To think this qualifies as bad science?

[tamade]

The graphic below is from a Lancet article about the relative impact of cold or heat on excess death rates. Look at the scale(s) at the bottom and tell me that they were not deliberately chosen to mislead the casual observer.....

YABU climate is so important the ends justify the means
YANBU it is a bit dodge and all scientists should hold themselves to better standards
bonus YABVU its obviously an honest mistake and it makes the graph look better: stfu

http//doi.org/10.1016/S2542-5196(23)00023-2

The stats would say what they say, if they factually wrong it is wrong if not people should use the brain cells they were born with and not rely on being spoon-fed everything

I'd suspect anyone reading the lancet to be capable of reading a basic graph

@PuffingPuffin
Hopefully, yes.

But that's not really the point; the purpose of any figure should be to aid the interpretation of the raw data, this one actually distorts the magnitude of the two measurements being compared by a factor of 5.

So just because some or even most people reading the article could note the scales does not mean it should have been done.

YABU for starting an inflammatory thread like this when the problem is you. It's not really the authors responsibility to teach you how to read a graph.
Or the article itself for that matter, the study isn't about comparing the impact of cold vs heat...

"In this study, we did a comprehensive mortality impact assessment due to heat and cold in European urban areas"

.........

Did you actually read the article?
It's about measuring death rate between cities and age groups, that's the comparison they are referring to. It's very clear from the introduction and through the entire article

I have read the article and I think that you are the one who does not understand it. It is about the effect of temperature on excess mortality but the researchers have broken down the data by age and region and this allows them to observe and discuss points such as how Scandinavia and the Mediterranean countries may be more robust when it comes to lower or higher temperatures respectively. Or that the elderly are more at risk from either heat or cold.

I have re-checked the link and we did read the same thing so I dont know what else to say.
Oooh yes;
Apart from to point out that you are making a strawman argument anyway, the question is whether the using a scale which visually magnifies heat related excess mortality by a factor of five on a bar chart.

It is bad practice to change the scale on a graph axis.
That's basic GCSE maths.
They should have used 2 separate graphs.

It’s fine I think. The graph conveys the point it needs to make and no, I don’t think it’s misleading to have different scales for heat and cold related mortality rates as I think it’s clear from caption and text. Lancet has a very involved editorial process as well as tough peer review so it’s not like you can sneak something like this past them unnoticed somehow.

So if you think they are lying have you contacted them about it to have it fixed

That’s a spectacularly shit graph. Nowhere on it is ‘heat’ or ´cold’ defined so I have no idea what it’s meant to be saying anyway. Presumably it’s slightly clearer in context. The change of scale is unforgivable. And the ´total’ for each region stuffed into the graph at 4 separate points with no divisions is weird and another example of bad formatting.

This is a terrible graph, mainly because of the distorted scales.

We should want science to be as accessible as possible to all — including those who don’t, or can’t, analyse a graph carefully. Regardless of which way the data points in any given case.

For those saying it’s fine: would you be happy if the distortion was reversed (to minimise heat deaths)?

To me, distortion bad, regardless of which political message it supports.

Also — the reasons to be alarmed about global warming are not mainly about heat deaths in the present, but the medium/long term future beyond the point of no return, and the many present-day impacts of climate change that are not heat-deaths. Distorting the message makes it easier to dismiss, and is bad tactics as well as bad science.

I agree it’s a bad graph.

I agree OP. To my mind the whole point of a graph is a quick visual representation which this isn't due to differing scales! Yes I can read it and understand it but it definitely doesn't feel right!

I see your point, but I also see why they did it this way. There is some value in seeing the cities together so you don't have to flit between two graphs, and I think it's pretty clear personally - a consistent scale would've made age ranges indiscernible.

It's known as a 'gee-whizz' graph. And yes OP I agree that the visual representation is deliberately misleading to make the point the author wants to bang home.

I agree OP. Articles like this are often framed in a certain way n order to generate media attention and headlines.

But you are using a distorted scale to conclude on how "bad science" tries to influence public opinion, when there is 0 evidence of that in the article.

But if you like making up conspiracy theories, have it your way!

Not a scientist or a remotely science y person but I do work as a (relatively new to the field) Data Analyst I'm I'm pretty sure there would be conversations if I presented data in this way at work. The entire point of data visualisation, rather than in a table or body of text is to make trends etc immediately obvious almost at a glance.

Also, there are other charts they could have used which would give the same effect without the need to change the axis in this way.

@HoneyIShrunkThePizza
“a consistent scale would've made age ranges indiscernible.“

That could be a valid reason to do it I suppose, but if I was most concerned about showing that aspect clearly I would chose consistent colours for age group in the stacked bars, rather than shades of red or blue.

@Berklilly
I have made an observation and come up with a theory which fits the facts. You have not offered a better theory neither have you offered a fact would make my theory inconsistent with the world.

Those age groups for 'Total' do not make any sense

Seems reasonable to me as the focus is presumably on national differences, although it would be helpful to see the caption too. Plot the graph the way you think it should be and see how insightful you find it.
Heat and cold are location specific thresholds defined for each place, and the intended audience will know that. Probably you can find a lay summary of the paper, but I wouldn’t expect every graph to be accessible to the general public.

Completely disagree. The graph allows comparison between different cities and age groups. The numbers are such that, if you used the same scale for heat as for cold, those differences would not be discernable. This is in the Lancet, not the Daily Mail - the readers should be perfectly capable of interpreting this graph correctly.

Sorry, countries, not cities...

The entire point of data visualisation, rather than in a table or body of text is to make trends etc immediately obvious almost at a glance.

Exactly and the altered scale on one side is intended to give a distorted impression of excess deaths due to heat. This kind of distortion happens regularly in tabloid newspapers I don't expect to see it in the Lancet.