Real Analysts Know Where Their Data Comes From

This is the final straw… well maybe… let’s ask a 9 year old… wait, what?

Decision-First AI
Course Studies
Published in
6 min readJan 30, 2018

--

The United States uses 500 Million straws every day! What an astounding statement of waste! In the last few days it has been widely reported by every major news organization you can think of… but only one, that I am aware of, asked where that number came from.

Well, that’s not completely fair, it came from the National Park Service — that mecca of analytic research that they are. I guess that didn’t raise any flags? Only the NPS didn’t research that number. They were provided it.

More on Milo here.

Milo Cress, according to what is now popular journalism, is the leading expert on straw usage in America. A picture of Milo might be nice, but he is only 16. He is an amazingly driven and inspired young man. But his expertise on straw usage comes from a handful of phone calls he made in 2011 (at age 9) to a few straw manufactures… and then he guessed.

That’s right. The US is using 500 million straws per day based on the guess of 9 year old! Now that is great journalism!

It is certainly not great analytics. And to be clear — I am not faulting Milo. He is very transparent about it all. He did his best to set an assumption. Milo rocks! Just not at analytics… but he has plenty of time for that when he is done saving the world.

Analyst know thy data!

Milo setting an assumption is fine. Journalist reporting that number as if it was anything other than the guess of an amazing (but unscientific) 9 year old is … should be — unfathomable. The Democratic majority leader of the California lower house using it to impose a $1000 dollar fine on restaurants is ludicrous… expected… unenforceable… typical politics. Can we give Milo that guys job?

Are straws really a big problem?

Perhaps. They do seem rather wasteful. At California’s Coastal Clean-Up day, they are reported (along with stirrers) as the #6 leading source of debris. With 736,595 being tallied over a 25 year period. This begs a lot of great analytic questions that so far no one is asking. Like:

  • how accurate is any of this counting
  • what defines a “stirrer” many are thin sticks and straight pieces of plastic
  • what is the ratio of straws to stirrers
  • if you find a bottle with a straw in it, how is that counted
  • what sort of sample is this any way? Guiness touts it as the largest garbage collection. So that is something, but hardly scientific.

There are plenty more questions. There always are. But, these are all things an analyst should consider. To do that, you have to understand where your data came from. It has a lineage. It has a process. Unfortunately, it does not have a blockchain — but you can reverse engineer one.

You can also use techniques to vet even the most speculative data. It isn’t that hard… a journalist could do it! Ouch.

Triangulation — Because Data Is Often Speculative

So now I (well, I should credit Reason) have uncovered two sources of data for which we have some sense of lineage and process. What do they tell us? Let’s triangulate… yes, we are one source shy — be patient.

500 M straws per Day in the US vs 737 K “straws” per 25 years in California

Which translates to:

500 M in the US vs 80 in California

All on a straws per day basis. Or a rate of -

0.000016%

Sorry — is that waste or a rounding error. Let’s adjust for the difference in sample area using the percentage of coastal population 39% and California’s share of coast line 6.7%. Google it.

0.0006% or 6 in every 1 Million straws used (likely produced given our context on the data)

Sorry Milo, representative Calderon, you would have a better case with a much lower number of straws per day. The data just isn’t meaningful… ignore the validity for a second. Real analysts need more data.

National Geographic gives us this wonder piece of journalism that starts with the estimates of 9 year olds and moves on to a single viral turtle video — what a great scientific institution they’ve grown into.

This article quotes another, which via its own nebulous and poor analytics claims that 8 million tons of plastic flow into the world’s oceans every year. That is horrifying! Is it accurate — I doubt it, but I don’t have time for that right now.

Triangulation — Benchmarks

Taking this article/study at its word, we can do some simple comparisons as we now have a benchmark. Our 8 million tons of plastic would include approximately 14 tons of straws (feel free to check my math). That is -

0.000175% —of the pollution is US straws. It just got less significant…

Yes but — 14 tons is a lot!

Opportunity Cost

Real analysts prioritize. Yes, fourteen tons is a lot. But it is not significant. There are better targets. There are more important targets. At least assuming your goal is to stop a massive source of plastic pollution… is that really the goal? Just saying…

Unintended Consequences

I am not a big fan of straws, although my dentist recommends them for sugary drinks. But hey, what is a few more cavities to save the ocean — right? My issue is that the glasses at most restaurants just don’t appear that sanitary. So dump the straw — just give me a disposable cup! Wait… what?

Real analysts don’t ignore their data, ignore opportunity costs, or ignore unintended consequences. Real models should never ignore any of these things.

Perspective — Great Analysts use it.

One final note, but I am warning you — this will offend some. Eight million tons of plastic per year is one hell of a carbon sink! Wait… what?

The low biodegradability of much of this plastic traps carbon removing it from the carbon cycle (atmosphere). It is a carbon sink on par with the swamps that produced our fossil fuels. Granted — landfills are even better as both carbon sinks and controlling pollution. You won’t find either in any global warming model — but then they exclude most undersea volcanoes, too.

Even ignoring the models … or is it the models that ignore… anyway. This should make you further question the intelligence of pursuing an initiative like this in the name of the environment armed with speculative data… at least if you are a real analyst.

Help Milo. Help Woodsy. Avoid straws. Recycle them. Go make your own carbon sink. Or better yet, start worrying about the other 99.99% of the problem.

Just don’t confuse any of these crazy numbers with real data. Don’t confuse any of this “journalism” with real analytics. Better yet, try to learn something from this story. This is all far more common than most people realize.

Either way — thanks for reading!

--

--

Decision-First AI
Course Studies

FKA Corsair's Publishing - Articles that engage, educate, and entertain through analogies, analytics, and … occasionally, pirates!