BI Duck Posted February 26, 2011 Report Share Posted February 26, 2011 Hi, I have two questions1.- So I am wondering if there is a way by which the uncertainty is equal to the absolute error?2.- During a lab report, in which I measured the lenght of a piece of wire, surprisingly all of my classmates and I registered 1.67 m exactly! (with a flexometer)Would it be wrong if I wrote in my report no absolute error? X_X Reply Link to post Share on other sites More sharing options...
nametaken Posted February 26, 2011 Report Share Posted February 26, 2011 Hi, I have two questions1.- So I am wondering if there is a way by which the uncertainty is equal to the absolute error?2.- During a lab report, in which I measured the lenght of a piece of wire, surprisingly all of my classmates and I registered 1.67 m exactly! (with a flexometer)Would it be wrong if I wrote in my report no absolute error? X_X1. I'm not completely sure. Uncertainty is usually defined by the absolute error.2. I don't think it'd be wrong. Reply Link to post Share on other sites More sharing options...
Keel Posted February 27, 2011 Report Share Posted February 27, 2011 (edited) Hi, I have two questions1.- So I am wondering if there is a way by which the uncertainty is equal to the absolute error?2.- During a lab report, in which I measured the lenght of a piece of wire, surprisingly all of my classmates and I registered 1.67 m exactly! (with a flexometer)Would it be wrong if I wrote in my report no absolute error? X_X1. The random uncertainty (in the form of a +/- value or a +/- precentage) will only be the same as the absolute error if the data has not been combined. E.g. Volume Z: 10 +/- 1 (+/- 10%)Volume X: 10 +/- 1 (+/- 10%)Volume Z+X= 20 +/- 2 (+/- 10%)---------No change in % error but a change in the absoluteDistance: 100 m +/- 1 (+/- 10%)Time: 20 sec +/- 1 (+/- 10%)Speed = distance / time = 5 m/s +/- 20% -----you add the % errors when dividing or multiplying errors The error documents here might help:2. You would be absolutely wrong if you stated there was NO ERROR. Just because everyone got the same answer it doesn't mean there is not absolute error in the equipment, it just means that human error was unlikely. Your random uncertainy should always be half of the smallest unit on the piece of equipment you are using. I would presume +/- 0.005 m in the case of the flexometer. Edited February 27, 2011 by Keel Reply Link to post Share on other sites More sharing options...
Bob Posted February 27, 2011 Report Share Posted February 27, 2011 Hi, I have two questions1.- So I am wondering if there is a way by which the uncertainty is equal to the absolute error?2.- During a lab report, in which I measured the lenght of a piece of wire, surprisingly all of my classmates and I registered 1.67 m exactly! (with a flexometer)Would it be wrong if I wrote in my report no absolute error? X_X1. The random uncertainty (in the form of a +/- value or a +/- precentage) will only be the same as the absolute error if the data has not been combined. E.g. Volume Z: 10 +/- 1 (+/- 10%)Volume X: 10 +/- 1 (+/- 10%)Volume Z+X= 20 +/- 2 (+/- 10%)---------No change in % error but a change in the absoluteDistance: 100 m +/- 1 (+/- 10%)Time: 20 sec +/- 1 (+/- 10%)Speed = distance / time = 5 m/s +/- 20% -----you add the % errors when dividing or multiplying errors The error documents here might help:2. You would be absolutely wrong if you stated there was NO ERROR. Just because everyone got the same answer it doesn't mean there is not absolute error in the equipment, it just means that human error was unlikely. Your random uncertainy should always be half of the smallest unit on the piece of equipment you are using. I would presume +/- 0.005 m in the case of the flexometerumm I wouldn't say you're a hundred percent correct about the concept of an absolute errorIt should be something like the difference between the measured value of a quantity and its actual valuefor example, i measured a banana for its length with a ruler with uncertainty ±0.5cmIt is measured to be 25cmbut then it is said that the banana is actually 30cm!even although the uncertainty is ±0.5cm or 0.5/25=2%the absolute error is 30-25=5cm or (30-25)/30=17%search for 'absolute error' on google, you would find definitions easilyso the answer is yes both questions,1with the same ruler and the same banana,you measure it and you get a value of 30.5cmin the case both the uncertainty and the absolute error is 0.5cmalthough the percentage values differ (0.5/30.5 and 0.5/30)2you get no absolute error only when the measured value and the GIVEN VALUE equalin your case there is not a given value so you can't say that there is no absolute error, but there is no way to determine an absolute error as well Reply Link to post Share on other sites More sharing options...
Keel Posted February 27, 2011 Report Share Posted February 27, 2011 (edited) for example, i measured a banana for its length with a ruler with uncertainty ±0.5cmIt is measured to be 25cmbut then it is said that the banana is actually 30cm!even although the uncertainty is ±0.5cm or 0.5/25=2%the absolute error is 30-25=5cm or (30-25)/30=17%search for 'absolute error' on google, you would find definitions easilyI must disagree with you and protest against your statements with the strongest of terms!An 'Absolute error' is the error which is on the piece of equipment eg. a ruler has a absolute error of +/- 0.05 cm.A 'random uncertainty' is the inaccuracy created as a result due to the accumalation of absolute errors. ie 25 cm has a absolute error of +/-0.05 this can also be converted into a precentage 25 cm +/- 0.2%. therefore the random uncertainty of the result is 0.2%If the banana is truely 30 cm, that is a literature value!The error is therfore NOT in the absolute errors or random uncertainty but in SYSTEMATIC ERRORS.What you have done is found the % difference between the obtained value and the literature value, which is a method for determining how much of the inaccuracy was due to systematic errors. e.g. Obtained value: 25 cm +/- 0.2%Literature value: 30 cm% difference between obtained a lit values: (25-30)/30 x 100%= -16.7% ie the obtained value was 16.7% less than the lit valuetherefore, error due to systematic errors = total inaccuracy - total random uncertainty = 16.7 - 0.2 = 16.5%therefore, systematic errors account for 16.5% of the experiment and random uncertainties accounted for 0.2%Edit: I URGE you once again to have a look at the error documents: Edited February 27, 2011 by Keel 1 Reply Link to post Share on other sites More sharing options...
Bob Posted February 27, 2011 Report Share Posted February 27, 2011 for example, i measured a banana for its length with a ruler with uncertainty ±0.5cmIt is measured to be 25cmbut then it is said that the banana is actually 30cm!even although the uncertainty is ±0.5cm or 0.5/25=2%the absolute error is 30-25=5cm or (30-25)/30=17%search for 'absolute error' on google, you would find definitions easilyI must disagree with you and protest against your statements with the strongest of terms!An 'Absolute error' is the error which is on the piece of equipment eg. a ruler has a absolute error of +/- 0.05 cm.A 'random uncertainty' is the inaccuracy created as a result due to the accumalation of absolute errors. ie 25 cm has a absolute error of +/-0.05 this can also be converted into a precentage 25 cm +/- 0.2%. therefore the random uncertainty of the result is 0.2%If the banana is truely 30 cm, that is a literature value!The error is therfore NOT in the absolute errors or random uncertainty but in SYSTEMATIC ERRORS.What you have done is found the % difference between the obtained value and the literature value, which is a method for determining how much of the inaccuracy was due to systematic errors. e.g. Obtained value: 25 cm +/- 0.2%Literature value: 30 cm% difference between obtained a lit values: (25-30)/30 x 100%= -16.7% ie the obtained value was 16.7% less than the lit valuetherefore, error due to systematic errors = total inaccuracy - total random uncertainty = 16.7 - 0.2 = 16.5%therefore, systematic errors account for 16.5% of the experiment and random uncertainties accounted for 0.2%Edit: I URGE you once again to have a look at the error documents: Is it something to do with the wording? I study in Hong Kong, and I believe it is the same in the UK, that absolute error is (quoted from wikipedia) the magnitude of the difference between the exact value and the approximation.and ruler does not and will never have an absolute error - if it does, and the error is 'absolute', why don't we correct it by cutting a bit off its tip? Reply Link to post Share on other sites More sharing options...
Drake Glau Posted February 27, 2011 Report Share Posted February 27, 2011 Keel right on the absolute error. The absolute error is the error due to the precision of the instrument. A rule can only measure to 1mm precision so the absolute uncertainty would be 0.5mm or 0.005m or 0.05cm (as he said). This is to ensure that the measurement you recorded, with your eyeballs, includes that part you can't measure and by having a 0.5mm uncertainty it accounts for the possibility of the true length being a little in between the mm lines that mark the ruler. Hope this made sense 1 Reply Link to post Share on other sites More sharing options...
Bob Posted February 27, 2011 Report Share Posted February 27, 2011 I still hold my opinion... an absolute error is 'absolute' because it is the actual amount of error in a calculation.yes it is related to the precision of the measuring apparatus, but that is when you ignore the accuracy of ita ruler usually has a constant uncertainty.if and only if the measurement is accurate, then the uncertainty is the absolute error.And lol, I did a quick search on google just then and found that all definitions similar to Keels is a chemist one... hmmmaybe I am wrongand, wikipedia is not always wrong, especially when it comes to the sciences Reply Link to post Share on other sites More sharing options...
Drake Glau Posted February 27, 2011 Report Share Posted February 27, 2011 (edited) Yea, first off I said "error" in my post, what my post describes is absolute uncertainty, sorry. As far as I know there is no "absolute error" simply because there is no "absolute value" that you can wrong from. There is however the % error based off accepted values that have been repeatedly tested and hold true or are an average that is now accepted by most people Edited February 27, 2011 by Drake Glau 2 Reply Link to post Share on other sites More sharing options...
Bob Posted February 27, 2011 Report Share Posted February 27, 2011 Yea, first off I said "error" in my post, what my post describes is absolute uncertainty, sorry. As far as I know there is no "absolute error" simply because there is no "absolute value" that you can wrong from. There is however the % error based off accepted values that have been repeatedly tested and hold true or are an average that is now accepted by most people Yes this is what I was trying to say, thank you. I am not sure about not having an absolute error; but obviously it can't be determined without having a given valueIMO it depends on the nature of the measurement... Reply Link to post Share on other sites More sharing options...
Drake Glau Posted February 27, 2011 Report Share Posted February 27, 2011 Yea, first off I said "error" in my post, what my post describes is absolute uncertainty, sorry. As far as I know there is no "absolute error" simply because there is no "absolute value" that you can wrong from. There is however the % error based off accepted values that have been repeatedly tested and hold true or are an average that is now accepted by most people Yes this is what I was trying to say, thank you. I am not sure about not having an absolute error; but obviously it can't be determined without having a given valueIMO it depends on the nature of the measurement... Mhmm that's why our teachers constantly tell us to say how/why we chose our uncertainties. Always good to include this when you're doing your sample problem in your DCP for the P part And yes, it matters for points Reply Link to post Share on other sites More sharing options...
dessskris Posted February 28, 2011 Report Share Posted February 28, 2011 I still don't get how to deal with error propagation...What I know is how to deal with the precision of the values of variables we have.When adding&substracting, the answer needs to have the same no of decimal places as the numerical value with the fewest decimal places.When multiplying and dividing, the answer needs to have the same no of sig fig as the numerical value with the fewest s.f.Do the same rules apply in propagation of uncertainties?When adding&substracting the numerical value, the absolute unc are just added up, right? Then do we round the answer up to the least d.p. or to the least s.f.?When multiplying and dividing, the percentage unc are just added up, right? Then do we round the answer up to the least d.p. or to the least s.f.?When converting from absolute unc into percentage unc, do we round the answer up to the least d.p. or to the least s.f.? I don't get it Reply Link to post Share on other sites More sharing options...
Keel Posted February 28, 2011 Report Share Posted February 28, 2011 I still don't get how to deal with error propagation...What I know is how to deal with the precision of the values of variables we have.When adding&substracting, the answer needs to have the same no of decimal places as the numerical value with the fewest decimal places.When multiplying and dividing, the answer needs to have the same no of sig fig as the numerical value with the fewest s.f.Do the same rules apply in propagation of uncertainties?When adding&substracting the numerical value, the absolute unc are just added up, right? Then do we round the answer up to the least d.p. or to the least s.f.?When multiplying and dividing, the percentage unc are just added up, right? Then do we round the answer up to the least d.p. or to the least s.f.?When converting from absolute unc into percentage unc, do we round the answer up to the least d.p. or to the least s.f.? I don't get it A common protocol is that the final total percent uncertainty should be cited to no more than one significantfigure if it is greater than or equal to 2% and to no more than two significant figures if it is less than 2%.You do not need to round your errors during the calculation. It is the final answer which needs to follow the rule above.So state all your observed values to the rules you have stated your self, apply the calculation (like in maths) with no rounding) but round the final answer. Reply Link to post Share on other sites More sharing options...
Tilia Posted February 28, 2011 Report Share Posted February 28, 2011 If I haven't forgotten this completely, we were told to give the final uncertainty without decimals. Reply Link to post Share on other sites More sharing options...
Keel Posted February 28, 2011 Report Share Posted February 28, 2011 If I haven't forgotten this completely, we were told to give the final uncertainty without decimals.If the uncertainty is equal to or greater than 2% you will give it to 1 sig fig (eg. 12.3% becomes 10%) if it is less than 2% you will give it to 2 sig fig (eg. 1.296% will become 1.3%) Reply Link to post Share on other sites More sharing options...
blskye Posted March 1, 2011 Report Share Posted March 1, 2011 Question! so I'm a litle confused on the uncertainty of a stopwatch b/c some people were saying that its uncertainty is ±0.1s b/c human reaction is that speed, but aren't you supposed to account for the uncertainty of the device instead? (which is supposed to be ±0.01s); which is right? Reply Link to post Share on other sites More sharing options...
Drake Glau Posted March 1, 2011 Report Share Posted March 1, 2011 You need to determine your own uncertainty for human reaction. The Absolute Uncertainty would be 0.01s, yes, but the actual uncertainty will likely be larger. And WOW, 0.1s reaction times, I've never seen that, ever, never. How about 0.5s? Reply Link to post Share on other sites More sharing options...
Bob Posted March 1, 2011 Report Share Posted March 1, 2011 If I haven't forgotten this completely, we were told to give the final uncertainty without decimals.Really, I think different teachers have different opinions on this, so the best way to score high in your IAs is to talk to your own teachers and ask them for help.You need to determine your own uncertainty for human reaction. The Absolute Uncertainty would be 0.01s, yes, but the actual uncertainty will likely be larger. And WOW, 0.1s reaction times, I've never seen that, ever, never. How about 0.5s? I agree, 0.5s would be more like it. But since we (at least I) have to justify our errors and uncertainties in our IAs, so I think its alright when you can provide reasons. Reply Link to post Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.