Icon (Close Menu)

Logout

When Facebook Inhibits Knowledge (Sarah Quintanar Commentary)

3 min read

In a given day, Google receives more than 5 million search queries and Facebook users share more than 4.75 billion pieces of content, according to annual search statistics.

Retailers and corporations use complex data mining systems to collect information about their consumers, information that ranges from products they view to products they actually purchase, as well as what their social networks’ interests are. Both data and research have become easily accessible in today’s world of “big data” and analytics.

As an empirical economist, I love data. Detailed information to quantify a complicated question and the ability to propose a potential solution supported by hard facts and statistics — what could be more exciting?

However, given the massive amounts of data available in today’s world, often, regardless of whether facts or conclusions are correct, they are massively published because of the effortless ability to spread information in a split second.

One of my goals is for students to be able to identify reliable research claims, and conversely, to be able to explain why some are unreliable. Increasingly, I’ve been able to find examples for my classes with zero effort, just from browsing my Facebook feed: generally inflammatory headlines followed by conclusions based on limited samples or no reliable study at all. So, why is this a problem, and how do we correct it?

This massive influx of information is certainly useful, but problems can arise because it is relatively costless to spread information using social media, emails and even text messages. Likely the idea of verifying facts in this instance is not even a consideration: Assumedly, it is being spread for knowledge or entertainment with no perceived harm or associated risk.

However, this becomes an issue if businesses or consumers make decisions based on inaccurate or nonexistent data. There is a loss of efficiency with these decisions, sometimes going so far as to influence policy or social welfare in conjunction with products and entertainment.

A current example can be found in the inflammatory statements perpetuated by a blogger who goes by the pseudonym “Food Babe” and who has been able to create an empire of followers on issues related to food safety without credentials or reliable sources on which to base her claims.

Nonetheless, pressure from followers has caused a number of businesses to alter their products, despite outright refutation of her claims as false from the broader scientific community.

This is an extreme example, but one that very well illustrates the influence of demand in a market. The loss of efficiency I mentioned previously lies in the impact of these changes: likely increased costs for the products and businesses involved, which increase prices for consumers and results in little to no health benefits insofar as the scientific community is concerned.

This is not meant to single out Food Babe, because many other examples abound, even those that are carefully executed and peer-reviewed scholarly works. However, the results are not as generalizable as the headlines imply, and herein lies the problem.

Internal validity implies the researcher has conducted a reliable study where results hold within the study itself, while external validity is generally what the public as a whole is interested in: Can these results apply outside of the study itself, and what are the broader implications?

For external validity to hold, the researcher(s) must have a large, random, representative sample or very specific data for the question at hand, and those results should be easily replicated. For example, using findings from a study on elderly women to impart a dietary change for college students is a poor application of data.

When we are making decisions and using the massive amounts of available information, there has to be some proven method of assessment and valuation so that we aren’t just following a crowd over a cliff. In today’s world of big data, and nearly limitless possibilities, it is increasingly vital to have the ability to verify accurate data and ensure that it truly represents what it claims to.

Sarah Quintanar is an assistant professor of economics at the College of Business at the University of Arkansas at Little Rock. Email her at SMQuintanar@UALR.edu.

Send this to a friend