Square Roots: Estimate, Divide, AverageDate: 02/04/97 at 12:25:05 From: Dorothy L. Mason-Schweitzer & Dan Schweitzer Subject: Square roots Hello! When I went to school about umpteen zillion years ago (pardon the inexact expression, but it was back some time before the pyramids), I recall a longhand-division-and-guess method used to determine square roots, at least until we ran out of papyrus. Nowadays, the only answer I can seem to find is, "Well, you hit this button on the calculator..."; not quite what I'm looking for. Even math teachers only seem to know which buttons to push, including an engineering math teacher in a technical school! Do you know this method, or perhaps know some mathematically inclined archaeologist I could ask? Thank you much! Dan Schweitzer Date: 02/06/97 at 10:50:45 From: Doctor Mitteldorf Subject: Re: Square roots Dear Dan, If you need to find a square root and have only a paper and pencil, the easiest way is probably to Estimate, Divide and Average. Here's how it works. Suppose you want to find sqrt(10). Estimate that the answer is about 3. Divide 3 into 10 to get 3.333333. Average 3 with 3.33333 to get 3.16666667. To get more and more accurate, repeat the process, using your last answer as your next estimate: Divide 3.166667 into 10 to get 3.157894. Average that number with 3.1666667 to get 3.162280. The actual answer is 3.162277, so you're already very close. If you go through another round of this procedure, you should have 5 more correct decimal places. The reason this works is that if y equals sqrt(x), then x/y=y; but if y is a little less than sqrt(x), then x/y will always be a little more than sqrt(x), and vice versa. ---------------------------------------------------------------------- There's another way that's a little harder to understand and to remember. It's based on the formula (x+a)^2 = x^2 + 2ax + a^2. Use the formula with x being the part of the number you've got so far and a as the new digit. We start by writing the square root out like a long division problem, but with no divisor. For an example, I'm going to write ________ /10.00000 because I can't e-mail the right symbol for square root. Just as when you do long division, since you get each decimal place by making an estimate, and sometimes your estimate turns out wrong and you need to erase it and try a bigger or smaller one, you have the same problem in doing square roots in this method. The problem is worse, however, because the estimate you're making is more complex than the one in long division. Here's an example of how to do the square root of 10. Try 3 as your first guess. _3______ /10.00000 Then square the 3 and subtract. Carry down TWO more digits from the top line: _3._____ /10.00000 9 ----- 1.00 Guess the next digit. This is the hard part. We guess "1" because we know it's the right answer; in practice, you often will guess high or low, go through the next step, and guess better the next time. For the number you've got so far, write the "3" as "30" by tacking on one zero. Multiply the new digit by TWICE the number you've got so far (twice 30). 1*2*30=60. Also, subtract the square of the digit (1^2=1). Your next step looks like this _3.1____ /10.00000 9 ----- 1.00 -60 ----- 40 -1 ----- 3900 When you "guessed" the 1, you had to have this whole process in mind. If you'd guessed too high, then the numbers you'd be subtracting would be too large, and your "remainder" would have been <0; this part is just like long division. If you guessed too low, the way you would know is that the number on the bottom line would have one digit too many (in this case, it would be 10000. I've brought down two more digits (zeros) from the top line. I'll show you one more step. The next digit is 6 - found by guessing, or trial and error. 6*2*310=3720, and 6^2=36: _3.1____ /10.00000 9 ----- 1.00 -60 ----- 40 -1 ----- 3900 -3720 ----- 180 -36 ----- 144 Hope that's enough for you to get the idea. -Doctor Mitteldorf, The Math Forum Check out our web site! |
Search the Dr. Math Library: |
[Privacy Policy] [Terms of Use]
Ask Dr. Math^{TM}
© 1994-2013 The Math Forum
http://mathforum.org/dr.math/