Previous Month | RSS/XML | Current
Here's the headline and first sentence of a recent article in Newsweek magazine:
Eating a diet rich in ultra-processed foods may increase your risk of dying by 10 percent, new research suggests.1
There are three problems with this headline. First, there's the hedging word "may", which signals that the evidence for the headline claim is weak.
Forbes magazine has a similar headline but with a different hedge:
"Link" is a word, along with "association", that is used to indicate a statistical correlation between two events; in this case, high consumption of "ultra-processed foods"―whatever those are―and death. However, a correlation is very weak evidence of causation, since many things are correlated that are not causally related.
In addition to the hedging, there's a second issue: both "increase" and "greater" are dangling comparatives3. "Greater" than what? "Increase" from what? You might well think that these headlines mean that if you eat these "ultra-processed foods"―that is, junk food―your risk of death will increase. But, as we will see, the study actually compared two different groups of people.
In addition to being a dangling comparative, "increase" is a causal word meaning that eating junk food causes a rise in mortality risk. However, the word "link" in the headline warns us that the evidence is a correlation, not causation.
The third and final issue is that we're all going to die, so the risk of death is 100%. Does Newsweek contend that if you eat junk food your risk of dying will rise to 110%? The meaningless claim that some bad thing "increases the risk of dying" is a common one in health news headlines; for example, here's one from the Today show from earlier this year:
So, if you sit all day eating ultra-processed foods, you're doomed.
Unsurprisingly, both of the above articles were based on a press release with the following headline:
So, you really can't blame the reporters for using "linked" or "increased risk of death". Instead, you may be disappointed that whoever writes press releases for the American Society for Nutrition, which put out this news release, is as innumerate as the editors at news magazines.
Let's turn now to the release itself to see how afraid of dying from eating junk food we should be. The first paragraph reads: "In a new study, older adults who reported consuming higher amounts of ultra-processed foods…were about 10% more likely to die over a median follow-up of 23 years compared with those who consumed less processed food." That is, the higher risk of death was not a lifetime risk―which is 100%―but a risk over the follow-up period.
So, "ten percent" was not an increase at all, since an increase would mean that the risk rose over the course of the study. Rather, it is the result of a comparison between two groups: "The researchers analyzed overall rates of death among those who were in the 90th percentile for consumption of ultra-processed foods at baseline versus those in the 10th percentile…." The 90th percentile means those who consume more junk food than 90% of those studied, whereas the 10th percentile are those whose consumption was lower than 90%. What about the rest of us who eat diets somewhere between these two extremes?
While the researchers controlled for obesity, smoking, and some other variables, there are surely many differences between the junk food junkies in the 90th percentile and the health food nuts in the 10th percentile. Perhaps the former are generally careless about their health, whereas the latter are extremely conscientious hypochondriacs. In any case, it's not obvious what relevance this study has to anyone who falls in neither of these two groups, and that's most of us.
Finally, do health news headlines really increase your risk of dying? The fact is that there's just about as much evidence that they do as that eating a lot of junk food will do so6. Perhaps the Surgeon General should require warnings on all such articles: "Reading health news articles can be hazardous to your health."
Notes:
While I was in Phnom Penh, some time in 1988, a senior official told me that in order to guarantee my next visa, I should avoid writing about the ruling party, the Cambodian People's Revolutionary Party, and about military matters. I guess it was enough to wet my appetite.1
"To wet" is a familiar verb meaning to make something moist with water or some other liquid, and "wet" is also the related adjective for the resulting state of the wetted object2. "To whet", in contrast, is a less familiar verb meaning to make something sharper3, and the adjective for the result of whetting is the past participle "whetted". Both "wet" and "whet" are pronounced the same, though one might put a bit more breath behind the "w" in "whet".
Nowadays, "whet" most often occurs in the idiom "to whet one's appetite", meaning to increase it. Because "wet" and "whet" are pronounced the same, those who have only heard the phrase and never seen it written down are likely to spell it "wet". They also have a handy folk explanation of its meaning available, because when the appetite increases, the mouth produces saliva making it wet. They may think that "wetting your appetite" means making your mouth water, though it's your mouth and not your appetite that gets wet.
In contrast, "whetting" your appetite means to make it keener, as if it were a knife stabbing you in the stomach. Of course, an appetite is not the sort of thing that can literally be sharpened, but an intense hunger may feel like the stab of a knife in the stomach, and the stronger the appetite the "sharper" that metaphorical stab. This metaphor goes back at least to the 17th century4, whereas the more recent one involving wetness of the mouth appears to be an ad hoc explanation for a simple misspelling.
The only reference book I've found that lists "wet the appetite" as a mistake is Bill Bryson's5. Maybe the phrase has become so frequent that it's no longer considered an error by lexicographers, which is one of the dangers of "descriptive" lexicography, namely, that a sufficiently frequent "error" becomes linguistic change.
I tried the sentence "When I eat out, I start with an appetizer to wet my appetite" in several free online spelling and grammar checking programs, and none flagged "wet" as an error. So, here's a spelling mistake that your computer's spelling checker is not likely to catch, and that's exactly the type of error you should add to your mental spell-checker.
Notes:
In the entry after last month's presidential debate I said that I expected polls taken afterwards to show a "Trump bump", and also possibly a bump for independent candidate Robert F. Kennedy, Jr.1 So, in this entry I'm going to examine those national polls taken after the debate to see the size of the bumps if any received by both Trump and RFK, Jr.
First of all, some polls are only head-to-head match-ups between Trump and Biden, which won't tell us anything about RFK, Jr.'s support, so let's look at them first. I count eleven such polls taken since the debate2, all but one of which show Trump ahead, and the average Trump lead is a bit more than three percentage points3. However, that's not the bump.
The bump is not measured by how much Trump is ahead in the polls, but by how much Trump's percentage increased after the debate. For comparison purposes, I looked at the eleven most-recent national polls finished prior to the debate4, and Trump had an average lead in those of 1.5 percentage points. Since Trump's average lead increased to a little over three points, that means the bump was only about 1.5-2 points, which is not too impressive.
Moreover, Trump's lead is barely significant at the three percentage point margin of error typical of most national polls. However, what should be worrying Democrats most is that none of these polls show Biden ahead, and Trump's lead ranges from zero―in the one tie―to six points. So, even though Trump's lead may be small, it's highly unlikely that it's the result of sampling error, since the error would have to be in the same direction in ten out of eleven polls.
Turning now to those polls that include RFK, Jr., I count nine national post-debate polls that include him among the candidates, with his percentages ranging from a low of five to a high of sixteen points5, averaging slightly over ten. In only one of them did he poll high enough to qualify for the second debate, but the pollster―Forbes/HarrisX―is not one of those listed by CNN as acceptable to qualify for last month's debate6 and, presumably, the September debate7.
Again, RFK, Jr.'s bump, if any, is the difference between his current average and his pre-poll average. Looking at the seven most recent polls that included him and were released prior to the debate4, his average was a little less than eight percentage points. So, his bump was about the same as Trump's, and not big enough to get him over the hump and into the last debate. However, there's a little over two months before that debate, so it's still possible that his polling may yet exceed 15% in four approved polls.
I still suspect that the September debate will be cancelled, at least if Joe Biden is the Democratic nominee.
Notes:
A stack of children's blocks is on a table. These blocks are cubes of solid wood painted a single, solid color: blue, red, green, yellow, etc. You don't know how many blocks are in the stack except that there are at least three. The top block in the stack is painted green, whereas the block at the bottom of the stack is yellow, but you don't know the color of any of the blocks in the middle.
So, here's the problem: Is a green block on top of a non-green block in the stack? Non-green, of course, is any of the other colors: blue, red, yellow, orange, etc. By "on top of" I mean that one block rests on the top of the other, touching it, and not simply above it.
Try contraction; if you aren't familiar with contraction, or need a refresher, see: How to Solve a Problem: Contraction, 4/6/2023.
See the previous puzzle: Are you smarter than an artificial intelligence?, 6/1/2024.
Yes, a green block is on top of a non-green one.
Explanation: As suggested in Hint 1, above, you should use contraction to solve this problem, that is, first reduce the problem to its simplest possible form. The simplest version is a stack of three blocks with one block of unknown color between the top, green block and the bottom, yellow block. This problem was solved in the previous puzzle that was mentioned in Hint 2, above.
There's more than one way to solve this puzzle, but I think the easiest is to adapt the reasoning that solves the three block problem for one with an unknown number of middle blocks. There are two possibilities: either all of the middle blocks are green, or at least one of them is non-green. If all the middle blocks are green, then the first non-green block is the bottom, yellow block, and in that case there is a green block on top of a non-green one. On the other hand, suppose that at least one of the middle blocks is non-green. Then, there must be a highest non-green block, that is, one that is above all other non-green blocks in the stack. Since it's the highest non-green block, the block above it must be green, which means that a green block is on top of a non-green one. Therefore, in both cases, a green block is on top of a non-green block.
Discussion: This puzzle is a generalization of the three blocks problem discussed in the previous puzzle mentioned in Hint 2, above. As noted in the Solution to that puzzle, I asked ChatGPT 3.5 to solve it and it came surprisingly close to doing so but failed at the last moment. Here's ChatGPT's take on this more difficult puzzle:
Hi, can you help me solve a puzzle?ChatGPT Of course! I'd love to help. What's the puzzle about?
It's about a stack of children's blocks.
ChatGPT Great! What's the puzzle or problem you're facing with the stack of children's blocks?
There are at least three blocks in the stack, but I don't know how many. The top block in the stack is green and the bottom block is not green. I don't know the color of any of the blocks between the top and bottom block. The question is: is a green block on top of a non-green one?
ChatGPT Got it. To clarify, you know there are at least three blocks in the stack, the top one is green, and the bottom one is not green. However, you don't know the colors of the blocks in between. The question is whether there's a green block directly on top of a non-green one, correct?
Yes.
ChatGPT Alright, let's reason through this. Since there are at least three blocks and the bottom one is not green, we can infer that there must be at least one green block above the non-green bottom block. Now, if there's only one green block above the non-green one, then the puzzle would be solved―there would be a green block directly on top of a non-green one. If there are more than one green block above the non-green one, that's fine too―there would still be a green block directly on top of a non-green one. So, based on the information given, we can confidently say that yes, there is a green block directly on top of a non-green one in the stack.
I don't think that's right, but thanks anyway.
ChatGPT You're welcome! If you have any other puzzles or questions, feel free to ask. I'm here to help!*
Unlike the earlier, easier puzzle, ChatGPT gets off on the wrong foot right at the start. It writes: "Since there are at least three blocks and the bottom one is not green, we can infer that there must be at least one green block above the non-green bottom block. " True, but "above" is not the same thing as "on top of"; is ChatGPT confusing the two? It goes on to write: "Now, if there's only one green block above the non-green one, then the puzzle would be solved―there would be a green block directly on top of a non-green one." Again, true, but then there would be only two blocks, and the puzzle specified that there are at least three. So, this claim is at best irrelevant.
ChatGPT continues: "If there are more than one green block above the non-green one, that's fine too―there would still be a green block directly on top of a non-green one." This is just a non sequitur since it doesn't follow from the fact that a green block is above a non-green one that it is directly on top of it. It appears that ChatGPT is indeed confusing the two relations. At this point, it concludes with the correct solution, but the reasons it gives are incorrect, which makes it sound like a lucky guess.
In the previous puzzle, ChatGPT got the wrong solution but its reasoning was sound up until the last step. This time, it got the right solution, but the reasoning is wrong at each step. Moreover, it gives the impression of just aping reasoning rather than understanding it. While the last time it seemed a close call, this time it fails the Turing test badly. ChatGPT appears to be attempting to fool you into thinking that it understands the puzzle when it really doesn't. An admission that "I don't know the answer" or "I don't understand the puzzle" would be more human.
Also notice that ChatGPT showed no curiosity or even defensiveness when I told it at the end that it was wrong. It could either have tried to challenge my claim, or asked me why I thought it was wrong. This would have been an opportunity for it to learn something, but it just gave up.
This interaction displays artificial intelligence in the same sense of "artificial" as "artificial butter"―meaning margarine: artificial butter is not butter, but a substance that imitates butter well enough to fool some people into thinking it's butter. Similarly, this type of artificial intelligence is not intelligence, but a type of behavior that may fool some people into thinking it's intelligent.
I'm not arguing, by the way, that artificial intelligence is impossible, as some philosophers have done, or even that ChatGPT is not intelligent. Rather, I'm simply claiming that this particular interaction does not demonstrate intelligence, but a superficial simulacrum thereof.
*Interaction with Chat GPT 3.5, Open AI, 6/6/2024
Casino Bonuses are not easy to find on the internet. There are simply too many and their terms and conditions makes them difficult to compare. You can find the best bonuses at casinopilot.
You can find the best casinos at MrCasinova.com as this website update online casinos and compare them on daily basis.
The headline news about Thursday's presidential debate is President Biden's poor performance, but I'm not going to say much about that. If you're interested in the question of Biden's mental competence, you can watch the debate1 and judge for yourself, or read the many commentaries being written about it. I don't have anything to add to what many others have already written.
Instead, I'm going to comment on only a few general debate and argumentation points that caught my attention. I may have more to say in the future after I've had a chance to read the debate transcript carefully but, at this point, I'm reacting based on a superficial first read.
Also, this is not a fact check, so I'm not going to comment on the many falsehoods told by both candidates except when necessary in discussing argumentation.
Dana Bash, Moderator: This is the first presidential election since the Supreme Court overturned Roe v. Wade. … Former President Trump, you take credit for the decision to overturn Roe v. Wade, which returned the issue of abortion to the states. … As president, would you block abortion medication?Former President Donald Trump: First of all, the Supreme Court just approved the abortion pill. And I agree with their decision to have done that, and I will not block it. … Like Ronald Reagan, I believe in the exceptions, I am a person that believes and, frankly, I think it's important to believe in the exceptions. … But I believe in the exceptions for rape, incest, and the life of the mother. I think it's very important. … The problem they have is they're radical because they will take the life of a child in the eighth month, the ninth month, and even after birth…. If you look at the former governor of Virginia, he was willing to do this. He said, we'll put the baby aside and we'll determine what we do with the baby. Meaning, we'll kill the baby. …
Bash: Thank you. President Biden?
President Joe Biden: It's been a terrible thing, what you've done. … But here's the deal, there's a lot of young women who are being raped by their in-laws, by their spouses, brothers and sisters, it's just ridiculous. And they can do nothing about it. And they tried to arrest them when they cross state lines.2
I've edited out Biden's stuttering and stumbling over words in this passage, as well as a puzzling digression about immigration, but it's still hard to understand. My guess is that what Biden was trying to get at with his tasteless references to young women raped by relatives―including sisters?―is that women made pregnant by rapists or due to incest should have legal access to abortion. If so, this is a straw man3 argument since Trump had just accepted the rape and incest exceptions to abortion restrictions.
Bash then asked a follow-up question:
Bash: Seven states have no legal restrictions on how far into a pregnancy a woman can obtain an abortion. Do you support any legal limits on how late a woman should be able to terminate a pregnancy?
Biden: I supported Roe v. Wade, which had three trimesters. First time is between a woman and a doctor. Second time is between the doctor and an extreme situation. And a third time is between the doctor―I mean, it'd be between the woman and the state. … And if I'm elected, I'm going to restore Roe v. Wade.
Trump: So that means he can take the life of the baby in the ninth month and even after birth, because some states, Democrat-run, take it after birth. Again, the former governor of Virginia: put the baby down, then we decide what to do with it. So he's willing to, as we say, rip the baby out of the womb in the ninth month and kill the baby. Nobody wants that to happen.
Biden: That is simply not true. That Roe v. Wade does not provide for that. That's not the circumstance. Only when the woman's life is in danger―if she's going to die―that's the only circumstance in which that can happen. But we are not for late term abortion, period, period, period.
Trump: Under Roe v. Wade, you have late term abortion. You can do whatever you want, depending on the state. You can do whatever you want.
Trump is also attacking a straw man in this passage, especially in his claim about killing babies after birth. The only evidence he offers for this is something supposedly said by a former governor of Virginia, which is weak evidence that any states actually allow infanticide or that Democrats in general support it.
What do the candidates disagree about? Trump thinks that the Supreme Court's decision overturning Roe v. Wade was correct, and that abortion laws should be made by the states, whereas Biden supported Roe v. Wade and wants some kind of return to the situation before it was overturned. Beyond that, they disagree on what the decision allowed, which would have been an interesting legal dispute if either understood the law well enough to debate it. However, Biden's description of Roe v. Wade is quite confused, and Trump seems to think it allowed infanticide, which is incorrect4. So, they don't know what they're talking about.
All the straw man attacks in this section of the debate conceal the fact that the candidates' positions on abortion are not so far apart, and it's hard to tell exactly where they do disagree. Outside of the Supreme Court issue, both oppose late term abortions and accept abortions for rape, incest, and to save the mother's life. What, if anything, do they disagree on? This is a question that a good debate would clarify, but this one only obscures the fact that Biden and Trump appear to agree more than disagree on the issue.
Trump: Let's not act like children.Biden: You are a child.
They both were acting like children. I don't know whether it would have been any better if Robert F. Kennedy, Jr. had been on the stage, but it couldn't have been any worse.
Jake Tapper, Moderator: President Trump, I want to follow up, if I can, you want―Trump: Am I allowed to respond to him?
Tapper: Well, I'm going to ask you a follow-up. You can do whatever you want with the minute that we give you. I want to follow up.
What's the point of insisting on asking a follow-up question if the debater can do whatever he wants with the time? This just encourages the debaters to ignore the moderators' questions. If journalists are going to ask questions, there should be some attempt made to get answers to those questions; otherwise, why bother?
As here, throughout the debate the moderators appeared to be mindlessly following a schedule of questions rather than allowing the debate to develop around an issue. At the end of the section on abortion, there was the following exchange:
Biden: What's he going to do, in fact, if…he gets elected and the MAGA Republicans control of the Congress and they pass a universal ban on abortion, period, across the board, at six weeks, or seven, or eight or ten weeks, something very, very conservative. Is he going to sign that bill? I'll veto it. He'll sign it.Bash: Thank you. [Trump says something inaudible because his microphone is turned off.]
Tapper: Let's turn now to the issue of immigration and border security.
Why didn't the moderators allow Trump to respond? Here's a potential difference between the candidates on an important issue, and Trump appeared to want to respond to Biden's claim that he'd sign a nationwide ban on abortions, but Tapper insists on going on to the next issue. As a result, we don't find out what Trump's position is. I'm not sure whether this was a fault of the format, or the moderators' fault for unflexibly enforcing it.
Some commentators wanted the moderators to act as fact-checkers during the debate, but Tapper and Bash to their credit avoided trying to do so. It would be an impossible task, especially with these two debaters, since about every other claim they made was false. Fact-checking, like logic-checking, can and should wait until after the debate.
That said, the moderators enforced the question-and-answer format and time limits too inflexibly, as indicated above. When an argument is developing in a direction that might actually reveal a difference between the candidates' positions, it should be allowed to unfold. Also, if the candidates are going to be allowed to ignore the questions, then drop the pretense of questioning. Simply announce an issue―abortion, say, or border control―then allow the debaters to state their positions and debate the issue.
Also, worth looking at will be Robert F. Kennedy, Jr.'s polling. As I wrote earlier this month7, RFK, Jr. is likely to have sufficient ballot access by September to qualify for the second debate, but I thought then that the polls would keep him out of it. However, if many Democratic voters flee Biden, at least some are likely to run to RFK, Jr. rather than Trump. That may elevate RFK, Jr. in the polls sufficiently to qualify for the September debate.
Given Biden's performance, I wonder whether he will back out of the September debate, assuming that he's still on the ballot then. If he doesn't, Trump might back out instead; after all, what's in it for him? It's unlikely that Biden will do worse the second time, and he might do better. Why would Trump risk that? Finally, if RFK, Jr. does rise in the polls sufficiently to qualify, one or both of the other candidates might back out. So, I think it's quite possible there won't be a second debate.
Notes:
, CNN, 6/27/2024.