Presidential: Choosing your candidate on the web, a good idea?
Contrary to what one might think by following the polls which have followed one another and have been alike for months, the presidential election is far from over: generally, a fifth of readers decide on the weekend of the election.
On Slate we are told that the presidential election is not played on the internet, which during the campaign was less a tool of persuasion than of mobilization of already convinced voters. It may have been true throughout the campaign, but what about the last-minute indecisors? How many of them will turn to the internet to make their decision, and above all, what will the internet tell them?
Program comparators in spades
No doubt because they are very popular, we no longer count the comparators of programs like Voxe, but especially the tests or quizzes that are supposed to help you find out which candidate you are closest to, or who you should vote for.
Some are very primitive, like that of Rue 89 (which dates it is true from last year) where by answering “yes” or “no” to a series of less than ten proposals, you will end up being nominated for a candidate. At the other end of the spectrum, we find the Comparactor of politique.com which submits to you 427 proposals drawn from the programs of the candidates and proposes to you to note them on a scale with 5 graduations going from “Never of the life” to “Absolutely”. At the end of the day, we gives a percentage of affinity with each of the candidates (positive or negative percentage).
Fact checking, themes … other presidential applications
It’s not just tests and comparators: the site Presid-apps lists around twenty applications developed for the presidential election.
Among these applications, there are several dedicated to fact-checking. Their goal: to repeat the declarations of the candidates and to verify, with supporting evidence, that they were not mistaken (intentionally or not) in the facts or figures. The Veritometer developed by OWNI for i-tele, goes further by proposing a ranking of the main candidates according to their credibility. At present, the classification follows a classic left-right graduation surprisingly well, placing Mélenchon in the lead and Le Pen in the back of the pack.
The other major trend among these applications is thematic analysis: the Speech Observatory offered by Le Monde or the Thema-Tweets application propose to use automated semantic analysis to reveal the themes of the candidates’ speeches. .
Tools that are not quite neutral
To our knowledge, no campaign has yet had the idea of setting up its own “oriented” comparator. In 2017, will a candidate dare to propose such a test, all of the answers to which ultimately lead to the same result: vote for me?
However, this leads us to ask ourselves the question: are all these tests, all these sites really neutral? Without being able to take a look at their algorithm, difficult to say, but it is enough to have the smallest bases of marketing or communication to know that the articulation of a question, in particular a question with multiple choice, necessarily directs the answer given.
Polls are often criticized for manipulating public opinion, but by submitting oneself to these questionnaires, one lets oneself be manipulated voluntarily, as pure as the intentions of their authors may have been. Try to submit yourself to several different comparators and you will probably find that you will probably not always recommend the same candidate in the end.
Should we not trust the net?
Taken alone as a tool for electoral decision-making, the internet is probably neither more nor less dangerous than any other medium. There is very focused information and more objective analyzes, manipulations and revelations. What you should not believe is that because you are using a programmed application, it is purely objective. Behind every program, there is a programmer, who is neither more nor less neutral than you and me.
Politics is more than a series of proposals (which will or will not be respected after the election) to which we adhere or not, or than a set of speeches of which we can count every word. Far from being uninteresting, all these applications can prove useful as decision support tools but should certainly not be taken as the ultimate rational solution to political questions, otherwise we might as well let the machines vote for us. .