Submitted by marshmallow on Mon, 03/27/2017 - 18:18
In this binomial expansion question, I was able to get the fraction approximation for i)
However I am not clear on why the error is given by the first neglected term and how I would use this to find the error of my approximation of root(11)?
I got the first neglected term as -1/16 (x)^3, and I subbed in x=1/100 into that term. I am not sure if that is the error or whether I need to use that term divided by approximately 80000/24000 to get the error in this approximation?
Thank you
The reason why the error is
The reason why the error is approximately the first neglected term has to do with the theory of power series and their convergence, an undergraduate topic that you aren't meant to know: hence the question telling you to assume it.
You already know that $\sqrt{1-x} \approx 1 - \frac{x}{2} - \frac{x^2}{8}$ with the next term being $-\frac{x^3}{16}$.
You use the first three terms to evaluate $\sqrt{11}$ by first evaluating $\sqrt{\frac{99}{100}} = \frac{3}{10}\sqrt{11}$.
So the error in evaluating $\frac{3}{10}\sqrt{11}$ is $-\frac{x^3}{16}$ with $x=\frac{1}{100}$.
But that means the error in evaluating $\sqrt{11}$ is the error in evaluating $\frac{3}{10}\sqrt{11}$ divided by $\frac{3}{10}$. After all, the multiply a quantity that has an error attached multiples the error along with it. Hence the necessity for division here.
This works out to be $$\displaystyle \frac{-\frac{x^3}{16}}{3/10} = -\frac{10}{3}\cdot \frac{x^3}{16} \, \mathrm{with} \, x=\frac{1}{100}$$
I'm not sure why you're using $8000/24000$ instead of just $10/3$ or why you think it's 'approximate'. You need to divide by exactly $10/3$ to recover the approximate error in $\sqrt{11}$ from knowing the approximate error in $\frac{3}{10}\sqrt{11}$. I hope this makes sense, if not - please let me know and I'll do my best to clarify.
Thank you!
This is very clear and helpful.
For some reason I was confused and was trying to use 8000/24000 because I got the approximate of root(11) as 79599/24000.
I now realize that I don't need to use that value at all when calculating the error as I know what I multiplied the expansion by.
Thank you again :)
Awesomesauce, glad you
Awesomesauce, glad you thought it was helpful. :-)