Cognitive selection

For the time being, communication technologies, like social media, enhance the consumption and producing of information. In his paper, ‘The Dark Side of Information Proliferation’, Thomas Hills highlights that within information-rich environments, information is placed under forces of cognitive selection, which may produce negative outcomes, such as extremism or hysteria. To test Hills’ ideas on negative and belief-consistent selection of information, I decided to conduct research on how people choose textual articles and what sort of information they tend to learn from them.

According to Hills, we are prone to weight advantages over disadvantage. Therefore, negative information is more likely to be selected than balanced content. When we think about this fact, it seems sensible: negative information has more significance for survival. Indeed, you probably won’t doubt that for ancient people it was more important to know which berries are poisonous than how plants propagate. Despite the fact that in most modern societies, life is comfortable and safe, we are still prone to negative selection because it’s part of our evolutionary heritage.

Moreover, in the ocean of information within the internet, we find a sea which is similar to what we know and which confirms our existing beliefs. Hills explains belief-consistent selection by our propensity to defend the knowledge which make up our worldview and shape our identity. More than that, due to our biology, we understand and recall information in relation to causality structures that we already understand more efficiently.

To test Hills claims about these tendencies in information selection, I’ve conducted a small study. I’ve created a Google Form with several short textual articles. On the first page, participants should have read nine headings and opt for one of them. Respondents have been given four pairs of the same texts which differ from each other only in headings. While one article has got a negative heading, another one has been entitled in a more positive way. The ninth article’s heading has been made neutral. According to the first hypothesis, the majority of participants should have opted for articles with negative heading which would have confirmed negative selection tendency.

After reading, respondents should have answered on the question: “Which option does describe the meaning of this text best?” As you remember, the texts in pairs were absolutely the same and so were the offered options. Among four options, a half contained only negative information from the text while the rest consisted of positive ideas. All options had equal chances to be chosen because ideas were represented in controversial texts approximately in the same way. However, I expected to see the correlation between headings of articles (except for the neutral article) and answers. This would have supported the idea of belief-consistent selection.

• 

The results confirmed both hypotheses. People opted for articles with negative headings more often. 25 participants (approximately 57 percent) chose negative headings while 15 people (approximately 34 per cent) read articles with positive ones. The rest was attracted by the neutral article. Also, there was a correlation between what participants expected to see in a text and what they interpreted as its main idea. For instance, the text devoted to privacy concerns about the use of Amazon Echo equally introduce both perspectives on the problem. But those who read “Amazon Echo could spy on users” decided that text criticized Echo while people who chose “Amazon Echo uses confidential information only with the knowledge of users and in their interests” thought that the article defended the gadget.

References

Thomas T. Hills , “The Dark Side of Information Proliferation”, Perspectives on Psychological Science 1 –8, 2018