Episode 10

(00:00) Big Data Gets Bigger

If you took a piece of letter-sized printer paper, and folded it in half repeatedly, it would take 7 folds to be as thick as an average notebook.

How tall do you think it would be if you folded it in half 50 times? The empire state building? Mount Everest? Big Bird? Try the distance to the SUN. It sounds crazy but it’s true.

Exponential growth is a hell of a drug. Don’t believe me?

Modern technology means we’re able to collect data about literally everything. Big Data is so big we now have to measure it in Zettabytes. What’s a Zettabyte? Well if you’ve ever used an external hard drive, those can usually hold about 1-5 terabytes. Which is quite a bit...A Zettabyte is A BILLION terabytes.

This number has become so outrageous for the same reason folding paper can take you 93 million miles into space. It doubles about every two years.

Okay so long way of saying...today’s show is about making paper airplanes and flying them to the sun. No unfortunately that’s not true, maybe next season.

The amount of DATA that exists in the world is beyond our comprehension. Data streams have turned into rivers, lakes, and now oceans that can’t be measured because we’ve never been to the bottom. We quite literally have the grand sum of humanity’s knowledge at our fingertips. Surely that knowledge is making us smarter, right? Surely, we’ve gotten better at feeding that information through our carefully calibrated, objective, cool-headed craniums, right?

Of course not….Where’s the fun in that?

There’s so much to unpack here, this topic will be a two parter that we’re calling Data vs Goliath.

Our question for part 1: Do facts really change our minds?

Playing the role of Goliath: Human nature.

My Big Data is bigger than your Big Data. I’m Scott Hermes, this is Working Better.

Are we as data-driven as we think?

There’s no problem data can’t help solve right?. Worried you’re spending too much time staring at your phone? Bam! Track your screen time down to the second. Feel like the thing that’s really missing from your life is a detailed report of your blood oxygen levels? Bam! Get an Apple watch. Want the government to track where you are at every given moment? Bam! Get a cell phone.

Whether it’s finding the best deal on a flight, where to live, what to eat, what car to drive, where to work, what to buy, how much to spend, your likelihood of getting COVID when you go out to get that burrito grande, there’s not a decision we encounter that we can’t take a deep breath and say “well let’s look at the data.”

But if we all made data-driven decisions, bacon wouldn’t exist, we’d all lock our phones in a safe while we sleep, and I would stop betting on the Bears to win the Super Bowl. But 500 to 1 to win next year. I’d be crazy not to bet on them...

(03:16) Confirmation Bias

One reason is due to what’s called confirmation bias. Confirmation bias means we seek out and favor information that confirms what we already believe. 

One of the most famous studies of confirmation bias came from researchers at Stanford University in 1979. Participants were divided into two groups: one group supported the death penalty as a deterrent to crime. The other opposed it, believing that the death penalty had no effect on crime. Each group was then presented with two studies. The first study confirmed what group A believed - that the death penalty deterred crime, supported with big, juicy data points. The second study confirmed the opposite, with it’s own equally compelling data. Participants then rated the credibility of the data.

Both studies were entirely made up, and both had the same effect. Those who were pro-death penalty rated pro-death penalty crime deterrence data as highly credible

Those who opposed the death penalty said that that data was bogus, and that data saying crime wasn’t affected by the death penalty was all but conclusive.

We favor the data that supports what we already believe. You see it all the time. At work, people will worship data that supports their argument, but will turn into Jack Bauer and interrogate the HELL out of information they don’t like.

Confirmation bias appears to be deeply ingrained in us. Even if we are made aware of its existence, we still tend to do it. The theory is that it allows us to rapidly process information and quickly avoid danger. As humans evolved, it saved us from having to think in life-threatening situations.

Cognitive scientists Hugo Mercier and Dan Sperber also point to the fact that in hunter-gatherer groups, social standing was the most important survival mechanism. So when it came to convincing others that they should go hunting for saber tooth tigers while you reorganize the berries at home...winning an argument was more important than being accurate.

(05:45) Pushing for Data Literacy

The human mind is wired to put up a fight against facts. It’s our nature. But it also used to be in our nature to communicate only via spoken word. For almost all of our existence as a species, around 200,000 years, no one wrote. For thousands of years, reading and writing was a relatively niche skill set of the highly educated.

So is the problem that we don’t yet have the tools and understanding to harness the data that surrounds us?

The discipline of DATA SCIENCE has exploded so rapidly, it’s easy to forget that it’s still a relatively novel idea in mainstream society.

Don’t believe me? We tracked the Google search volume for “Data Science” and “Fax Machines.” In 2020, “data science” was 36x MORE likely to be searched than “fax machines”. When do you think was the last year that Fax machines were on top? The last year searches for FAX MACHINES were more popular than DATA SCIENCE? (pause)

Anyone. (pause) Come on. (pause). Take a guess. (pause). I can wait all day. (pause). Oh right. This isn’t live.

2009.

Barack Obama had just begun his first term as President and fax machines were as relevant a business asset as data science? We’ve come a long way, but the growing pains are real. In our race to inject scientific thinking into virtually every facet of life, core principles of GOOD science can fall by the wayside.

So what - we’ll all have to define what a regression analysis is in order to renew our driver’s license? That we’ll finally start flossing twice a day once we understand how to calculate when gingivitis becomes statistically significant?

Not exactly. Some of the push for “data literacy” is simple good citizen, good consumer sort of stuff. Check your sources, consider the broader context a data point might belong to, don’t get all your news from an Instagram account called The Tinder Blog. You know, the basics.

It’s easy to take data at face value. Asking questions to dig a little deeper is harder. Take the debate about genetically modified food. The Oklahoma State University Department of Agriculture Economics asked consumers whether foods made with genetically modified ingredients should be labeled as such. 80% said yes. Seems like a good rationale for a law then right?

BUT...80% of those same consumers also said that labels should explicitly indicate if food contains DNA.

For those keeping score at home, yes you heard that right. And yes all living things have DNA. So if we avoided food with DNA, we’d….you know, die.

The first number on its own seems to make a compelling case. But a little more context reveals the data might be skewing towards people who have incomplete knowledge or are just terrified of acronyms. Now that doesn’t mean that it is a bad idea to label food as GMO BUT you can’t use that data point to prove it is a good idea.

(09:02) Bridging the Literacy Gap

KARL:

We all need to think about how we engage with data better.

That’s Karl Hampson CTO of AI & Data from Kin + Carta Europe. Karl says improving our relationship with data and facts means being naturally skeptical and curious about the information we encounter:

KARL: 

"If I handed you a bottle of water and the cap was open, you're going to question that. Why don't people do that with data and information?

Other research suggests that the less we know about a topic, the stronger our biases actually tend to be.

In 2012, the US Supreme Court ruled to uphold key provisions of the Affordable Care Act, aka Obamacare. After the ruling, the Pew Research Center conducted a survey to gauge public reaction. 36% were in favor of the ruling, 40% opposed it, and 24% expressed no opinion. Then they were asked what the actual ruling was. Only 55% were able to. So 76% had no problem giving an opinion, but only 55% actually understood the question.

One school of thought is certainly a logical one: Increase the actual understanding of a topic, and those biases are less likely to get in the way. It’s often referred to as the “deficit model.” So in theory, if people spent less time being outraged by the news around an issue, and more time understanding the issue itself, arguing with your Aunt on facebook might be replaced by...arguing with your aunt on the phone. Baby steps.

Dr. Jenny Rankin is an award winning educator, author and data scientist. She says we should work towards “over the counter data.” That data should be something that can be safely and properly used without an expert. We asked Karl - is that a good idea?

KARL:

I love the idea of trust and the idea of packaging data in a way that you can take it, consume it for a particular purpose. It's documented. All that good stuff for me is actually productizing data. That's really the bigger idea, I think, to take away.

I'm handing you a bottle of water, you trust it in the same way, because you look at it immediately, interpret it. It's from a recognized brand. The integrity is there.

Imagine thinking the same thing with data, and therefore, you can take that and build upon it, do the same, productize it, add more value, and put that back into the product ecosystem.

Make this data, these GOBS of information, easier to access and understand, and we might begin to close the gap here. Karl underscored what side of the equation needs the most help.

KARL:

So we don't need more technology really at the moment with data. What we have is a people issue.

(11:53)The Power of Culture

“What we have is a people issue” sounds like planet Earth’s mechanic explaining what’s making all that racket. It’s us. We’re the problem. (sigh) Again. Because while we absolutely can improve our ability to interpret, interrogate and understand raw data - there are still bigger forces at play. Another reason facts and data have a hard time changing our beliefs? That’s not how they’re formed in the first place.

In their book, the Knowledge Illusion, authors Steven Sloman and Philip Fernbach argue that our beliefs are forged through powerful cultural and contextual factors that make them nearly impossible to change.

With that in mind, we’d like to go on record here at Working Better, to not shy from taking a stance on a controversial topic, with a bold belief of our own. And though we may lose a few subscribers and sponsors so be it. (clear throat): The Earth is round. There I said it.

When we’re talking about how groupthink can beat well established facts, the flat earth movement is a topic that’s hard to avoid. Even the most conservative estimates measure the growing number of “flat earthers” in the millions, just in the US alone. There’s lots of ways to explain the phenomenon, but undoubtedly one is that people discover a sense of community. They attend conferences and gatherings around the world, they forge friendships, and find identity and meaning in the movement, in the pursuit of what they believe is the truth...and that becomes strong armor against any evidence that says otherwise.

Here’s a clip from a National Geographic Documentary from 2019:

FROM CLIP:

“Your belief in the earth being flat flies in the face of hundreds of years of scientific evidence that the world is round. But not only that - we have satellite imagery, photos from space that prove that the earth is round…."

“Right and nobody here believes any of that anymore.” (Nat Geo)

We ALL can be guilty of this type of thinking in one way or another. We instinctively ignore or discredit data that threatens a part of our identity. A chef is much more likely than an average person to be skeptical over pizza-making robots. It’s not that Blockbuster didn’t have data indicating more and more people were consuming media online and ditching their DVDs...the facts threatened the core of who they were, so they were effectively ignored.

Someone wiser than I once said, “It’s hard to get someone to understand something when their salary depends on them not understanding it.”

(14:38) Interview with Jeremy Hoffman

With that in mind, I’m thrilled to be joined by someone whose job DOES depend on them understanding data. Dr. Jeremy Hoffman is the Chief Scientist at the Science Museum of Virginia and a professor at the L. Douglas Wilder School of Government and Public Affairs at Virginia Commonwealth University.

Jeremy specializes in Earth science communication, data-driven and community based participatory science, and science center exhibit content development. He’s been highlighted in the Grist50 and has been written about in publications like the New York Times, NPR, STEM jobs Magazine, UPWORTHY, and yes folks, the Working Better podcast. Jeremy talked with us back in Episode 5 about how cities are affected by climate change. We’re thrilled to have him back to talk about data and the importance of science-based education

SCOTT:

Can you tell us a little bit about what you've been working on at the science museum?

JEREMY:

The Science Museum of Virginia likes to think of itself as the marketing agency for science. Everything that we're doing these days is trying to center the experience around COVID-19 and the ongoing pandemic. From last year, focusing on the most breaking science, most reliable science we could share about things from masking to the change from surface... Focusing on cleansing surfaces to maintaining distance and wearing a mask. And then now it's shifted into reinforcing the safety and reliability of vaccines to alleviate ongoing pandemics. So that's kind of been coloring the whole background of work these days.

The science museum is dedicated to communicating climate change and its impacts on the Commonwealth of Virginia for the last several years. Working on explaining how something as big and seemingly far away in space and time as climate change is in our backyards. How does something so seemingly so far away in space and time impact me in my day-to-day life? And then finally, just broad brush, we're trying to integrate real-world science all the time into every aspect of the institution, from our exhibits to our social media content, to our external communications. So it really is a job that allows me to live, breathe, and explain science throughout the week and the years.

SCOTT:

I think that as a consumer of that data, that information coming from the scientists that is either frustrating or hard, is that the research is happening in real time. Right? The message keeps changing. And how do you decide how to collate that information in such a way that you're not seeming to spin 180 every other day?

JEREMY:

The best and most clearest example of how reliable scientific information has needed to be updated as we discovered new things was our experience at the beginning of the pandemic to where we are now. And what do I mean by that? Originally it was continue to wash your hands religiously between every single touch of any surface. Which is good practice, that's good public health practice to begin with. But then things like making sure that you're cleaning every single surface. It's become clear through laboratory experiments and observations and studies based on how COVID-19 spreads among people is that that's not a really viable way to catch COVID-19. It's much more about avoiding cramped, crowded, and poorly ventilated spaces.

SCOTT:

Can you talk about some examples that you can give where you feel like you've been successful in taking something which could be maybe hard to grasp and making it really more impactful to people?

JEREMY:

What we started to do in my early times as a science communicator was trying to figure out how we get people to remember the simple things around risk mitigation should an earthquake exist or happen?We started doing flash mobs around earthquake science communication, and we called it flashmob science because what we would do is recruit a large group of people to pretend as though an earthquake was going on.

SCOTT:

What do you feel, you think changing data literacy or science literacy could help maybe people understand better the severity or the urgency behind climate change?

JEREMY:

I think that data can be very, very strong for particular people. Whereas on that other side of the spectrum where people's values don't align with the sorts of things that need to happen in order to address the climate crisis, data doesn't matter anymore.

JEREMY:

It's taken the climate science communication field a long time to start to recognize that it truly is about connecting with communities through trusted voices. Who are you hearing about this information from? Is it someone that looks like you? Is it someone that engages in the same activities as you? I imagine someone like a sports caster or something on Monday Night Football talking about the heat exhaustion that players of the future might encounter because of climate change. That might turn some heads. As well as something like doctors having conversations with their patients about how a certain climate stresser disproportionately affects them because of where they live in a city. I think that those are the sorts of trusted voices, both in public eye as well as professional life that can be having these conversations and helping to move the needle on public awareness of the climate change impacts that will affect them in their day-to-day lives. Without that it truly is in large part just kind of extra data until you start to connect it to someone's backyard or their front porch.

SCOTT:

Do you feel like, as a scientist, that's made you more aware of your own biases or able to sort of realize, at least in retrospect, that you may have perhaps brought a bias to the table?

JEREMY:

I certainly think that understanding your own personal biases is a very introspective piece of work and scientists are not trained any differently than the rest of us in how to be introspective. The scientific mind frame provides you with is the ability to understand progress. And again, that changing information through time, being comfortable with changing your worldview based on that kind of new data being incorporated into your understanding. It's also being experimental and figuring out what works and what doesn't under certain situations. And how do I start to incorporate this information into my day-to-day life? So while I don't think scientists are any more prepared to investigate their own biases than others, I do think that we bring, and we have a certain training in incorporating that information into our day to day lives to produce a more positive outcome.

Of course everyone has blind spots, things that even if we do as much introspection as possible, we're never going to uncover them without conversations with others and seeking out opinions and understandings of the world that are different from ours. So introspection is the first step and then being able to hear and listen to other's experiences and how that relates to your own understanding of the world is also as important as doing the own introspection to identify your own biases.

SCOTT:

Is there anything else you want to add before we go?

JEREMY:

As of April, 2020, the FDA lists 85 different vaccines that are licensed for use in the United States for various diseases. Doctors currently recommend 16 of these by your 18th birthday. And to put that into perspective, there are 237 vaccines that are in some level of development for COVID-19 alone, according to the World Health Organization. And I think that that kind of scaling, it really identifies the magnitude of the scientific endeavor that's going into identifying safe and effective vaccines for this illness.

Conclusion

Huge thank you to Jeremy again for a great conversation, and to Karl Hampson for talking with us.

So. What have we learned today

There is an overwhelming amount of information out there.

It is rarely presented to us in an easy to digest format.

When it is, we are terrible at understanding numbers and probability

Furthermore, we only pay attention to those facts that support what we believe in

We are all going to die

Learning how to better understand and utilize raw information starts with better understanding our own innate biases. It can make us better collaborators when we can recognize why our coworkers seem to be clinging to one piece of data. When our instinct is to discredit information that doesn’t support our point of view, we can be more aware that our old friend confirmation bias might be poppin in to say hello.

Our complicated relationship with the truth isn’t new by any means.

But clearly the internet is magnifying those biases. As the rate of new information intensifies, so does our hunger for it. So we jump quickly from article to article, video to video, which means stopping to fact check becomes even less likely.

AND even when we try, the technology of disinformation is making it more difficult to separate fact from fiction in the first place. Which brings us to part two of Data vs Goliath. Listen next week as we ask: Has technology broken the truth?

If you liked the show, please remember to subscribe. Reach out to us on Facebook,Twitter, Instagram, or LinkedIn to give us more data about how we are doing so I can cherry-pick the information that reinforces my belief that I am CRUSHING it. If you don’t believe that social media exists, you can always send us a message by giving to one of the friendly mole people who live inside our hollow earth. They are slow, they don’t say much, but they can always be trusted. See you next episode!