Humans struggle with statistical thinking, and behavioural economists have written about this widely. An intriguing way this plays out is in our understanding of societal events. Many fail to view data at an abstracted level and are more easily swayed by compelling narratives.
We easily fall prey to anecdotal evidence from those closest to us.
A recent conversation with an Uber driver triggered this essay. He was convinced that COVID was a conspiracy and that all the deaths were basically "hospital homicides". His evidence was that only people who went to the hospital died, corroborated by his cousin, who thankfully escaped the hospital's clutches. Ignoring the many other problems with this argument (a lesson in Venn diagrams might be helpful), he would not entertain alternative explanations.
Notice too how journalists and media outlets hook our attention with "human interest stories."
For example, you see a campaign to raise funds to feed 100 children suffering from malnutrition in an orphanage. Something in your mind allows you to detach from the high-level problem, and you're less likely to give.
But think of little Anna, who lost her mother and father and will go hungry tonight without your donation to the advertiser. Campaigns like these have been shown to be substantially more effective. (Imagine they documented all 100 human interest stories. You would likely be overwhelmed.)
We need to be aware of our shortcomings in statistical thinking rather than dismissing them as evolutionary flaws.
Before jumping to conclusions, be careful not to extrapolate anecdotal instances or forget about the incidental base rates. Conversely, remember that behind every number, there is a human.