Statistics 3: The Search for Statistics 2
Statistics. We hear many of them every day, and most of them sound trustworthy. But sometimes you hear a statistic and you think that it sounds a bit weird. Or maybe you just wonder where that statistic came from. Following this train of thought can lead all sorts of places. Sometimes you just can’t find a source. Sometimes you can only find a secondary source.
This is the situation I found myself in recently. Back when EnduraMark was an engineering project, we found a statistic that says that 400 million whiteboard markers are thrown out in the United States every year! That’s a lot, and it made sense as a sufficiently large number back when we at EnduraMark were sleep-deprived college students. However, we were talking about that statistic in a recent team call and, this time, something sounded off about it. 400 million markers is a lot, but it’s close to the number of people in the USA (about 330 million). So, on average, each person in the USA only consumes one marker each year? That sounded weird to us, so I decided to get to the root of this.
I started by looking at our project report bibliography to see where exactly we got that stat. We got it from a competitor's website, and they didn’t cite their source. This may look like a dead end, but I still had a few tricks up my sleeve. I went on Google and searched for “400 million” markers and “400 million” whiteboard markers. The quotes around “400 million” are so that Google only returns results with the exact phrase “400 million”.
With that search, I mostly found articles using that same figure. Some of those articles even had a source listed! Unfortunately, that source is the same source we found already, so that doesn’t get us any closer to an answer. I also found someone on reddit who said that they learned that stat and they cited Crayola’s now-paused ColorCycle program, which collected used markers from schools. The page for the ColorCycle program doesn’t mention the 400 million stat, but I used the Internet Archive’s Wayback Machine to check the ColorCycle page at various past dates and still couldn’t find that stat. The pieces of the press briefing announcing this program that I saw in my search didn’t mention this stat.
From here, I searched Google for “400 million” crayola colorcycle, and I found a few articles that mentioned that stat and the ColorCycle program. I assume that they got their stat from Crayola, but they could’ve also got it from the original source we found.
Interestingly, one of the websites I found in my initial search, the Green Schools Initiative, says that “Some estimates suggest that 400 million used whiteboard markers -- that's 25,000 tons -- are thrown out in U.S. schools every year, ending up in landfills.” There are two interesting caveats in this quote. First, saying that some estimates suggest. This acknowledges a level of uncertainty with this statistic, and it also implies that other sources have different estimates. In my searches for how many whiteboard markers are thrown out every year, including searches that don’t include the 400 million figure, I was unable to find other estimates. Second, saying “in U.S. schools”. The other instances of this stat I’ve seen said that 400 whiteboard markers are thrown out in ALL of the US every year, not just in schools. That caveat changes so much! If 400 million are thrown out JUST in US schools, then there are some thrown out in the US OUTSIDE of schools, which makes the total US number bigger. Failing to include a few little contextual words can completely change the meaning of a statistic!
Going back to an earlier point, the word “estimate” leads to a big question: what methodology was used to get this number? What assumptions were made? What data was collected to make this estimate? Without this information, and without a definitive primary source of this statistic, this 400 million statistic doesn’t mean much. I don’t even know WHICH year this 400 million markers a year statistic is from. Based on when we initially found it (2017), the launch of ColorCycle program in 2013, and the fact that our unnamed competitor claims to be at least 15 years old in 2023, I estimate that this estimate is from sometime between 2007 and 2017.
With all this ambiguity from one statistic and about 50 tabs in my internet browser later, what do we do now? For now, we've decided that 400 million markers a year in the US sounds like a reasonable enough estimate. We're going to continue looking for other estimates and we might even try to make our own estimates to double check this.