Hi all,
I started reading Fractions,Decimals and Percents Fourth Edition manual, and have some serious doubts on "rounding to the nearest digit" problems. Namely, it is said in the book that when rounding to the nearest digit, let's say tenths, you should ignore all other right digits of the right one and only concentrate on that one, namely hundredths digit. If it's five or greater, you should round up, else, round down. I have a serious issue with this...

For example, I would round 2.5489 in the following manner:
Step 1) 2.5489
Step 2) 2.549
Step 3) 2.55
Step 4) 2.6

So, for me, the rounded to the nearest tenth value would be 2.6.

On the other hand, the book says that the value should be 2.5 (just take the hundredths digit, it's 4, and leave the tenths digit as it is, because 4<5). For me, this implies that you are rounding not the number 2.5489, but the number 2.54, and they are not the same numbers... One can not just ignore the rest of the number as if it doesn't exist....

Please, is this some kind of weird GMAT convention I should get used to? Thnx in advance!

Regards,
Srdan

nope.
this is not a "weird gmat convention"; it's the ONLY rule that is EVER used for rounding.

i can guarantee you that you will not find the method you're using here - rounding each consecutive digit, one at a time - as standard in any reputable source, anywhere.
it produces some results that are just plain absurd - for instance, under your system, rounding 1.4445 to the nearest whole number will give 2. that doesn't make any sense (the nearest whole number is clearly 1, not 2).

ALL rounding is done in ONE STEP.

--

if they wanted you to do what you're talking about here, then they would have to spell it out, one step at a time.
i.e., "x is rounded to the nearest thousandth, and then the result is rounded to the nearest hundredth, and then the result is rounded to the nearest tenth"

Open [Microsoft] Excel and type this: =ROUNDDOWN(2.5489,1)
=ROUNDDOWN(2.5489,n) where n = number of places upto which you require the solution after decimal.

--
Umesh

Unfortunately, we won't have Excel to help us out on the GMAT!

_________________
Ben Ku
Instructor
ManhattanGMAT

Hi Umesh!

Thank you for the reply, it is quite clear to me what you want to say, but I was convinced that this explanation is going the wrong way (it is justified as the best way in sense of minimum error of the numerical method, and I hope you agree that the excel software can't be a taken as a proof of a mathematical consistency:-)) ).

I did my research regarding this issue, and have found only a few (unofficial) sources that claim the rounding should be made from the last digit (irrational numbers therefore could not be rounded at all.:-)) ). The conventional and only right method (in sense of best approximation of the number) is surely the standard procedure.

I'm sorry if I made a stupid and rude question, but somewhere in my head I was sure I was taught by this method.:-) Obviously not.:-)

Thank you once more for your help and answer.

Srdan

Also, I'd add that Excel has several rounding functions: Rounddown, Roundup, and Round. GMAT rounding is like the ROUND function.

Beware ROUNDDOWN, Umesh. 1.55 rounded down to the nearest tenth is 1.5, but rounded to the nearest tenth on the GMAT it becomes 1.6.

_________________
Emily Sledge
Instructor
ManhattanGMAT

Srdan,

thank you for asking this question. I'm going through guide 2 fourth edition and had the very same question. Thank you as well to all who replied.

Vivian.

there should really be no confusion here. there is one universally-applied rule for rounding, and the GMAT adopts that rule. just learn the rule and make use of it.. :)

_________________
Tim Sanders
Manhattan GMAT Instructor