effect of violent video games on (short / long term) violent behavior is very complex and has no definite answer (yet). Multiple factors such as characteristics, environment, genetics, and behavioral features may all play a role for aggressive behavior, but the link between violent media and violent behavior seems very easy to make for a lot of people. This can particularly be seen by the interpretation of research which is reported about in popular media. For example, this Dutch article with the headline: “Everyone is addicted to games”, which is about Maria Haagsma’s dissertation. The title, however, is heavily dramatized and suggests that her research was mainly about this topic. Her research was actually about ‘problematic gaming behavior’ which does not equal addiction! Furthermore, the last sentence of the image states that 3 percent of Dutch gamers are addicted. That doesn’t really match with the headline, does it?

uk-newspaper-runs-series-yellow-anti-video-game-articles.w654

Another example is when a researcher states: “We now have conclusive evidence that playing video games has harmful effects on children and adolescents.” And the next day you will see a headline with “Video games dangerous for children!” and an article that presents it as a causal effect. Research by Dr. Craig Anderson has been mentioned quite some times in the media, but in my opinion, his research is biased because he is part of an active anti-gaming movement. Nevertheless, journalists and other fellow activists love using the strategy where they can say that scientific sources have been used (without having a critical look).

SiGray.Recombu-6375Think about this proposal: what if aggressiveness was measured by letting someone hit a human dummy before a soccer match and during the match? Don’t you think the player would hit the dummy harder during the match when he is filled with excitement and adrenaline? But can we then say that person will become more aggressive because of soccer? Is that a viable conclusion? The same might apply to gamers. It is nothing new that while playing action video games one’s heartbeat and adrenaline rises as they are trying to achieve their goal. So in some of the previous research, the critical question is: what exactly has been measured and is it right to state that it makes people more violent? In my opinion, many people tend to stop thinking too soon, which comes in handy for anti-gaming movements and journalists who want to use scientific sources to increase their credibility.

A personal look at agenda-setting and framing as a gamer
As an avid gamer myself, it’s interesting to have experienced framing and agenda-setting about a topic that I am familiar with. Over the recent years, I have indeed noticed an increasing ‘fuss’ about violent video games (although the amount of attention seems to decrease again and more positive messages can be heard – maybe because more research is done?). Because I’ve played games all my life, I could be biased (just as the anti-gaming activists), but I see that the experience allows me to have a critical look regarding things that are mentioned in the media. Especially because I noticed that many people who don’t have gaming experience use a lot of arguments that have been presented by journalists and news stations. Therefore, as I have mentioned at the beginning, agenda-setting and framing are like brother and sister, because news is always reported in a type of context where a certain opinion is presented, either directly or indirecty.

Data visualisation – making the invisible visible, or maybe not?

“It feels like we’re all suffering from information overload or data glut. And the good news is there might be an easy solution to that, and that’s using our eyes more. So, visualizing information, so that we can see the patterns and connections that matter and then designing that information so it makes more sense, or it tells a story, or allows us to focus only on the information that’s important.”  – David McCandless (TedGlobal, 2010)

Many would agree with expressions such as “seeing is believing” and “a picture is worth more than a thousand words”. Research has also shown that visualisation provides cognitive aids as it augments human memory to provide a larger working set for thinking and analysis (Kerren et al., 2008). Therefore, data visualisation is a very effective method to turn abstract, complex or a large amount of information into something insightful and meaningful, making it a valuable part in the field of journalism. However, data visualisation also has a dark side and in this post we are going to discuss both good and bad data visualisations by showing some examples.

DEFINING GOOD VISUALISATION
But what defines good visualisation? Professor A. Cairo mentions some good points. Visualisation should be functional, meaning that the right type of visualisation should be chosen to present the information (e.g. map, chart, infographic, cartogram, network, scatterplot, etc.). For example, the picture below shows multiple ways of presenting the same data. It is your task as a journalist to not pick your personal preference, but to choose the one that tells the story easily, efficiently, accurately, and meaningfully. For starters, here you can find a functional chart (see what I did there ;)) that shows some suggestions for choosing a visualisation type.

tableau-visualization-example_illustration

Visualisation also should be beautiful. They don’t always need to have a high aesthetic value (such as in scientific papers), but in journalism, however, the attractiveness of the presented information does contribute to whether a reader will interact with it or not. Some guidelines that can be followed on aesthetics are the Gestalt principles making visualisations both aesthetically, as well as cognitively attractive. Furthermore, visualisations should be insightful. They should reveal things that are unexpected, surprising or extremely important. In other words, things that cannot be spotted just by looking at the raw data should be made visible in the visual. Finally, great visualisations are usually enlightening, meaning they should be newsworthy by revealing things that the reader didn’t know about before.

INNOVATION IN DATA VISUALISATION
In an insightful and amusing Ted Talk, statistics guru Hans Rosling presents data in a new and playful way. By using his Trendalyzer software (now owned by Google), he is capable of showing an amazing global development. The gif below shows a fascinating part of his presentation where the y-axis represents the life expectancy (in years), the x-axis shows the size of families (number of children) and the bubble size represents the size of the population of one country. What is amazing is that it allows you to see a global development trend in just a matter of seconds. While most third world and developing countries are located in the lower right corner in the 1960’s (large families, short life expectancy), we can instantly see that their life expectancy and family size becomes similar to those of developed countries (small families, high life expectancy). This is an enlightening result, as many people might still think that the current situation (2003 in the graph) is similar to the one in 1963 – while it is obviously not! You might wonder, what are those green dots that stay behind? There is an insightful explanation for that: HIV. Those green dots represent African countries and in the 1990’s the epidemic caused the life expectancy in those countries to drop again, while other countries kept moving up. If data visualisations like these excite you, make sure to check his presentation for even more examples where he uses different visualisation types to show other global trends, such as child mortality and poverty rate.

HansRosling_TEDTalks_Animation (1)

DATA VISUALISATION – HOW IT SHOULD NOT BE DONE
The previous example showed some great and advanced data visualisation techniques which are particularly valuable in the case of enormous and multiple data sets. But of course, this doesn’t mean that less advanced visualisations, such as graphs and tables, will disappear from the table. Everyday, many people are confronted with business, sports or politics related statistics and just as with any kind of information, data visualisations can be false or misleading. Let’s have a look at several examples which show that data visualisation can be misleading and how it should not be done.

“Charts, graphs, maps, and diagrams don’t lie. People who design graphics do.” – Alberto Cairo

If a picture says a thousand words, the two images below should instantly show the problem: a truncated Y-Axis. By using different scales for the Y-Axis, this technique makes differences look much larger than they really are; the interest rates in the left graph seem to be increasing almost exponentially, while if you look at the graph on the right, you get a totally different view of these rates.

misleading1_yaxis

People are used to having the (0,0) coordinate in the lower left corner of a graph. Because the following examples ignores this convention, it seems to present the opposite of what is really going on. The graph suggests that deaths caused by gun violence in Florida are decreasing, while they are actually increasing.

misleading3_deaths

3D pie charts are often used for aesthetic reasons, but the third dimension does often not improve the reading of the data, which causes false 3D interpretation. While item C looks larger than item A in the left pie chart, in actuality, it is less than half as large, meaning it doesn’t represent the quantities accurately.

325px-Misleading_Pie_Chart 325px-Sample_Pie_Chart

Sometimes, certain values can be removed from the data, for example, to show a more positive result, which is called omitting data. Have a look at the following example, which could be coming from a financial report. When some values are excluded, it makes the outlook of the left graph look more positive, creating a more favorable visual impression.

350px-Scatter_Plot_with_missing_categories.svg 350px-A_scatter_plot_without_missing_categories.svg

DATA VISUALISATION: IMPACT AND CHALLENGES
Data visualisation is capable of changing the way we see the world. Rosling’s example may have changed your view about the global life expectancy and size of families. The thing is that, absolute numbers, such as the amount of soldiers in a country, do not provide the entire picture. For example, China tops the list with 2.1 million soldiers, but drops down to the 124th place if you look at the soldier-population proportion. And this is what data visualisation is allowing us to do these days: connecting vast amounts of different data into one understandable figure. The fact that more wonderful design tools and more interesting data become available, allows us to have other ways of looking at the world that was impossible before. Therefore, the more important it becomes to pay extra attention to the presentation of this data. Not only should these tools pay careful attention to the design of data presentation, but also try minimizing the chances of visualising it in a way that causes confusion and misinterpretation (which I think is the most important part!). Therefore, I couldn’t agree more with the following citation about the need in data visualisation software, especially if you keep in mind the problems of data interpretation while reading it:

“What’s most needed in the field of data visualization, as in other fields, is not always what’s most exciting or not even what’s particularly innovative. Sometimes we simply need to make it easier to do those things that work. One example of this is the effort of a few software vendors to build data visualization best practices right into the tools, such as in the form of defaults, thereby making it easier and less time-consuming to do what works and harder and more costly to do what doesn’t.” – Stephen Few (2013)

Sources used in this article:

http://www.ted.com/talks/david_mccandless_the_beauty_of_data_visualization
http://www.win.tue.nl/~vanwijk/infovis_springer.pdf
http://www.interaction-design.org/encyclopedia/data_visualization_for_human_perception.html
http://data.heapanalytics.com/how-to-lie-with-data-visualization/
http://extremepresentation.typepad.com/blog/files/choosing_a_good_chart.pdf
http://en.wikipedia.org/wiki/Gestalt_psychology
http://www.ted.com/talks/hans_rosling_shows_the_best_stats_you_ve_ever_seen#t-148297
http://en.wikipedia.org/wiki/Misleading_graph

Data in journalism – the good, the bad, the ugly

Be able to analyze statistics, which can be used to support or undercut almost any argument.” Marilyn vos Savant

Dealing with numbers, statistics and big data is of no value if one does not know what to do with it. Being able to work and interpret data provides advantages in almost all thinkable fields and has become a valuable skill in journalism as well. It allows us to tell stories that are backed with  numbers, to see things that would not have been visible otherwise, and encourages trust as then it should be able to (partly) replicate a story that was based on those numbers. So in terms of verification, which was discussed last week, this can be seen as a huge advantage in the field of journalism.

Data journalism in the early days
One of the classic examples that shows the value of data journalism happened way back in the 1850’s. John Snow (no… not the one from Game of Thrones, but a physician from England), was very skeptical about theories and rumors regarding the transmission of cholera. These theories did not allow him to understand what spread the disease. By making a map that showed dots of all people who died due to cholera or were infected, he was able to make a remarkable discovery. The public water pump was identified as the source and the discovery contributed to the end of the outbreak in London. In other words, by visualizing data – because that’s what he did by creating the map – he was able to make a discovery that might have gone unnoticed otherwise.

Snow-cholera-map-1

Data are not sacred
Why has data journalism become such a hot topic? Why now? First of all, technology allowed us to produce and store a vast amount of data of which many are available to anyone. Second of all, many tools are available that can be used for making maps and visualizations without requiring any coding skills. However, data should not be taken as a sacred truth. Anyone who knows even a little bit about statistics, also knows that interpretation mistakes can be made easily.

How to question numbers
It is important to keep in mind that data, like any type of information, can be wrong. Just because numbers might seem objective, this does not mean that they should not be questioned. So how can one question data? Here are some suggestions:

  • The Who: Just as with any acquired sources, journalists should always be skeptical towards the information they obtained and especially source they acquired it from. Furthermore, if you know where the data came from, this provides transparency for your story which increases your trustworthiness as a journalist.
  • The What: Great, you’ve acquired yourself some data sets! The question is, what do you want to do with it? What are you capable of telling with these numbers? In this part, your job as a journalist is to translate raw data sets into something meaningful. Not everyone has time or has the knowledge to interpret statistics so many are relying on you as someone who correctly interpreted the numbers.
  • The When: This is about checking the recency of your data and asking yourself whether it is still valid. By the time you get the data, the numbers might apply to a situation about a year ago. Therefore, many journalists prefer using methods that are more likely to provide up to date data, such as near real time data, social media data, sensor journalism and crowdsource reporting.
  • The Where: One of the things that can be done with statistics is to look at data that originates from different places around the world and perform checks into something that you are interested in. For example, one could make map of countries that do and do not legalize marihuana and the amount of drug related violence.
  • The Why: The Where question, however, might cause some serious errors. On the one hand, data visualizations can show what is going on, but a correlation is not always a causality. Therefore, dealing with data sets requires some knowledge about statistical tests and analysis. Otherwise, mistakes can easily be made when it comes to reporting casualties or relationships that seem to make sense, but in reality are false.

Speaking of false correlations, check out this page that shows some funny correlations between two variables. How would you respond when reading a headline that says “Divorce rate linked to margarine consumption”? Our common sense tells us this does not make much sense, but it becomes problematic when it does seem to make sense (such as bullying and suicide rates) and is then reported as a causality.

screen shot 2014-05-09 at 11.52.35 am

“Antidepressant use in England soars”
Let’s have a look at a serious example: an article that was written on theguardian.com. The main topic addressed in the article is that antidepressant prescriptions have increased by more than a quarter in three years, suggesting that more people in England are suffering from depression. The journalist had at least two data sets about current prescription rates and those of three years ago, which were then compared. The thing is, he could be right, but he could also be very wrong. In this case, comparing data set A to data set B is not sufficient, as other factors may play a role in the increase of prescriptions. What if the age group that get’s mostly affected by depression has increased? In that case, the number alone does not say much. It’s the proportion that matters. Or is it possible that doctors are writing more frequent  prescriptions in smaller amounts for the same amount of patients? Such questions show that more numbers are needed to conclude that the amount of people who suffer from depression has indeed increased.

Journalists need to embrace statistics
Some, however, question whether data journalism can be considered journalism. I find it very difficult to understand why someone would not consider it to be journalism. We are living in a technological era that provides us with a lot of data and this can be seen as new and emerging sources that are capable of providing evidence or to serve as a supportive factor in a bigger story. Data provides new and additional ways of verifying news stories, which will always be an important part of journalism. It does, however, require to keep some ‘W-questions’ in mind that have been discussed before. First of all, journalists should bridge the gap between raw data and the readers. Raw data often has no meaning. This needs to be created. Therefore, journalists should ask themselves critical questions about what can be told with the data they acquired to create meaningful stories. Second, this meaning and interpretation should be correct, of course. With so much data available about so many variables, it is not a difficult task to find correlations. Although people tend to be scared of numbers, (digital) journalists are not allowed to be scared anymore. No matter at what educational level someone acquired a degree in journalism, it is almost unacceptable to not teach anything about statistics to future journalists as they will, most likely, need to deal with numbers more and more often.

————————————————————————————————————————————-

Sources used in this article: 

A love and hate affair with news on social media

Although most of us probably know that the amount of data shared on the internet is enormous, seeing the actual numbers is quite amazing. For example, every minute around 278.000 Tweets are send, 72 hours of video material is uploaded on YouTube and 1.8 million likes are generated on Facebook. According to Eric Schmidt (Google’s Executive Chairman):

“Every two days, we create as much information as we did from the dawn of civilization up until 2003.”

infographic-resized1 

Social media and changes in our news acquisition
Due to this phenomenon, one can think of several ways our news acquisition has changed over the recent years. People are no longer limited to consuming news from traditional media institutions, but they can also select and follow news that matches with their personal preferences. For example by joining communities and following specific pages on social media that report about topics that personally interest someone and which can be read at any time of the day. This allows users to find out more about topics that are not often discussed in traditional media sources. Individuals and non-media organizations are now capable of highlighting issues because they feel a need to spread the word about a certain topic, for example about abuse in animal factory and fur farms. This is actually something I personally experienced, which changed my attitude towards the use of animal products entirely and I think it’s quite safe to say that I am not the only one whose attitude towards a certain topic has changed due to social media, simply because it allowed us to witness information from a new or different perspective.

Everyone can be a reporter
Apart from possible attitudinal changes, social media allows everyone to share news online. It can give the audience a glimpse and updates about live events that may not be accessible to everyone (yet). Today, anyone can capture and share valuable, up to date, images. Remember this tweet that went viral in 2009?

twitter-hudson

It was made by a bystander, Janis Krums, who happened to be on the “right” place at the right time. Within minutes, news networks started contacting him for interviews and LA Times stated that:

This may be among the most striking instances yet of instant citizen reporting”.

A journalist’s dilemma
The example above shows how valuable citizen reporting can be to journalists and media institutions. Therefore, it isn’t strange to see numbers (like in this report) that show that many journalists consider social networks to be very important to their work and daily spend some time on these networks, preferably microblogs such as Twitter, to check for breaking news (Willnat & Weaver, 2014). With so much content and news being shared every second, the problem is, however, that it is difficult to check the validity a post, which is one of the challenges in digital journalism. It can be a hard task finding the original source of a news post and people may simply share fake or incorrect information to cause confusion or to obtain “likes” and new followers. Maybe you remember some of the fake hurricane Sandy pictures where two photos were combined which made a location look like a set from the movie The Day After Tomorrow. Or the hashtag #BaldforBieber that encouraged fans to shave their heads to support the pop star who was diagnosed with cancer, according to the false rumor. Or, a more serious hoax, the threat of a tube bomb that discouraged people from using the London Underground as you can see below.

_77295295_nestatext_77296115_btp_twitter1

Dealing with the verification problem
Journalists have always dealt with verification problems, but verification in the digital age might have become even more difficult as many techniques can be applied to make a picture look real, to make a source seem trustworthy or to present facts in a way that seem to make sense. Let’s look at some elements that often need to be checked and confirmed to verify a piece of user generated content.

  • Confirming the authenticity of a piece of content: As mentioned before, journalists prefer scanning messages on microblogs for breaking news. However, tools such as www.lemmetweetthatforyou.com make it possible to easily fake a tweet. One of the things that can be done is checking whether the Twitter account is a verified one. This can done by hovering over the blue “check” icon, next to someone’s name, which should then say “verified account”. Whenever the information is accompanied by a picture, reverse image search methods can be used such as TinEye or Google Images, which allow to search for previous use of a particular image.
  • Confirming the source: This is about finding the original uploader of the content. The ideal goal would be to get in touch with the original source. This is something that gets easier as many people have multiple social media accounts where some contact information is usually shared. Whenever a journalist manages to get to talk to someone of whom he thinks he is the original uploader, the questions that should be asked are: where were you standing when you took the footage, what did you see and what type of camera did you use to record the footage? The key is to ask direct questions, because these type of questions seem to lead to admitting fake news.
  • Confirming the date: Confirming the date of an event can be quite difficult and one of the tricks that can be used is to search for other sources that report about the same event or check if the weather in the video or photo corresponds with the actual weather. WolframAlpha is one of the available tools that provides weather information on a particular date.
  • Confirming the location: Fake videos or images might have errors in characteristics of the surrounding. For example, a fake photo might not display a building on the background that should have been present on the date the photograph was taken. This step is about checking whether the surrounding of an event corresponds with the actual surrounding. Some useful tools that can be used for this are: Google Maps, Google Earth and Wikimapia.

The time of a broadcast publishing model seems over and digital journalism is facing many difficulties and challenges. On the one hand, social media play a valuable role in quickly acquiring important pieces of information that are part of a larger puzzle, but on the other hand new techniques need to be applied and invented to find out whether that piece of information is actually true. Furthermore, social media also made news quicker. Something critical that occurred at 8pm might have went viral on Twitter already but one might read about in the newspapers 12 hours later. The quickness of today’s news may put even more pressure on journalists to find and publish more in a shorter amount of time. The question that arises is what this pressure does to the validation of news sources which can be a time consuming activity. At the moment, when trying to validate a piece of information, to me the process seems inefficient and difficult because it requires journalists to search for information on separate websites and separate social media networks. Based on the amount of confirmed information, there comes a point where one has to decide to consider a source to be trustworthy or not as a verification of 100% is nearly impossible. I believe that this is not only the case for journalists who want to write a news story, but also for news consumers who need to have some form of responsibility when it comes to figuring out who to trust. Especially in today’s media landscape where journalists are working under high pressure and where many people are daily confronted online with dozens of news messages that do not originate from official reporters.