At first sight, the title of this article might seem a bit strange. After all, Meriam-Websters dictionary defines fact as:

- something that truly exists or happens: something that has actual existence
- a true piece of information

while the Oxfords English Dictionary says:

A thing that is known or proved to be true.
- Information used as evidence or as part of a report or news article.
- Used to refer to a particular situation under discussion.
- Law [mass noun]The truth about events as opposed to interpretation.

The way in which these honorable bastions of linguistic knowledge phrase their definitions varies but the intrinsic meaning is the same. Facts are true pieces of information and are commonly used as evidence during a discussion or argument.

As a silly example I might present the fact: “I like the color blue” as justification for wearing a specific blue shirt.

On the other hand, opinions are almost the opposite of facts:

A view or judgement formed about something, not necessarily based on fact or knowledge.

according to the OED and, in the words of MWD:

A belief, judgment, or way of thinking about something : what someone thinks about a particular thing

By now you’re probably thinking I’ve gone off the deep end. How can Opinions be Facts if they are (apparently) opposites? The answer is simple and lies at the heart of how facts are used to backup a given argument or opinion. You choose the facts that backup the argument you are trying to make but

the facts you choose to present and how you choose to present them reflect the argument you are trying to make.

This is true in a wide range of contexts, ranging from coffee shop discussions to scientific articles and, of course, political speech. This is perhaps clearer in the case of data visualization, so let’s take a look at a quick example:

https://wattsupwiththat.com/2011/04/05/global-temperature-still-headed-down-uah-negative-territory/

This (now old) plot of Global Temperature Anomaly shows how the Global Temperature has fluctuated around the mean from 1979 to 2011. Looking at the trend, it is clear that the temperatures have been increasing. However, if the author had shown the values only from 1998 until 2011 it would show that the temperatures have been decreasing. Indeed, this period is commonly used to show that there is no global warming (see http://www.skepticalscience.com/global-warming-stopped-in-1998.htm for a longer discussion). It is a fact that temperatures after 1998 were lower for a long period of time. It is also a fact that 1998 was a particularly warm year, what one would call an outlier in statistical parlance. It is also a fact that global temperatures have steadily increased (with fluctuations) from 1979 and 2011. If you choose your facts carefully you can use them to justify any position you want.

This example is particularly obvious and with clear intent, but there are many contexts where it is not so clear and where even if you have the best of intentions you may fall into this trap. A classical example of this became famous as the “UC Berkley gender bias case” where a simplistic analysis of the admissions data seemed to indicate a clear case of gender bias in university admissions that could potentially result in a law suit to be brought against the university. Only when the data was further analyzed, in a paper published in Science magazine, did it become clear that this was just a case of “Simpson’s Paradox”. This kind of statistical paradoxes and unintentional mistakes in data analysis will only become more likely as dataset size increase thanks to the so called “Big Data” revolution.

Another situation where this became clear to me was during my stint teaching Physics at Aix-Marseille Université in Marseille, France. While preparing the classes I had to make many decisions regarding how to present the material to the students. I could use this example or that one, use analogies and visualizations or stick to the purity of the mathematical expressions. Every example or analogy would (hopefully) help to make a specific point clear, but it also had its own limitations. Over time, I noticed that some analogies worked better than others. In some cases, an example or analogy I used, that seemed completely clear to me, caused only confusion or misunderstanding despite being factually correct.

In summary, the way you present your arguments and the facts and examples you use make a difference and may reflect an internal and perhaps unacknowledged bias or even an explicit intention to deceive. Keeping in mind that “Facts are opinions too” has helped me try to look at my arguments and analyses from different perspectives and to try to poke holes in them. It has also helped me to pay more attention to what I’m reading or hearing and trying to find what opinions and biases might lie under the facts that are presented. Hopefully you’ll find it useful as well.

Data Science, Machine Learning, Human Behavior

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store