The fact that the height of adult males with ADHD is greater than that of males without ADHD (first figure/graphs, G & H) should indicate that there is no causative link that the medication has an impact. For females, the average does not show this, but if you look carefully, the variation is bigger. The same selection bias is possible in this case.
Instead, to me it looks like there is selection bias going on. Shorter males are more likely to get MPH prescribed is an equally valid explanation.
Luckily, the conclusion makes no assertions about height either, nor does the paper make conclusions about any causative relation.
I find this especially surprising because line 2 may be exactly the reason why line 5 is unreachable. E.g. if puts("A") contractually throws an exception you cannot just remove it.
I'd say that the needing rewrites is bad for the future clarity of the code. So much so that people conclude the 50% performance hit is worth it to use open polymorphism (interfaces, etc.) over closed (tagged unions, algebraic data types, etc.). So I optimize for the ability to reason about over the lifetime of the project above the current ability to reason and above performance (with exceptions).
What is adding a polygon going to do? Either back to the 'bad' interface to hide variable object size, extend to a tagged union with a size of the biggest datastructure - hurting cache performance each time it grows -, or an involved Object of Arrays structure - not good for clarity but great for performance.
All while having to remember which field, width or height, to use for the circles radius.
my 2 cents: I don't see algebraic data types as strictly superiour. It's just the other side of the polymorphic coin: there is open and closed ploymorphism. Open polymorphism happens with interfaces, inheritance, and typeclasses- the number is unlimited. Closed happens with ADTs - an enumeration of the cases.
Open is great for extensibility: libraries can be precompiled, plugins are possible. Changes don't propagate - it is ""forward compatible"" which is great for maintaning the code.
Closed on the other hand is great for matching. Finite is predictable & faster. Finite is self-contained and self-describing because it exposes the data types without shame.
The purpose of the visitor pattern is now clear: it closes the open polymorphism for a finite set. Great, now we only need one kind and we still get matching. Or is it the worst of two worlds? Slow & incomprehensible and all changes propagate everywhere.
So which one is better? Neither. But the reality is that all old imperative languages with polymorphism chose the open kind - the kind that adds more features because it was needed for shared libraries. Leaving you to build any pattern matching yourself and to burn yourself with the unmaintainable code.
If people get burned, they learn. First they say don't do that and only then they replace the gas stove by induction.
The analogy with sexuality and gender bothered me. It is a sore point and fragile. The solution for them is (foremost) equality, the solution for people with disabilities is egality.
Yes, society changed to help people as much as possible. With changes your living-mate can interact with society as well as possible. Sure, the missing thumbs are disabling, but somebody insisting on hand-written instead of typed notes is an asshole.
At some point, people who ignore a disability or disorder because of general ignorance or worse are the problem.
It is the beginning of remote-learning science. Anecdotes allow estimations of impact and the creation of hypotheses. If nobody talks about a problem, does it exist?
Instead, to me it looks like there is selection bias going on. Shorter males are more likely to get MPH prescribed is an equally valid explanation.
Luckily, the conclusion makes no assertions about height either, nor does the paper make conclusions about any causative relation.