Habit of Mind 4: Analyzing Answers, Problems, and MethodsChecking and analyzing answers. When we were kids, checking meant only doing the problem again, or working backward from the answer—checking a subtraction with an addition, for example. Checking has another, richer meaning—reviewing an answer for reasonableness. You can tell when students are trying to make sense when you hear chains of thought like these:• I added two numbers, but my answer is less than one of them, so something must be wrong.• My answer is three-and-a-third busses. That makes no sense!• The area of rectangle A is 24 square inches, and I got 50 square inches for rectangle B’s area. This looks OK because B looks much bigger. Oh, and if I place A here on top of B, there is about enough space for another copy of A next to it, so B is about twice as big as A. My answer seems correct.Students must see that even where computations are required, common sense remains central.Tinkering with the problem. When a student has found, justified, and made sense of an answer, it’s time to step back and see how the whole problem (statement, solution, and answer) fits in a bigger picture. Can other problems be solved with the same method?How can the answer to this problem be applied elsewhere? Can this problem be extended or generalized to address new situations? This, in turn, leads to one of the most powerful ideas of all: that even (in fact, especially!) in mathematics, one can ask what-if questions. The what-if has to be taken seriously to see what its consequences really are. But only as children see that this is ok—that asking what-if questions is a genuine part of the “mathematical game”—can they feel completely free to suggest such questions themselves (see Goldenberg & Walter, 2002; Brown & Walter, 1990).Creating and analyzing algorithms. There is, today, intense debate about whether children should be learning standard algorithms or making up their own. We’re inclined to think that the reason the debate rages on is that there’s wisdom on both sides: neither answer is sufficient by itself. On the one hand, a standard algorithm6 gives students a general method for solving certain kinds of problems. On the other hand, the process of inventing approaches of their own involves students in analyzing the problems and answers and in developing the kinds of ideas—generally quite algebraic (though without the specialized notation)—that may help them understand and appreciate a general algorithm. In fact, creating a good algorithm requires reflecting on the steps required for solving a problem, then generalizing these steps in a way that can solve a class of similar problems. Each approach contributes something important.These reasons for standard and student-generated algorithms are given from the teacher’s perspective. There is also something we want students to realize as they work6 “The” standard algorithm is quite another story. Algorithms vary from society to society, and some are more efficient than the ones we call standard. For example, the addition algorithm we typically call standard in the US works from right to left, and is safe and efficient on large columns of multidigit numbers. If only two or three multidigit numbers are to be summed (the much more usual situation facing people now) a left-to-right approach is equally efficient, makes mental computation easier, and ties in more closely with other mathematical ideas and curricular goals such as estimation and rounding.