I'd explain it this way:
Pretty much all math, from basic 2+2=4 up to differential calculus, is based on the addition of numbers. To understand addition, you have to understand what is actually being done to the two values in question.
Take 2 and 2. Both of these are magnitudes of an arbitrary unit. They could represent houses, planes, or internet geeks who post too much on silly topics. The best way to conceptualize what is happening when you add two and two, is to draw a number line. On the number line, you start with your pencil tip at zero, and draw a line from zero to two. Your new origin is 2 now and you go two more to the right (default direction). The total displacement is the answer, and you have traveled across 4 units.
When you want to move to subtraction, the concept of direction comes in and it can be taught a number of different ways. You can introduce negative numbers and positive numbers, and show that negative is the left direction and positive is the right direction.
So if you want to subtract 1 from 2, (2-1), you start at 0, draw a line from 0 to 2, and then draw a line ONE unit in the negative direction from 2, or left from 2. Your total displacement from 0 is 1 unit.
This is absolutely the same as saying (2+(-1)), as the sign only indicates the direction you travel on the number line. The important thing is the value, and the total displacement. On the number line, you cannot have a negative distance traveled (as you can't have a negative length of your line from 0 to your end point), but you can have a distance traveled in the negative or positive direction.
So when someone says -5, and you look at it with respect to the number line, it is simply saying that you travel five units in the negative direction. The negative sign tells direction, and the 5 tells distance. The two are clearly separate entities.
This becomes an issue in multiplication (which is really fancy addition). So, you have 2*5. Start on the line at 0, and go five in the positive direction, twice. Total displacement is 10 units.
Now try -2*5. You can do this another way, by going 2 in the negative direction 5 times, or 5 in the negative direction twice. Clearly, the direction is detached from the values.
So now you have (-5)^2. You factor it out to (-5)(-5) and it looks confusing because you have two direction components. But just like double negatives, they cancel each other out because they are SEPARATE from the numerical values themselves. Two neg's make a positive, and you have (5)(5) and you go 5 units in the positive direction five times.
So, basically, it should be obvious that the sign and the value are separate entities UNLESS they are explicitly grouped together.
-5^2 is basically the negative direction of the result of 5^2, or (5)(5).