October 2021

Decision making in digital life

Good decision making has become crucial in the modern era. We are exposed to a great deal of data every day and filtering, digestion and remembering the useful information are becoming more challenging. Being aware of your intuitive bias and relying more on facts than your emotions are absolutely necessary to make a good decision. 

This essay discusses how we are influenced by the digital world when we make decisions. We are constantly exposed to various data in the digital world, which we need to process, degist and remember. Our decision mechanism on what we eat, who we vote and what we dress is heavily affected by the data we are imposed everyday. The Web has become the leading platform to publish random data without a factful filtering. Disinformation, misinformation, which are intentionally or unintentionally generated erroneous information are different public enemies of decision making. The question is if we can trust our intuitive decisions we make or should we give more importance to factful factors.

Important decisions

The judicial system is designed to be neutral and not biased against any race or gender. However, even judges’ decisions can vary based on the lunch they ate [1]. The research published in 2011 about the bias in judicial decisions demonstrated that judges’ decisions can vary up to ≈65% by what they ate for breakfast. This research naturally raises a question of credibility of decisions even at the highest levels of institutions. 

In 2011 among 500 respondents; 18 years and older; randomly selected from a national internet panel are asked which of the following influences your final decision when voting for/about political candidates and issues and 64% of the respondents said they are influenced by the television while 40% is influenced by the internet. The result gives a hint about how we are influenced by the programs and data sources we follow for making an important decision. [2]

Figure 1. The result of the survey done in 2011. Which of the following influences your final decision when voting for/about political candidates and issues?

Nigeria had cognitive biases against Western Polio vaccine a decade ago because of historical and cultural reasons [3]. The Nigerian religious community did not want to get any vaccine from the Western world believing that it is just another trick of a Western countries to damage Nigerian people. After a great deal of effort, the problem was solved and the anti-vaccine community was engaged in the Polio eradication initiative. The efficient use of the digital world could have prevented the death of children in the early stages of the Polio epidemic.

Fact check

Hans Rosling, the writer of the book called The Factfulness, claimed that chimpanzees score better than most educated people in a knowledge test [4]. In 2017, they asked nearly 12 000 people in 14 different countries to take a 13-question multiple-choice quiz. They included various segments of people in the experiment: medical students, teachers, university lecturers, scientists, investment bankers, journalists and senior political decision makers. Each question offered one of three choices. However, most of the respondents, even Nobel laureates and medical researchers, gave wrong answers to the questions. The results indicated that most people get the world dramatically wrong despite their high intelligence and intellectual levels. Here are some of the questions and the corresponding answers:

1. In the last 20 years, the proportion of the world population living in extreme poverty has…

A: almost doubled

B: remained more or less the same

C: almost halved

2. What is the life expectancy of the world today?

A: 50 years

B: 60 years

C: 70 years

3. There are 2 billion children in the world today, aged 0 to 15 years old. How many children will there be in the year 2100, according to the United Nations?

A: 4 billion

B: 3 billion

C: 2 billion

4. The UN predicts that by 2100 the world population will have increased by another 4 billion people. What is the main reason?

A: There will be more children (age below 15)

B: There will be more adults (age 15 to 74)

C: There will be more very old people (age 75 and older)

5. How did the number of deaths per year from natural disasters change over the last hundred years?

A: More than doubled

B: Remained about the same

C: Decreased to less than half

The correct answers:

1:C,2:C,3:C,4:B,5:C

The percentage who answered the first question correctly is 25% and Sweden shares the first place with Norway with this score. Only 2% of people in Hungary answered this question correctly.

How would chimpanzees score? If you throw a bunch of bananas to a group of chimpanzees and each banana is marked with either A, B, or C that correspond to their multiple-choice options. Then every time chimpanzees eat a banana, they mark down a letter on the banana a chimpanzee chooses to eat. If, ever, some researchers conducted this experiment thousands of times, the chimps would score about 33 percent correct on each three-answer question, meaning their results would be random. Humans’ score worse than chimpanzees because of the intuitive bias influenced by the media, the internet or people around. Because disaster stories are more interesting, we pay more attention to an airplane crash, a murder, a celebirty interview or a political scandal. Digital media is diverse and obviously more independent after the epic social media disruption. However, we still perceive the data selectively and dramatically. 

Figure 2. Illustration image for the dramatic attention filter. 

Data and decision

An individual has been described by a neighbor as follows: “Steve is very shy and withdrawn, invariably helpful but with little interest in people or in the world of reality.  A meek and tidy soul, he has a need for order and structure, and a passion for detail.” Is Steve more likely to be a librarian or a farmer?

The association between Steve’s character and stereotypical librarian hits everybody without any doubt. Nevertheless, the number of farmers in the U.S is about 3.2 million [5] and the number of librarians is around 170 thousand [6]. Therefore, statistically speaking, Steve is more likely to be a farmer than a librarian. In an experiment, the participants disregarded the statistics and completely went with their stereotypical views by answering the question above as Steve being librarian. The researchers proposed that their reliance on heuristic (mental shortcut which uses emotion to influence the decision [7])  caused predictable biases (systematic errors) in their answers [8].

Importance of data collection

Data collection for creating AI models is crucial because it affects the accuracy of decisions AI makes. For example, if you are building an AI model that decides if an object in a picture is a dog or something else, firstly you need to train the model with proper dog pictures. After the training, the software can decide if a given picture is a dog picture or not. The better data set is used to create the model, the better accuracy the software annotates dog pictures. Therefore, the data set collection and preparation directly affect the accuracy of the decisions of AI software. 

How do we prepare a good dataset to train the AI to make better decisions? We filter out irrelevant data from the data chunk and keep the dataset clean. It is like the physiology of the human body, the nutrition intake is hugely important and affects the entire body functionality. Naturally, we can use the same analogy for data and decision mechanisms. The more we feed our brain with factful and healthy information, the better decisions we can make. 

Filtering out irrelevant information

Contrast with “misinformation,” which refers to misleading people unintentionally.  The media companies have attempted to address this issue. Information overload from the Internet will make it increasingly difficult to separate fact from fiction. Speaking of misinformation, Twitter announced  a pilot program called Birdwatch, a community-based approach to fight against misinformation [9]. The community-driven approach allows people to give feedback about the Tweets they think are misinformative. Users can write notes that provide informative context to prevent spreading disinformation. 

Conclusion

The bias of intuition can drastically vary based on the digital data we process throughout our lifetime. Do not trust fully intuitive decisions, try to take the facts into account while making a decision. People who are privileged or conscious about accessing the factful data and make their decisions accordingly, shouldn’t blame the ones who do not have luxury or environment for factful data but help them to find the simplified factful data. Every human has the right to access facts and should demand to access factful data.

References

1. Extraneous factors in judicial decisions Shai Danziger, Jonathan Levav, Liora Avnaim-Pesso. Proceedings of the National Academy of Sciences Apr 2011, 108 (17) 6889-6892; DOI: 10.1073/pnas.1018033108

2. Media influencing final decisions about politics in the U.S. in 2011. Statista Research Department. https://www.statista.com/statistics/223191/media-influencing-final-decisions-about-politics-in-the-us

3. Nasir SG, Aliyu G, Ya’u I, Gadanya M, Mohammad M, et al. (2014) From Intense Rejection to Advocacy: How Muslim Clerics Were Engaged in a Polio Eradication Initiative in Northern Nigeria. PLOS Medicine 11(8): e1001687. https://doi.org/10.1371/journal.pmed.1001687

4. Rosling H., Rosling O. (2020)., Rosling Rönnlund A. Factfulness: Ten Reasons We’re Wrong About the World–and Why Things Are Better Than You Think

5. United States Department of Agriculture National Agricultural Statistics Service (ACH12-3/May 2014). Farm Demographics U.S. Farmers by Gender, Age, Race, Ethnicity, and More. U.S. Department of Health and Human Services, https://www.nass.usda.gov/Publications/Highlights/2014/Farm_Demographics/Highlights_Farm_Demographics.pdf

6. Number Employed in Libraries. ALA Library Fact Sheet 2 [Fact sheet]. American Library 

Association. https://www.ala.org/tools/libfactsheets/alalibraryfactsheet02

7. Heuristic. In Wikipedia. https://en.wikipedia.org/wiki/Heuristic

8. Daniel Kahneman (2011). Thinking, Fast and Slow.

9. Keith Coleman. Introducing Birdwatch, a community-based approach to misinformation. https://blog.twitter.com/en_us/topics/product/2021/introducing-birdwatch-a-community-based-approach-to-misinformation