Comparing countries
- El Pollo Diablo
- Stummy Beige
- Posts: 3324
- Joined: Wed Sep 25, 2019 4:41 pm
- Location: FBPE
Re: Comparing countries
And also, the dumbest thread ever is clearly the coffee one, this is at worst second dumbest
If truth is many-sided, mendacity is many-tongued
Re: Comparing countries
Is the coffee thread someone saying "I like coffee X" and someone else saying "I like coffee Y" and someone else saying "You are both WRONG"?
Awarded gold star 4 November 2021
Re: Comparing countries
Well, how are we supposed to compare the dumbness of threads?
Even the coffee thread is not without value; I went back to it after my sister gave me a kilo of coffee beans for Christmas, because I knew people had discussed coffee grinders there and I obviously now had to spend significantly more than the cost of her present buying the equipment I needed to use her present.
Okay, strictly that probably means the coffee thread has a negative value, but negative numbers are numbers too. Don't sign shame.
Even the coffee thread is not without value; I went back to it after my sister gave me a kilo of coffee beans for Christmas, because I knew people had discussed coffee grinders there and I obviously now had to spend significantly more than the cost of her present buying the equipment I needed to use her present.
Okay, strictly that probably means the coffee thread has a negative value, but negative numbers are numbers too. Don't sign shame.
Re: Comparing countries
As I've said, "You'll have to define your objective(s) much more precisely to arrive at any cromulent conclusion".El Pollo Diablo wrote: ↑Wed Jan 11, 2023 6:57 amI said "comparing", not "ranking".KAJ wrote: ↑Tue Jan 10, 2023 6:41 pmYou'll have to define your objective(s) much more precisely to arrive at any cromulent conclusion.El Pollo Diablo wrote: ↑Mon Jan 09, 2023 10:26 amIf you were to try to build up a picture of the relative state of a country compared to others, what measures would you use? What measures would you avoid? <snip> I'm thinking here across a whole range of aspects of life, such as economic, politics, health, quality of life, and so on.
Some ideas include the following:
- GDP per capita
- Human Development Index
- Gini coefficient
- Democracy index
- Health outcomes / healthcare system (life expectancy, % women dying in childbirth, etc)
- Quality of life
- Happiness
- Religiosity
Any others? And any of the above which are so problematic that they should be ignored (I reckon there should be a high bar for this)?
With "a whole range of aspects" you'll have to decide how to combine and weight them (and/or the underlying "measures") to get a single overall value for each country to allow ranking. That is non-trivial. With a number of "measures" comparable to the number of countries you can choose weightings to get any specified ranking.
As I've said, this is a non-trivial topic. This Wikipedia article gives a flavour.
Are you saying that the comparison is not desired to allow a conclusion that one country is "better/worse" than another? You did say (emphasis added) "the relative state of a country compared to others".
But you haven't said what you do want to do. "Compare" isn't very precise You did say "the .. state of a country" (singular) and "what measures [to] use" (plural) suggesting you did want to combine numbers (measures) to arrive at a single value - maybe non-numeric, maybe multi-dimensional, but you didn't specify.El Pollo Diablo wrote: Well not really. I'm fully aware of how to combine numbers using weighting, it's just not what I said I wanted to do.
And I think "how to combine numbers using weighting" might be more complicated than you think. Creating a weighted sum is trivial, choosing appropriate weightings less so.
Re: Comparing countries
Isn't a thread arguing about which threads are dumbest the dumbest thread? Is there an infinite series of increasingly dumb threads where the next dumbest thread refers to the previous dumbest threads? I'm sure Godel had something to say about, or possibly Cantor.El Pollo Diablo wrote: ↑Wed Jan 11, 2023 9:01 amAnd also, the dumbest thread ever is clearly the coffee one, this is at worst second dumbest
- El Pollo Diablo
- Stummy Beige
- Posts: 3324
- Joined: Wed Sep 25, 2019 4:41 pm
- Location: FBPE
Re: Comparing countries
Yes, I know. I've been involved in international benchmarking of railways, I'm fully aware of the complications involved in pitting one complex multifactorial system with metrics measured in slightly different ways against another complex multifactorial system with metrics measured in slightly different ways.KAJ wrote: ↑Wed Jan 11, 2023 11:58 amAs I've said, "You'll have to define your objective(s) much more precisely to arrive at any cromulent conclusion".El Pollo Diablo wrote: ↑Wed Jan 11, 2023 6:57 amI said "comparing", not "ranking".KAJ wrote: ↑Tue Jan 10, 2023 6:41 pm
You'll have to define your objective(s) much more precisely to arrive at any cromulent conclusion.
With "a whole range of aspects" you'll have to decide how to combine and weight them (and/or the underlying "measures") to get a single overall value for each country to allow ranking. That is non-trivial. With a number of "measures" comparable to the number of countries you can choose weightings to get any specified ranking.
As I've said, this is a non-trivial topic. This Wikipedia article gives a flavour.
Are you saying that the comparison is not desired to allow a conclusion that one country is "better/worse" than another? You did say (emphasis added) "the relative state of a country compared to others".But you haven't said what you do want to do. "Compare" isn't very precise You did say "the .. state of a country" (singular) and "what measures [to] use" (plural) suggesting you did want to combine numbers (measures) to arrive at a single value - maybe non-numeric, maybe multi-dimensional, but you didn't specify.El Pollo Diablo wrote: Well not really. I'm fully aware of how to combine numbers using weighting, it's just not what I said I wanted to do.
And I think "how to combine numbers using weighting" might be more complicated than you think. Creating a weighted sum is trivial, choosing appropriate weightings less so.
Quantitative comparison using arbitrarily chosen weightings for measures is as, well, arbitrary, as qualitative comparison across those same set of metrics. I was deliberately not very precise when saying "compare" because comparison of countries is very not precise, especially when across a wide set of measures.
I'm not looking so much at establishing whether specific country A is better or worse than specific country B according to said arbitrarily chosen measures, as to get a feel for where countries sit relative to others in different fields of interest.
For instance, the USA tends to do pretty well in many economic measures, often coming top (outright GDP being the most obvious), but on many social measures it's rubbish and aligns more with much poorer countries (e.g. intentional homicide rate or religiosity), which is interesting and worthy of discussion.
Similarly, everyone here tends to slag off constitutional monarchies as an undemocratic aberration, despite the fact that of the top 20 countries in the world for democracy index, half of them are constitutional monarchies, and only seven directly elect their head of state (the sense there isn't causative - if any of the major economies which are constitutional monarchies became republics, we would obviously expect their democracies to remain very good, and certainly we don't expect any countries in the world to go from a republic to a monarchy any more, but also it's perfectly possible to have something rated as a strong democracy with an unelected head of state, as most of the "best" democracies in the world demonstrate).
(And, clearly, any metric which implies that it's a bit more complicated than many here have considered clearly has flaws too overwhelming to allow )
Like I say, comparison, rather than ranking.
If truth is many-sided, mendacity is many-tongued