Showing posts with label proof. Show all posts
Showing posts with label proof. Show all posts

Thursday, December 20, 2012

Junior's first number theory proof

Back in my undergrad days, when I was taking a number theory class, I noticed that all squares that weren't multiples of three were always one more than a multiple of three, never one less.

16 = 15 + 1

25 = 24 + 1

100 = 99 + 1

After I thought about it for a while I realized that this had to be the case and there was a simple algebraic proof that showed it (there's a also a simple geometric proof -- think about cutting the square into two smaller squares and two rectangles -- but that can wait for another day).

I'd assign this with the following hint:

think about (x + 1)(x + 1) and (x - 1)(x - 1)

Here's the proof. All natural numbers can be written as:

I. 3k

II. 3k + 1

III. 3k - 1

We're talking about squares that aren't multiples of three so we can skip the first case and look at II and III.

(3k + 1)(3k + 1) = 9k^2 + 6k + 1 = 3(3k^2 + 2k) + 1

(3k - 1)(3k - 1) = 9k^2 - 6k + 1 = 3(3k^2 - 2k) + 1

I like this problem for a few reasons.

First, it's simple. The question is easy to state and the proof is at the right level for a beginner.

Second, it gives the student a chance to do something interesting with polynomials.

Third, it demonstrates an important problem solving technique -- breaking the problem down to cases.

Fourth, it introduces number theory and it gets the students thinking about numbers in a different way.