3 ms·

Flatly absurd. These are the kinds of connections only a schizophrenic mind could make. "For example, surgeons performed 484 (31%) fewer procedures to remove

Flatly absurd. These are the kinds of connections only a schizophrenic mind could make. "For example, surgeons performed 484 (31%) fewer procedures to remove something from a child’s nose in 2022 compared with 2012." So your sample size is so absurdly small that your conclusions could not possibly have any meaning. What a waste of time this article was. -- EDIT: The population of the UK is 68 million people and this is an entirely retrospective assessment. There's probably a reason they just didn't link to the RCSE page itself as it's speculation is far more reserved: https://www.rcseng.ac.uk/news-and-events/media-centre/press-releases/drop-in-children-needing-operations-for-items-stuck-in-their-nose-and-mouth/ https://www.rcseng.ac.uk/news-and-events/media-centre/press-... That reads far less like a paper trying to "nudge" people in a preferred direction.

dehrmann 2y ago
I wish it had better data to back it up, but it's reasonable to think that coins are a popular thing for kids to play with.
MajimasEyepatch 2y ago
Is sample size even relevant here? If they’re working with NHS data, presumably they have pretty reliable population data (i.e. not randomly sampled) to work with. And regardless, even if it was a sample, I don’t see how a sample of thousands would be too small to detect a very large effect like that, unless the variance was extremely high. (But again, I’m not sure that complaining about “sample size” even makes sense in this case.)
- timewizard 2y ago
  > Is sample size even relevant here? There are two sample sizes. Kids who put things in noses. People who use cash. One is vastly larger than the other. Drawing conclusions from one about the other _purely retrospectively_ is nonsense.
  - MajimasEyepatch 2y ago
    I agree that an observational study cannot prove that the reduction in these cases is due to the rise of cashless payments. The researchers themselves say it is “multifactorial.” However, that has absolutely nothing to do with “sample size.” There is no sampling here. The relative size of the population of people using cash vs. the population of kids sticking things up their nose also has nothing to do with whether is study is valid. The study is essentially making an observation and then presenting a hypothesis: in 2012, the use of cash began to decline in the UK, and over the next 10 years, there was a significant decline in the number of cases reported to the NHS where kids had to have a foreign object removed. Given that 75% of cases historically involved coins, the authors make a totally reasonable inference that the decline in the use of cash in the UK over this exact period likely played a big role. They also name several other factors that may have contributed. Your strong reaction to this study seems disproportionate and misguided. It’s not the most powerful or compelling study, but there’s nothing blatant wrong with it.
krisoft 2y ago
> So your sample size is so absurdly small that your conclusions could not possibly have any meaning. What do you mean by sample size? These are total numbers for the whole year. There is no sampling going on. > The population of the UK is 68 million people Ok? So what? > and this is an entirely retrospective assessment. Again, what does this mean? How else would you notice these trends? Proactively? > There's probably a reason they just didn't link to the RCSE page itself as it's speculation is far more reserved I read both, and they seem to say the same. Where do you see the difference?
- dullcrisp 2y ago
  Well can we really make inferences from the population of the UK to that of bigger countries? Maybe only people on tiny island nations have this problem.
  - Symbiote 2y ago
    The UK has the 21st-highest population of all countries. It's hardly 'tiny'.
Dylan16807 2y ago
More than a thousand hits is a very good sample size. The total population of the UK barely matters in this check for statistical significance. A very general rule of thumb is that you want 10-100 samples in each category. In this case it's about a thousand children in one category and millions in the other category. That's more than enough to measure a 30% difference.

dehrmann 2y ago

I wish it had better data to back it up, but it's reasonable to think that coins are a popular thing for kids to play with.

MajimasEyepatch 2y ago

Is sample size even relevant here? If they’re working with NHS data, presumably they have pretty reliable population data (i.e. not randomly sampled) to work with. And regardless, even if it was a sample, I don’t see how a sample of thousands would be too small to detect a very large effect like that, unless the variance was extremely high. (But again, I’m not sure that complaining about “sample size” even makes sense in this case.)

timewizard 2y ago
> Is sample size even relevant here? There are two sample sizes. Kids who put things in noses. People who use cash. One is vastly larger than the other. Drawing conclusions from one about the other _purely retrospectively_ is nonsense.
- MajimasEyepatch 2y ago
  I agree that an observational study cannot prove that the reduction in these cases is due to the rise of cashless payments. The researchers themselves say it is “multifactorial.” However, that has absolutely nothing to do with “sample size.” There is no sampling here. The relative size of the population of people using cash vs. the population of kids sticking things up their nose also has nothing to do with whether is study is valid. The study is essentially making an observation and then presenting a hypothesis: in 2012, the use of cash began to decline in the UK, and over the next 10 years, there was a significant decline in the number of cases reported to the NHS where kids had to have a foreign object removed. Given that 75% of cases historically involved coins, the authors make a totally reasonable inference that the decline in the use of cash in the UK over this exact period likely played a big role. They also name several other factors that may have contributed. Your strong reaction to this study seems disproportionate and misguided. It’s not the most powerful or compelling study, but there’s nothing blatant wrong with it.

krisoft 2y ago

> So your sample size is so absurdly small that your conclusions could not possibly have any meaning. What do you mean by sample size? These are total numbers for the whole year. There is no sampling going on. > The population of the UK is 68 million people Ok? So what? > and this is an entirely retrospective assessment. Again, what does this mean? How else would you notice these trends? Proactively? > There's probably a reason they just didn't link to the RCSE page itself as it's speculation is far more reserved I read both, and they seem to say the same. Where do you see the difference?

dullcrisp 2y ago
Well can we really make inferences from the population of the UK to that of bigger countries? Maybe only people on tiny island nations have this problem.
- Symbiote 2y ago
  The UK has the 21st-highest population of all countries. It's hardly 'tiny'.

Dylan16807 2y ago

More than a thousand hits is a very good sample size. The total population of the UK barely matters in this check for statistical significance. A very general rule of thumb is that you want 10-100 samples in each category. In this case it's about a thousand children in one category and millions in the other category. That's more than enough to measure a 30% difference.