But I agree it's badly worded because their is no new information being introduced in the second trial. The OP should have presented as one test or said the difference the between the first and the second test was that in the first the children can't see the other kids hats and in the second they can or something.
Not at all. The new information in the second test that makes it possible is that they are told "there is at least one red hat".
In the first test, the children aren't told how many red or black hats there are, so you can't use that method of deduction to determine the colour of your hat.
However, if you know that there is at least one red hat and all of the other red hats you can see have said "I don't know" and been eliminated, the only possibility is that your hat is red.
Someone still needs to explain why the first kids couldn't work it out.
The only difference between the situations is that the teacher in the second says "at least one hat is red", but that's completely irrelevant since the kids can already see that there are at least 6 (if they have a red hat) or 7 (if they have a black hat) red hats, so why even tell them that?.
It's a process of elimination that can't be implemented in the first test because the vital piece of information isn't given.
.
.