Thread by @tylerblack32

/1 Thread: C'mon @nytimes, you can you please get suicide statistics right?!!

We really need to address the NYT's propensity for not understanding, & worse, poorly communicating suicide statistics. A new article.

Sigh. Here we go again.

/2 (overall, the article a compelling read!)

But In it, we have college campuses reporting suicides, however the #'s are 1, 2, 3... this is not actually abnormal.

Let's zoom in on one: Worcester Polytechnic with 2 (possibly 4 after investigation, according to the NYC).

/3 Worcester Polytechnic has concern because of a cluster of deaths, and I understand that, but we actually have CDC provisional data for Worcester County for 18-22 year-olds.

The 2019-2020 vs 2020-2021 suicide frequencies are NOT different. And they are lower than 2018-2019's.

/4 Certainly, having multiple suicides at WPI causes concern (clusters do!), but this means that other uni's in the area had NO suicides (& no administrators ringing alarm bells). I get why this stat seems "truthy."

But this is why the expertise in suicide stats is necessary.

/5 (note, I cannot provide the numbers due to CDC privacy restrictions as they are very small)

Low frequency events such as suicides on campus are prone to "noise" that we call variance - a few suicides higher or lower dramatically increase the relative frequency.

/6 Worcester, MA has EIGHT universities. It must be so stressful on WPI's campus during this time, and my heart goes out. But there are 7 other universities in the area. Clusters ALWAYS suck, but the larger picture is clear: no significant difference in the county.

/7 Famously, the @nytimes made this mistake in a huge way in @EricaLG's article about Clark County - describing the near "doubling" of suicide rates among kids (And wielded to justify a superintendents crusade to reopen schools).

/8 This is Clark County for the past 6 years. um.... are we sure 2020-2021 was the abnormal year?

/9 And this is the past 21 years. We can clearly see the tremendous variability in the yearly suicide rate, and it is **NOT 2020-21** that is the outlier (like the NYT article claimed), it was the 2019-20 year that was **ABNORMALLY LOW**.

/10 Yet the entire thesis of that article and the justification the superintendent used for the policy decisions he made was based upon the "spike in suicides."

/11 CONSIDER THIS AGAIN: Read the headline, then look at the graph.

THERE WAS NO SURGE!!! It has been following a noisy, increasing trend, and **2019 WAS THE OUTLIER**.

By the way, 2019 numbers were released in December of 2020, and GUESS WHAT THE NYT *DIDN'T REPORT???*

/12 I called it out at the time but because of "scary=more clicks" and "it bleeds it leads", the NYT Clark County article with an... um... blind spot to the reality of the Clark County statistics was picked up in news outlets around the world.

https://twitter.com/tylerblack32/status/1353428208553664513

/13 Looking at it another way, the suicide rate between march-february has varied tremendously year to year in Clark County, and has been >+50% 6 times in the past 21 years. 2020 was NOT the highest increase. There was no school closures or pandemic to blame in 2009 or 2013.

/14 At the EXACT moment the @nytimes article about Clark County, an article by a local news agency in Kansas showed a 33% reduction in child suicides (equivalent to a 50% increase). It was shared ONCE. By me. Once. The NYT article? Picked up across the world thousands of times.

/15 The @nytimes stuff REALLY frustrates me though. They are widely cited and seen as the "paper of record." *They can hire, contact, or work with statistical experts who could correct these errors*.

/16 There are people with whole degrees (not me, though I dabble!) in statistics that can visualize & compare year-over-year changes.

/17 I am freely available (and some NYT reporters HAVE reached out to me and I've been more than happy to explain the nuances of suicide statistics).

/18 And to all who are going to be made to fear what the pandemic has "done" to kids suicide, please see my threads:

FIRST: What happened to "Kids 8-17" in America (short version: LESS SUICIDES DURING STRONGEST LOCKDOWN).

https://twitter.com/tylerblack32/status/1470785676316663820

/19 THEN: What happened to "Young Adults 18-24" in America (short version: no significant changes overall, but significant racial differences that are important)

https://twitter.com/tylerblack32/status/1471163163466997766

/20 I enjoyed @anemonanyc's article generally. It's just... The use of suicide statistics. It sucks.

There is good reason to read the it! I think the narratives are compelling, and I think about how challenging this has been on learners. IN NO WAY do I want to diminish that.

/21 The Pandemic has caused 1.3 mil children worldwide to lose a primary caregiver. Schools have shut down, reopened, and there are so many stressors to cope with. Please remember that we have no real ability to separate *pandemic* effects from *closure* effects re: kids.

/22 But, there has ALWAYS been a severe need for more child and youth mental health supports and the pandemic has EXPOSED the need, not created the need.

/23 The article:

nytimes.com/2021/12/22/us/…

The previous NYT article that made the same variance/historicity error re: clark county:

nytimes.com/2021/01/24/us/…

/24 FOOTNOTE:

Since the Clark County Article was written, the rate of suicide in 2021 for kids looks about the same as 2020 so far, in the first 5 months of 2021. So the "reopening of schools" in Clark County, NV did not (so far) result in a decrease in the suicide rate.