Hacker News Reader

Comment by mieseratte | Hacker News Reader

The answer to the clickbait headline is "pollution," and the town is Waycross, GA. Here's your "money" paragraph:

> In late 2015, the Georgia Department of Public Health said it could find no link among the children’s cancer cases. Then it backtracked and said more investigation was needed. In December, the federal government stepped in. The Agency for Toxic Substances and Disease Registry, an arm of the Centers for Disease Control and Prevention, said it would work alongside state officials in evaluating contamination at the railroad yard as well as an Atlanta Gas Light property that once held a power plant, which was torn down after closing 60 years ago.

View on HN

Added nuance: A government-sponsored investigation using a narrow data set generated by a company responsible for the cleanup of a single site, of which Waycross has multiple under investigation, decided there was no conclusive evidence of Waycross having abnormally high levels of cancer, excluding diagnoises of cancer that occured outside of the town even if the diagnosises were on town residents.

Also: A town newspaper did not cover this, and the head editor as well as the mayor believes this is a case of "Facebook moms" and overly-concerned parents using personal griefs to wreck Waycross's tenuous economic bounce back.

>excluding diagnoises of cancer that occured outside of the town even if the diagnosises were on town residents.

I would be floored if they did this. I'm a statistician who reports on cancer rates for another state, and the universal standard for assigning locations to cases is the patient's residential address. Looking at Georgia's 2016 report[1], they mention using the SEER*Stat software provided by SEER[2]. I have the same software, and I've never seen it offer any information about the reporting facility beyond a general category ("hospital", "lab", etc). They would've had to work outside the well-worn paths using a custom extract from the cancer registry. Not impossible, but also not reasonable.

It's true that epidemiologists, or whoever looked into the clusters, could've been a separate unit from the statisticians behind the annual reports. And maybe they're not used to cancer epidemiology and didn't ask about best practices. But I'd be surprised if they messed up this basic thing.

Also, I've heard of the study mentioned in the article where only one out of 428 cluster investigations proved true. But the way it was portrayed at a cancer registry conference, the low rate is a combination of small sample sizes and over-reporting of "clusters." Cancer's a rare and frightening event, which are two things people have trouble seeing rationally.

[1] https://dph.georgia.gov/cancer-reports

[2]https://seer.cancer.gov/

> excluding diagnoises of cancer that occured outside of the town even if the diagnosises were on town residents.

That's a pretty big exclusion. It's not as if pollution gives a damn about political boundaries.

it's also clearly irrelevant from a medical standpoint where the diagnosis was made... unless there's local bribing of doctors going on!

so why insist on only counting local diagnoses?

EDIT: perhaps there are regulations that mandate doctors to diagnose similar percentages, to prevent overdiagnosis... and perhaps this is being weaponized against discovering localized exposure to carcinogens?

> "On the cancer cluster question, ATSDR cites data from the Georgia Comprehensive Cancer Registry to make the case that there isn’t one. GCCR data from the past decade shows childhood cancer incidence has been below the state average here. GCCR’s Waycross-area data, however, "doesn’t include people who are diagnosed at hospitals in other cities", as Lexi, Harris, Gage, and Raylee all were."

Charitably, I guess the researchers just don't have resources to locate patients more accurately than by their diagnosing hospitals location. It's a huge amount of noise on top of patients actual location history, which would seem to make their survey only capable of detecting persistent hotspots that span the location of many hospitals.

>"On the cancer cluster question,"

Either it's on the cancer cluster question, but then you contact and ascertain the accurate patient location history.

Or the data never was on the cancer cluster question...

I note that it is missing from the Wikipedia page, as well.

I wouldn't call it clickbait. It's a longform piece telling the story in a way that reflects the uncertainty that the locals feel.

I called the headline clickbait. I never said TFA is garbage, but the headline certainly is.

The point is it's an investigation not a conclusion.

The author heavily focuses on pollution in the area being a possible cause, and points out inadequacy in investigation performed on and arguments against pollution being the cause. But all her evidence is anecdotal.

The point isn't to assert that she is certain of the cause, it's to provide a spotlight on evidence which she has seen which has not been given adequate examination and the resulting suffering from what could possibly be preventable.

The title is appropriate, not clickbait.

The title is clickbait, because it deliberately withholds key information to force readers to click. A non-clickbait title would be:

"Pollution suspected of causing rare cancers in Waycross, GA"

While that's a solid point, I think a critical part of why people look down on "clickbait" is the additional ingredient of "withholds key information to force readers to click ... and that key information makes the content substantially less interesting than the headline implied." The reason we call it clickbait is because it implies that it would clearly not be something you'd want to click, were the hook left unbaited.

Otherwise, it's not all that bad - just clever headline choice. (The point of headlines has always been to intrigue prospective readers, after all.)

Since this article is a substantial piece of journalism that remains interesting if you know the tl;dr, rather than an insubstantial chunk of #content, I don't think it really counts as clickbait.

When I was small we learned the inverted pyramid model of news, that the most important information should be at the top (the headline) and succeeding content should be ordered primarily in descending order of information significance.

Did this change at some point? Or is this just tabloid practices infecting journalism in the era of maximizing clicks? I find the old model infinitely better especially as people tend not to read past headlines in most cases.

The key information does reduce the expected level of interest. Controversy about cancer clusters and pollution is nothing new. But there are other possibilities, e.g. it might be caused by some genetic disorder that's found in a family who happens to live there, and the article might be about identifying and diagnosing that. It might be about some infectious disease that causes cancer like HPV. It might be about a cultural practice, e.g. eating carcinogenic foods like hijiki seaweed. Unless you actually live there, or know somebody who does, all these are more interesting to read about than the real "residents say it's pollution, industry says it isn't, everybody argues but nothing is really decided" story that's happened so many times before.

> The point of headlines has always been to intrigue prospective readers, after all.

Which would have been a better headline on, say 9/11?

A: “U.S. Attacked - Hijacked Jets Destroy Twin Towers”

B: “Something Awful Happened in NYC Today. Click here to find out what!”

The article seems well-researched and informative. It made me think about the balance between economic growth's positive and negative outcomes. If it takes a modicum of editorial trickery to get more eyes on such a piece, I think we're fine.

> A non-clickbait title would be: "Pollution suspected of causing rare cancers in Waycross, GA"

That's arguably worse than clickbait. It actually discourages people from reading the article who might find it interesting.

It suffers from being both too specific and too general at the same time.

It's too specific because Waycross, GA, is a town of under 14k people and as far as I can see there is no reason to expect that many people outside of Georgia to have ever heard of it. Naming it in the headline suggests that the story is aimed at those people who do recognize it.

It's too general because of course it is pollution. The other things that cause cancer don't tend to cause cancer spikes in small areas.

Both the town and the cause are actually incidental to this story. The interesting aspects of the story are how the various parties reacted when some started noticing there might be a cancer cluster.

considering something to be bait or not is subjective. so why argue on it ? :D wtf guys

The headline mirrors the article's portrayal of the quest to find out the truth. If the article was just a few paragraphs long, I would have felt slighted. But as such, it read more like a mini-novel/documentary.

And of course, it's pure coincidence that the headline is perfectly tuned to cause fear to small-town Georgia residents and to their friends and family elsewhere, and the extra page views and ad revenue from people worried if it's their town are of no relevance whatsoever.

If there were a viable news business that allowed local journalists to investigate the issue, perhaps this wouldn't have happened, or have been uncovered earlier.

it does not appear to have been a power plant, it was likely a coal gasification plant

"The power plant helped light Waycross from 1916 to 1953. Along the way, it discharged coal tar—a thick, black byproduct of converting coal to fuel—into the canal system. Prolonged exposure to coal tar has been linked to various types of cancer...."

The first half of the article also calls out multiple toxic organic chemical contaminants from several operations including a railroad and a waste-processing plant ... severe enough to require Superfund remediation.

In that era ('Better things for better living through chemistry') most people were unaware of the dangers of industrial chemicals in the air, water and soil. They got little press, at least until 'Silent Spring'.

Coal Tar is the prime byproduct of coal gasification- the timeline is right too, most coal gas plants were shut in 45-55 timeframe.

What's the town?

Waycross, GA

Reply on news.ycombinator.com