this post was submitted on 27 Aug 2023
-24 points (16.7% liked)

science

14892 readers
277 users here now

A community to post scientific articles, news, and civil discussion.

rule #1: be kind

<--- rules currently under construction, see current pinned post.

2024-11-11

founded 2 years ago
MODERATORS
m3t00@lemmy.world
Joleee@lemmy.world
laverabe@lemmy.world
DeadPand@midwest.social
-24
Is there a causal relationship between electricity consumption and obesity, or is it just an illusory correlation❓ (cdn.masto.host)
submitted 1 year ago* (last edited 1 year ago) by LabPlot@floss.social to c/science@lemmy.world
17 comments fedilink hide all child comments
 

Is there a causal relationship between electricity consumption and obesity, or is it just an illusory correlation❓

@science @dataisbeautiful @health

The plot and curve fitting made in @LabPlot, a FREE, open source Data Visualization and Analysis software. It works on #Windows, #Linux and #macOS.

➡️ https://labplot.kde.org/download

#Data compiled for 184 countries.

#FOSS #FLOSS #OpenSource #FreeSoftware #DataViz #Visualization #Obesity #Health #ClimateChange #ClimateCrisis #Climate #Food

all 21 comments
sorted by: hot top controversial new old
[–] ILikeBasil@lemmy.world 8 points 1 year ago

No

[–] Diabolo96@lemmy.dbzer0.com 7 points 1 year ago* (last edited 1 year ago)

I don't even know what the graph results are supposed to mean. The Lower consumption side is too jumpy and goes in the two extremes.

[–] edwiebe@mstdn.ca 6 points 1 year ago (2 children)

@LabPlot @LabPlot @science

One can fit a curve to anything. One should not.

https://blog.revolutionanalytics.com/2017/05/the-datasaurus-dozen.html

[–] LabPlot@floss.social -2 points 1 year ago* (last edited 1 year ago) (1 children)

@edwiebe @science

We agree. But still, a question is just a question, and you can always refine your questions.

Matejka, J., & Fitzmaurice, G. (2017). Same Stats, Different Graphs: Generating Datasets with Varied Appearance and Identical Statistics through Simulated Annealing.

BTW, the Datasaurus Dozen example is already available in @LabPlot via File > Open Example.

[–] LabPlot@floss.social -1 points 1 year ago

@edwiebe @science

If you are interested, please see also this thread on the importance of visualizing data (the Anscombe's quartet, Simpson's paradox are also included in @LabPlot):

https://mstdn.social/@onemoment/109692198312380103

#Anscombe #SimpsonsParadox #DatasaurusDozen #Visualization #DataViz

[–] ottaross@mastodon.social 5 points 1 year ago

@LabPlot @science @dataisbeautiful @health seems like a classic graphic for illustrating the wrong variables being compared. To stick a trend line in the there should result in a loss of coffee room privileges for a week.

[–] neptune@dmv.social 5 points 1 year ago

Statistics teaches how to measure the strength of a correlation. Also, science can help us understand cause VS correlation. You can't just do a curve fit and assume it means something.

[–] Spuddaccino@reddthat.com 5 points 1 year ago (1 children)

The answer is... kind of, but only really at the lower end.

Countries with very low (around 0) electricity usage are going to be places where food refrigeration is hard to come by, if even possible, and so stockpiling and transporting food becomes more difficult. These places, then, have to grow or hunt their own food, and it's often just enough to get by, especially considering how much hard work goes into it.

Once electricity becomes more prevalent and food refrigeration becomes common, people tend to be a bit freer with their food consumption. This doesnt mean that they all turn into fat slobs, but it does mean that they have the the option to do so that didn't exist before.

Once you hit that threshold, you start to notice things spreading out on the chart, whereas there are basically no obese countries at 0 kWh, outside of a few outliers. I'm kind of curious about which countries are up there at 45% obesity rate and no electricity.

[–] LabPlot@floss.social -1 points 1 year ago (1 children)

@Spuddaccino

For example: Tonga, Samoa, Kiribati, Nauru with electricity consumption per capita (the median) 548 kWh.

[–] Spuddaccino@reddthat.com 2 points 1 year ago

Ah, that explains a lot, then.

These are all island nations in Oceania that receive large amounts of their food supply from outside the country. This offloads much of the energy cost of refrigeration onto whatever nation owns the ship. I don't know if there's a good way of figuring out how much energy is spent shipping supplies to those countries, though.

[–] Hazdaz@lemmy.world 4 points 1 year ago (1 children)

With this kind of cherry-picking and manipulation of data, you have a bright future in the news room of Fox News.

[–] LabPlot@floss.social -3 points 1 year ago* (last edited 1 year ago) (1 children)

@Hazdaz @science

Is the act of distinguishing a question from an answer as difficult as recognizing spurious correlations?

The question has been raised earlier by others,. See for example this paper from 2021 (Measuring the effect of energy consumption on the epidemic
of overweight in Latin America and Caribbean countries):

https://dialnet.unirioja.es/descarga/articulo/8100043.pdf

[–] JoBo@feddit.uk 3 points 1 year ago

You've fitted a curve through a bunch of points and called it correlation. It is not.

If you want to do this properly, you've got a lot of reading to do.

[–] JoBo@feddit.uk 4 points 1 year ago

What correlation?

[–] WhoRoger@lemmy.world 1 points 1 year ago

Consumerist lifestyle of (some) richer countries leads to both of these, tho I don't see any direct relation. I guess obese people may be more likely to invest in a bigger TV, may turn up the AC more, and spend more time in hospitals hooked to machinery?

Generally speaking, electricity is mostly used up by factories and businesses.

[–] cuteprince@mastodon.social 1 points 1 year ago

@LabPlot @science @dataisbeautiful @health I mean... A bit? But the comparison does come across as "computers make you fat and lazy" which feel a bit aimless.