The abuse of science

In this case it all starts with a screenshot of a poll. Here it is.

I do not know more about this than that. A “science communicator” posts it online. Without a link to the source or the opinion piece / article it come from. I do a bit of searching for it, but come up empty.

EKOS seems to be a Canadian research institute, that does market research and opinion polls. They have no public research listed currently (end March 2022). Their staff is listed on their page. Their director lists their education as an MA, with PhD coursework submitted. That does not inspire confidence in me. I cannot be bothered to look into this further. I would have expected for a head of research institution to be able to complete a PhD, at least. It is not an unreasonable expectation. I agree that they might still be able to excellent research.

The topic of this post is the comments under the screenshot. Let’s break this down first. What do we know (if the science behind data gathering is sound):

  • There is a correlation across COVID vaccination status and attitudes towards Ukraine invasion.
  • … that is it. Data tells us absolutely nothing else.

The sample size is missing. Therefore this could be 15 people in a bar. Probably isn’t but proper scientists disclose their data files particulars.

The comments under the screenshot are roughly divided into

(1) people who defend the position of not getting the vaccine, while disassociating themselves from supporting Russia in the Ukraine invasion.

(2) people repeating Russian fake news and, in essence, saying “...of course it is horrible when Russia bombs maternity wards and bread lines, but, bear with me, this is excusable, because … reasons“. Reasons include (a) NATO building bases in vicinity of Ukraine, (b) the West started the invasion, pretending to be Russia, because they wanted to give weapons away to Ukraine, (c) Russia, by total coincidence, had 200.000 soldiers at the border of Ukraine, and has sent them in when it became clear that the Ukrainian people need to be liberated from the oppressive Zelensky regime, (d) Ukraine has biological weapons, so in order for Russia to defend itself, they had to invade and bomb nuclear power-plants.

(3) People who are conflating correlation with causation. They are variously saying that: (a) no one could expect anything else from anti-vaxxers. They are happy to see our children and economy suffer, so they are happy to see Ukrainian children suffer too. (b) The morons who do not know what is good for them, don’t know what is good the Ukrainians. (c) The anti-vaxxers swallowed Russian fake news when it came to COVID, and now they are swallowing their fake news when it comes to the war. Generally speaking, the implied idea is that not being vaccinated somehow caused the more lenient attitudes to Russia.

Correlation does not imply causation

I tell my students that when ice cream sales rise the frequency of forest fires goes up. It is clear that consuming ice cream causes forest fires. There is usually a confused silence after that.

I then tell them that correlation does not imply causation. In the above case, the rise in outside temperatures (or as we colloquially call it, summer) causes rise of frequency of both events. But one does not cause the other. Much like being unvaccinated does not cause people to be more lenient towards Russia.

This is the first issue I have with the above table. While it does not claim anywhere that there is a causal link, it implies it. There is an expression in stats in the Factor Analysis item creation – Garbage In, Garbage Out, implying that if you use garbage as an input for the analysis, the results will also be garbage. Calculating the correlation across two unrelated things will yield results, but the interpretation cannot be anything but garbage. I am absolutely convinced that one could find other non-sensical correlations too. Say, between the level of toe nail clipping and the purchase frequency of buying pedigree guinea pigs. I can already see the headlines – “Do not clip your toenails, unless you want to suddenly get buried in guinea pigs“.

Scientific misconduct

A proper scientist does not mislead their reader. There is more than one way to do this. The simplest way would be to just lie about it, fabricating the results. That gets discovered sooner or later by the scientific community and some time later by the general public, who historically speaking, now dislike science more, in general. They did not care about it earlier and mostly did not understand the nitty gritty in the first place. But they are told science’s bunk anyway, so now they dislike all science mumbo jumbo.

Another, more common way is to “leave the interpretation to the reader“. This is disingenuous at best. It is like handing a loaded gun to a caveman and saying: “Look this is a gun, it goes boom. Bye.” “I told them what it was, who knew that they would kill their partner by mistake?

This double speak is omnipresent. Slovene politicians keep saying things along the lines of: “I will let you decide by yourselves what these ‘facts’ mean, but it is certainly interesting that [something]“. Why is it interesting? Who is it interesting to? To you or to me? Why will you let me decide on my own? Don’t you have an opinion? Are you too stupid or too ashamed of it? Where did you get your facts? Why are you now sharing them with me? What is your agenda?

While politicians as well as scientists (when communicating with the general population) both manipulate in the same way, the standard for the scientists is in some ways higher. The issue may be that when academics talk amongst ourselves, we tend to indicate facts and/or results of something and expect the listener to connect the dots. However, we are experts in the field. To give you an example – when a fellow academic says to me that 47% of Europeans have been defrauded in the past (true stat, by the way). My response is notBut what does that mean?“, but rather “So, did you find out in what way this number is wrong?“. And their response would certainly not be: “I will leave it to you to decide” if they wanted to have any more conversations with me. I would expect them to say: “I also figured it was wrong. So I read the report where it comes from and their interpretation of constitutes a victim of fraud is just plain crazy.”

Communicating with the public is fraught with peril in this respect. Stating the “facts” and not commenting on them is problematic. Treating the public as if they have enough information to make an informed decision, as an academic would do, is foolhardy and, in my opinion, scientific misconduct. Feel free to communicate in facts and numbers with fellow academics, but do not hide behind them with people. Saying, oh I just presented everyone with this screenshot and I am not responsible for their interpretation of this is the very definition of gaslighting.

About David Modic (admin)

We'll do this later
This entry was posted in ENG, science. Bookmark the permalink.

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.