Previous Month | RSS/XML | Current

WEBLOG

HEADLINE!
October 5th, 2019 (Permalink)

Canadian Bakin'

Are you a healthy skeptic? By that I mean, do you have a healthy skepticism about what you read and hear?

What is a healthy skepticism? It's one that lies between an unhealthy skepticism and an even unhealthier gullibility. Both unhealthy skepticism and gullibility are all-or-nothing attitudes: in the case of gullibility it's "all" and it's "nothing" for the skeptic. An unhealthy skeptic tends to dismiss everything that he or she doesn't agree with, whereas a gullible person tends to believe it all, no matter how implausible.

Test the health of your skepticism on the following alarming headline1:

1 in 20 young Canadians are hospitalized every day for substance use2

If this headline doesn't cause your internal alarm to go off, then your skepticism needs a check-up. The first thing that should set that alarm ringing is a lack of clarity. What does it even mean?

The verb "to hospitalize" usually means to admit to a hospital, so in that sense the headline would mean that 5% of the population of young Canadians are admitted to the hospital each day. However, unless the hospital stays were very short, after a few days a large percentage of Canadian youth would be in the hospital. Surely, then, the hospitals in Canada would be overwhelmed. Have you read or heard any reports about hospitals in Canada being over-run by young patients? If they were, wouldn't you have? Moreover, given what you know about "substance use", is it plausible that so many Canadian youngsters have a problem severe enough to require hospitalization?

In addition to a lack of clarity, another basis for skepticism is implausibility. How do you check plausibility? You use what you already know: you know many things, perhaps more even than you realize. Ask yourself whether the headline fits with what you know.

Perhaps "hospitalized" just means "in the hospital", so that the headline means that one out of twenty young Canadians are in the hospital for substance abuse on any given day. 5% of the population of Canadian young people would be a large number of patients, though not so large as if 5% were admitted every day. For this reason, this interpretation is more plausible than the previous one, but is it plausible? If 5% are hospitalized, what additional percentage would have a "substance use" problem not so severe as to require hospitalization? Surely, for every person with such a severe problem there are several with less severe ones. If Canada were suffering from such an epidemic of "substance use", wouldn't you have heard or read something about it?

Given both its ambiguity and implausibility, it seems likely that something went wrong with this headline, but to find out what we need to read the article itself. The first sentence reads: "A new report that looked at Canadian youths aged 10-24 finds that some 65 of them are hospitalized every day for substance use issues.3" That 65 Canadian youngsters are admitted to hospital every day seems plausible, but not that that represents 5% of the youth of the country.

As Lawrence Mayes, who submitted this article, commented in an email to me: "I know Canada's population is small but I'm sure it's not that small.4" Here Lawrence was using what he knows to test the claim's plausibility: he knows that Canada's population is small, but it's large enough that there must be more than 1,300―20 × 65―young Canadians.

Another way to check claims for plausibility is through cross-checking, that is, checking related numbers against one another for consistency. For instance, in the next full sentence of the article we read: "Between 2017-2018, there were more than 23,500 hospitalizations5 among youth―or 1 in 20 of those ages 10-24―because of substance use." So, apparently 23,500 is supposed to represent 5% of the Canadian population aged 10-24, which is certainly more plausible than 65. However, these two numbers are inconsistent.

At this point, our plausibility checks have shown that something went wrong with this article, but we don't know exactly what. Given that 23,500 is 5% of the population of Canadian youth, the total population of young Canadians would be almost half a million. Is that plausible?

Now, if you are a Canadian or happen to know at least approximately what the population of Canada is, you could continue to check this against what you know. Unfortunately, as an American citizen, I don't know the population of Canada, though I do know that it's less than the United States. So, I had to do a little research.

According to Wolfram Alpha, the total population of Canada is about 37 million and life expectancy is over eighty years6. So, an age group of fifteen years should represent almost 20% of the total population, that is, around seven million people. So, a half million young Canadians seems far too low a number.

That exhausts what we can find out from this short article, so if we want to learn more we'll have to turn to the study that the article is reporting7. Among its "Key Findings": "1 out of every 20 hospital stays among youth age 10 to 24 in Canada in 2017-2018 were related to harm caused by substance use.8" It goes on to say:

In 2017-2018, there were 23,580 hospital stays for harm caused by substance use among youth age 10 to 24. This is the equivalent of 65 youth hospitalized every day in Canada.8

Apparently, this is the source for the headline. However, these claims refer to 5% of the hospital stays of young Canadians, not to 5% of the youngsters themselves. Given that people in that age range are usually healthy and seldom spend time in the hospital, no wonder only 65 a day were involved. The report also reveals the origin of this number: it's the result of dividing 23,580 by 365 days. So, the reality is much less worrisome than the headline: 23,580 is .3% of seven million9, not 5%.

This article is an egregious example, which is why I selected it as an exercise in healthy skepticism and plausibility checking. If you approach every news article you read with an alert mind, a healthy skepticism, and armed with the ability to apply what you already know, you'll seldom uncover such an extreme error. However, critical reading is not only useful when it uncovers errors: if an article passes your skeptical scrutiny, it should give you greater confidence in what you read.


Notes:

  1. Shraddha Chakradhar, "Morning Rounds", STAT, 9/19/2019.
  2. This is not an entry on doublespeak, but notice the euphemism "substance use" in the headline and the article itself. As we find out later in the article, the substances were drugs, with marijuana and alcohol making up 65% of those used. The article doesn't say what the remaining 35% of the substances were but, according to page four of the report that was the basis of the article, they were other drugs―see note 7, below, for the citation. Why not word the headline: "1 in 20 young Canadians are hospitalized every day for drug use"? Other problems aside, this would be both more informative and shorter. See also: Close Encounters with Doublespeak of the Third Kind, 9/8/2019.
  3. Same citation as the first note. What does "issues" add to this, other than an extra word?
  4. Lawrence Mayes, private email, 9/19/2019.
  5. Notice that what is counted here is "hospitalizations", yet the article treats the number as representing the patients hospitalized. Given that there were probably some cases of individuals hospitalized more than once in the same year, the actual number of patients hospitalized would be lower than the number of hospitalizations. See note 9, below.
  6. "What is the population of Canada?", Wolfram Alpha, accessed: 10/3/2019.
  7. "Hospital Stays for Harm Caused by Substance Use Among Youth Age 10 to 24", Canadian Institute for Health Information, 9/2019.
  8. P. 5; see the previous note for the source.
  9. This is assuming, falsely, that every hospital stay is by a distinct young person. We learn on the same page: "17% of youth who were hospitalized for harm caused by substance use were hospitalized more than once for substance use within the same fiscal year." So, fewer than 23,580 youngsters were hospitalized in the year covered by the report.

Previous Month | RSS/XML | Current