[danbooth]

Does anyone know if you round off numbers from a measureing device in experimental data?

I know you round off from a calculation with a calculator.

[Mitch]

Please give an exact question with example, don't ask for vague general help.

[Blueshawk]

Does anyone know if you round off numbers from a measureing device in experimental data?

I know you round off from a calculation with a calculator.

If you do round off, say to the nearest ml for volume...include it in you error analysis as a + or - 2ml.

Something to the sort...but you should try to be as accurate as you can.

But there is always a slight error in measurement.  I dont think anyone can be absolutely 100% sure on a measurement.

As Mitch said...plz be a little more specific.

[danbooth]

Ok, sorry, I was in a hurry this morning and didn't have time.  I hope this is better.


After recording an instrumental measurement, do you off experimental data?

[Blueshawk]

In some cases yes, but it depends on how accurate you want to be.

What are you measuring ??

[tompka]

You can generally assume that the last number on anything that gives you three decimal places is a guess.  If you have a very sensitive piece of equipment that gives 5 or 6 decimal places, the last 2 or so numbers are a guess.  

It depends on what you're doing, and it depends on how much you care.  If it meant the difference between good results and bad results, I'd be more specific, but even then, those last numbers are just a guess.

What exactly are you using?  A spec?  A measuring cylinder?

[Demegorgon]

I have a rounding question.  I work with a chemist who says i round wrong, I have never heard of his riules for rounding can you help PROVE who is right?

He say when rounding to 2 decimals you look at the 3rd decimal 6 and up rounds up 4 and below rounds down if a 5 go to the 4th decimal, and 6 is up 4 is down 5 the next digit .
ie  .5555555 would round to .55         .55556 would be .56  and .555554 would be .55

i was always told 5 rounds up
.55555 would be .56       .5549999 would be .55

[Borek]

He say when rounding to 2 decimals you look at the 3rd decimal 6 and up rounds up 4 and below rounds down if a 5 go to the 4th decimal, and 6 is up 4 is down 5 the next digit .
ie  .5555555 would round to .55         .55556 would be .56  and .555554 would be .55

Never heard of such rounding rule. And I don't like it, sometimes difference between full value and rounded one is same in both cases, but sometimes his rule introduces larger errors than your rule:

number	his rule	his error	your rule	your error
.55554	.55	-.00554	.56	.00446

However, I think I know where the problem emerged. Rounding numbers ending with 5 always up is wrong:

0.25 + 0.55 = 0.80

rounding to one digit:

0.3 + 0.6 = 0.9

Thus some use rule that says - if the preceding digit is odd, round 5 up, if even - round down (or opposite, can't remember right now). So after rounding our sum looks like:

0.3 + 0.5 = 0.8

and the final result is not overestimated in a systematic way.