You are here

Media in The Spinneys of Opinion Surveys

Properly performed opinion surveys are necessary. These can, for example, reveal the situation in the society at a given time, can identify certain trends, suggest intervention measures to maintain the social balance etc. The bad thing is that we cannot always check if a survey is carried out properly. Much worse is that we usually find out about the results of the surveys from the media, which most of the time present them in a distorted manner.

A first example: a news portal announced on October 12, 2015: ”(SURVEY) Moldovans no longer want to join EU; Eurasian Union, more attractive”. And then: If a referendum would be organized next Sunday and the citizens would have to choose between joining the European Union or Eurasian Union, 36,7% of them would vote pro-European Union, while 40,9% - pro-Eurasian Union. The results of the survey show that 11, 7% of those surveyed could not provide a concrete answer related to the option they would choose in a referendum. 6, 1% of respondents say they would not attend, while 4,6% of citizens answered they do not know what their option would be.

We do not think that the author of the news meant to lie to us. Probably, he did not understand the survey data and wrote about in the best way he could, however if 36,7% of citizens would vote pro-European Union, it does not mean that all ”Moldovans no longer want to join European Union”, as the news headline suggests. If the author analysed the survey data, he would have noted, maybe, that over 16% of respondents did not provide a concrete answer and do not know how to vote, yet. And if he considered the margin of error of about plus/minus 3%, he would have certainly written the news differently.

Another news portal, provided us with the following information on September 29, 2015: ”SURVEY: Renato Usatii enjoys the highest trust from the population”. A first question would be: from the population or from the respondents?! The question is reasonable, if the news communicates us that R. Usatii is credited with”… 31% of those who indicated a name. If we also consider the number of those who did not answer the question or indicated that they trust nobody, than Renato Usatii enjoys a confidence rate that falls to almost 9%. When the respondents were asked about the political personalities they trust more, Renato Usatii and Igor Dodon were the first two. It is important to note that over 70% of respondents did not indicate their name”. If so, how accurate the headline is?!

The results of the survey carried out by the Association of Sociologists and Demographers, presented on October 12, were widely broadcast. But in what way? Here are some of the headlines: ”A new survey: Usatii and Dodon, leaders of the electorate preferences” (of the electorate!? Are you sure?);
Survey: Radical changes in the Parliament. A new majority in case of new elections” (really?!).;

Let us compare two headlines and see the differences: “Plahotniuc, considered by the citizens the main culprit for the situation in the country”, and Plahotniuc, the main culprit in the Republic of Moldova // SURVEY”. In the news with the first headline the author insists that the citizens believe that Plahotniuc is responsible for the situation in the country, while in the second headline the author specifies: ”Most of the respondents in a survey organized by the Association of Sociologists and Demographers believe that the oligarch Vladimir Plahotniuc is responsible for the situation the Republic of Moldova is in. Did you notice the difference? This is what we must seek in a reporter’s work - exactitude and accuracy, just like in the following news, entitled: ”Survey: 79% of Moldovans opt for a direct election of the Head of the State”. Otherwise, we can easily become the target of manipulations, because not all news are written in good faith. Let us recall a news from the previous electoral campaign, published in June 22, 2015 with the headline: Survey: Zinaida Greceanii, the favourite candidate in the second tour in Chisinau”. We know who won the elections in Chisinau and this is another reason to handle such products presented as news with all due caution. On top of all that, there are many cases where the difference between the data provided by surveys and the results of elections is much bigger than plus/minus 3%. In Russia, for example, the difference in 1996 elections was of 15%!

There are no laws to forbid media to use survey data, just like there are no laws to prevent it from being very prudent. For example, the British Corporation BBC established on its own initiative specific rules every one of their journalists observes, namely:

  1. the broadcasting shall not begin with results of surveys;
  2. BBC will independently interpret the surveys results, without necessarily providing the interpretation of data offered by the survey authors;
  3. BBC will avoid using expressions that could reinforce the credibility of survey results, for example, instead of the research data “show”, the research data “presume” will be used;
  4. will emphasize the margin of error and if the difference in results of two candidates is within this margin, it will be particularly specified.

In addition, BBC proposed that in case survey data are analysed, additional information should be provided: the survey author, the period when the survey was carried out, the sample, geography, methodology, questionnaire etc.

In Netherlands, the most serious newspapers adopted a self-assumed rule – not to publish survey results on the first page.
These are good practices to take over, and while the media ignores them, maybe we, the readers should ignore as well the truncated or, even worse, the manipulating information about surveys the media provides us.

media expert

This material is published within the project "Freedom of expression and media development in Eastern and South-Eastern Europe and South Caucasus", implemented by CIJ during the period May-September 2015, supported by Deutsche Welle Akademie and financed by German Federal Ministry for Economic Cooperation and Development. The opinions expressed in this material belong to the authors and do not necessarily reflect the financer’s opinion.