Fake polls and statistics: how numbers and graphs are manipulated

Sociological data and marketing research are an important tool for analyzing trends in politics, business, and public life. However, the methods of collecting and presenting information can significantly affect the final conclusions, which necessitates careful checking and verification of published results.

Falsification of the results of sociological surveys is a serious problem that affects the development of social processes and decision-making. Today, special attention should be paid to the use of unreliable statistical data in the virtual space, especially in social networks and news resources, since the Internet has become the main channel for transmitting information, including scientific publications and sociologists’ reports, and news containing accurate data in the form of numbers and graphs is more likely to be perceived by many users as something truthful.

Reasons

One of the most common reasons for organizing so-called «fake» polls is political considerations. Polls of this kind are often conducted with the aim of forming a desired public opinion, supporting a certain political force or candidate, influencing the outcome of elections, increasing the level of public support for a specific politician or government course.

Example: In May 2016, on the eve of the Brexit referendum, The Telegraph published the results of a poll showing an abnormal lead of the camp against the country’s exit from the European Union. This publication provoked a wave of criticism from experts, who pointed out serious methodological flaws: a sample of only 800 people, a poll only on landline phones (biased towards older and wealthier respondents), as well as statistically insignificant representation of key demographic groups (for example, only 6 unemployed and 6 black respondents across the country). The goal was to create the illusion of an inevitable victory of the pro-European camp, demoralize opponents and influence the outcome of the vote, but the actual results of the referendum (the majority of Britons supported Brexit) confirmed the bias of this poll.

Social aspects also play a role in motivating dubious surveys. Sometimes interested parties conduct such an event to promote specific social ideas, directing the public to support certain socially significant programs or events. False figures and conclusions help create the impression of broad popular support for a particular initiative.

Example: Anti-vaccination activists and groups have promoted ivermectin as a “miracle cure” for COVID-19, relying on fabricated or methodologically flawed studies. More than a third of the key studies cited by the drug’s proponents contained serious flaws or fraud, including duplication of patient data, non-random assignment of groups, statistically impossible numbers, and lack of access to the original data. The goal was to create the illusion of scientific validity to influence public opinion, undermine trust in mainstream medicine and vaccines, and stimulate demand for the drug. This has led to real harm: patients have delayed seeking qualified care in favor of ineffective treatments, in some cases leading to dire outcomes. The episode highlights how pseudoscience and data manipulation can exploit public hopes in a crisis.

An equally significant motivation for conducting unreliable surveys is the desire to extract economic benefits. Companies often initiate such studies in order to improve their public image, increase the attractiveness of the brand for investors, stimulate demand for products or services, and maintain the necessary level of reputation in the market.

Example: US tobacco giants have spent millions of dollars placing ads in prestigious publications disguised as editorials. These publications promote the companies as “public health champions” and their new products (such as heated tobacco systems) as “safe alternatives” to smoking, despite the lack of independent evidence. The goal is to legitimize their activities in the eyes of regulators and the public by hijacking the “harm reduction” narrative, while discrediting critics by portraying them as purveyors of “disinformation.” This has allowed the industry to circumvent bans on direct advertising by using the credibility of major media outlets to disguise harmful businesses as progressive initiatives.

The use of pseudoscientific approaches to collecting and processing information can serve as a means of psychological influence on people. Processing the mass consciousness through the support of certain trends shown in research contributes to the formation of the desired behavior, maintaining a sense of the legitimacy of certain views or positions.

Example: The US oil industry (represented by the American Petroleum Institute) has consciously positioned itself as part of the climate solution, promoting natural gas as a “clean” alternative to coal and sponsoring massive advertising campaigns. At the same time, industry leaders have known for decades about the harm of their products, but have hidden the data and spent billions on lobbying and disseminating the necessary information messages in order to delay the adoption of regulatory measures, creating the illusion of environmental responsibility and influencing politicians through “donations” to the campaign, thereby sabotaging real climate initiatives.

In addition to deliberate manipulation of data, false statistics can also be the result of cognitive biases — systematic errors in thinking that cause people to misinterpret information and draw erroneous conclusions, even though they genuinely believe them to be true.

There is a separate satirical website on the Internet that collects allegedly related statistics. For example, it shows a “clear correlation” between the popularity of the name Ariel and the number of fashion designers in Michigan — of course, this is a joke by the authors.

Main goals of falsification

Creating an image of objectivity: One of the main effects of using falsifications in the digital environment is to create the appearance of objectivity and scientific validity of the information provided. Falsified data is presented as official results of authoritative research organizations, inspiring user confidence.

In 1998, a paper by Andrew Wakefield was published in the prestigious medical journal The Lancet. Using just 12 children as a sample, it claimed to show a link between the combined measles, mumps, and rubella (MMR) vaccine and the development of autism and bowel disease. The paper was presented as a serious scientific study. The publication caused mass panic, a sharp drop in vaccination rates, and subsequent measles outbreaks. Only years later, an investigation showed that the data had been falsified. The study was completely retracted from the journal, and Wakefield lost his medical license for breach of professional ethics.

Spread of fake news: Spread of fake statistics has a significant impact on the mass audience. User preferences, behavior, attitudes toward political processes and economic realities depend on information coming from online sources. Therefore, the targeted placement of disinformation can cause mass unrest, panic, and increased distrust of official sources of information.

A video of a 2016 interview is still circulating online, in which naturopath Peter Glyden falsely claimed that chemotherapy is 97% ineffective and that doctors have financial incentives. However, Peter Glidden is not a certified doctor and is known for his commitment to dubious pseudoscientific treatment methods, including homeopathy. At the same time, his statements are refuted not only by real statistics, but also by time itself: the overall mortality rate from cancer was significantly lower in 2016, and since 2005, according to the American Cancer Society, it has decreased by 31% due to improved chemotherapy protocols.

Manipulating audience perception: The Internet provides enormous opportunities for mass consciousness control through the controlled delivery of information. Unscrupulous organizations create entire campaigns built around false research aimed at changing attitudes toward certain issues. This practice is especially dangerous because users receive this information passively, taking it at face value.

Since 2012, a number of environmental NGOs (including Greenpeace) and activist groups have been conducting an aggressive campaign to create an irrational fear of GMO products in the public, especially in Europe, despite the scientific consensus on their safety. They have used frightening images (“Frankenstein food”), references to dubious and disproven studies, and “experts” with alternative opinions have been invited to the media, creating the false impression that there is a 50/50 debate in science about the safety of GMO products. Thus, millions of consumers have developed a persistent negative perception of GMOs, not based on scientific data, but based on successfully imposed fear. As a result, more than 100 Nobel laureates have appealed to Greenpeace with an appeal to stop.

Ways to manipulate numbers and graphs

Artificial creation of trends: Often, researchers try to form trend changes by retouching previous data, artificially increasing the volume of responses, or even creating non-existent values. This method creates the appearance of a stable development of the process, when in reality nothing of the sort is happening.
Violation of the representativeness of the sample: Surveying a limited number of participants or a specific audience allows us to obtain predetermined answers that differ significantly from the real picture of the population’s perception of an event or problem. Thus, the sample is deliberately formed in such a way as to ensure the desired result.
Substitution of terms: One of the simplest methods is to replace one term with another that is close in meaning, but has a completely different meaning. This technique helps to hide the real context of the issue and present the situation in the right light, creating the illusion of agreement of the majority of citizens with a certain position.
Manipulation of visual elements: Presenting numbers in the form of graphical images (graphs, charts) that visually emphasize the desired effect is widely used to mask the unreliability of the figures. With this approach, graphs are constructed in such a way as to emphasize the desired trends, using various techniques of exaggeration or concealment of scale.
Incorrect interpretation of final data: Often, the data obtained during the study are subject to arbitrary interpretation by the researchers. Such an incorrect explanation of the results allows the customer to reach the necessary conclusions regardless of the actual state of affairs.

Consequences

The consequences of using falsified data are very diverse and concern various areas of society.

Disorientation of the public. People are deceived, losing the ability to correctly assess what is happening and navigate events.
Reduction in the authority of scientific institutions. The publication of incorrect or falsified data reduces the level of trust in scientific organizations, calling into question the reliability of all the research they conduct.
Formation of false ideas about reality. Having become accustomed to perceiving information exclusively through the prism of official publications, citizens stop critically analyzing incoming data, preferring to trust the appearance of evidence rather than its essence.
Making unfounded management decisions. If the country’s leadership relies on falsified information, it risks making wrong decisions that can harm the economy, politics, or social development of the state.

Manipulation of numbers is not just deceit, it is a tool for managing reality. These are real decisions made by people deceived by false data: refusal to vaccinate, belief in “safe” harmful products. This is an erosion of trust in all institutions that are supposed to rely on facts. To protect against such manipulation, it is necessary not only to expose individual fakes, but also to systematically improve media literacy and transparency of research, so that data again serves people, and not those who forge them.