Percentage error, percentage uncertainty, percentage difference.

Can someone clarify, for once and for all, the difference between the three? I'm still a little hazy.

I'd be ever so grateful. :smile:

PS - any other equations it'd be handy to know for an A2 Physics practical, do tell. :smile:

Reply 1

SpikeX

This is what my revision notes look like:

% uncertainty
For a single measurement: (smallest division/your measurement) x 100
For a range of measurments: (range/average) x 100

% difference
For a single measurment: [(real value - your measurment)/real value] x 100
For a range of measurments: (range/median) x 100

Conclusions

- If % uncertainty is less than 5%, your value can be accepted as reliable

- If % difference is less than the sum of your % uncertaintities, the values can be accepted (e.g. for proving T(a) = T(b))

Reply 2

SpikeX

Just thought of an exception for smallest division; when using a timer always use 0.1

Even if your timer displays to 0.01s, still use 0.1 as this is human reaction. (My teacher says this is something you're expected to know and use).

For everything else, just use the smallest division (e.g. 0.001m for a ruler, 0.01mm for micrometer, etc. On digital readings its decimal places as shown on the display)

(Some people 1/2 the smallest division, although I've never done this. Can anybody explain why?)

Reply 3

Epitomessence

Ta very much. :smile:

How do you calculate the combined uncertainty?

Reply 4

Epitomessence

SpikeJx

(Some people 1/2 the smallest division, although I've never done this. Can anybody explain why?)

I think you use half the division if the divisions are clearly quite far apart - like a newtonmeter.

Reply 5

SpikeX

Epitomessence

I think you use half the division if the divisions are clearly quite far apart - like a newtonmeter.

So bascially if you are taking readings between the lines

Reply 6

SpikeX

As for combined uncertainty - just add them up

Although if the question involves some equation where the quantity is raised to a power, multiply the % uncertainty by the power (this usually seems to be in those "show that questions" so it will probably be faily obvious where to use this)

Reply 7

~Adel~

spikeJX, if you were to calculate the percentage uncertainty for which involves a few measurements, do you use range over average x 100 or do you use half of the range(largest minus the smallest measurement)over average x 100?

Reply 8

SpikeX

I use the first method you mentioned, although some people use the second method.

I assume either was is correct

Reply 9

~Adel~

OK, thanks a lot. I read through the past year papers with the answers given by my teacher and found that in most cases, he advised us to use the second method, but he did mark the first method correct in one particular paper.(i think i wrote that accidentally at that time). Good luck with your preparation.

Reply 10

SpikeX

~Adel~

good point, the mark scheme usually say something along the lines of "correct answer from range, 1/2 range, etc. method"

Reply 11

shady zax

Can smn please explain why the experimental value can be accepted when percentage difference is less than percentage uncertainty ?

Im i think u told it the other way around.. 😕

(edited 6 years ago)

Reply 12

Kingdestiny

It is correct because the aim of experiment is to sometimes minimise error. So if uncertainty is becoming note and more small the the precision also increases