I think it also depends on your background or coding style.
FP languages are often completely immutable so you'll fall into this pattern naturally. It's kinda hard to create a half-initialized thing and populate it later when you aren't allowed to mutate anything.
On the other hand, OO languages are much happier to let you initialize an object after it has been constructed, especially because object references can be null by default and they have no qualms about mutability... You could argue this contributes to why OO languages have a propensity for crashing due to an unexpected NullPointerException.
Don't take this all as gospel, though. It's just a way of coding, and there are plenty of places where it's the wrong way to code (e.g. because it unnecessarily constrains your code or information just isn't available until a later part of the program).