how do i use uncertainties in my IA?

[abelkoontz]

I calculated the uncertainty for raw data using max-min/2 (what my teacher told me) this is for systematic error

So for random error do i JUST calculate the % uncertainty for each apparatus (apparatus uncertainty/measured value x100) and then add them together? what else am I supposed to do? HELP

[Vioh]

1 hour ago, abelkoontz said:

I calculated the uncertainty for raw data using max-min/2 (what my teacher told me) this is for systematic error. So for random error do i JUST calculate the % uncertainty for each apparatus (apparatus uncertainty/measured value x100) and then add them together? what else am I supposed to do? HELP

Nope, the max-min/2 method is not for systematic error. It is in fact for the random error. Random errors are usually caused by random factors originated from the equipment, human errors (like when pressing the stop button of a stop watch), or just random changes in the environment, which means that random errors can cause a spread in the measurements of the same variable. And since max-min/2 (or range/2 method if you will) measures the spread of the data, so we use it to measure random errors.

Suppose that you have the formula V= 1/4 x pi x D^2 x H, and your entire experiment is to measure D and H such that you can then use them to calculate V. Now, to calculate the uncertainty for V, you first try to calculate the uncertainties for D and H, and then apply the propagation of errors technique on them. More details on how to propagate the errors here: https://www.ibsurvival.com/topic/35954-uncertainty/. Note that the uncertainty for V, D, and H, are all for the random errors, not the systematic errors! If you want to calculate the systematic error for the experiment, you have to compare the calculated value for V against the theoretical value that you get from the scientific consensus. Suppose that your experiment is to try to calculate the acceleration due to gravity, and you get the value of 9. Since the theoretical value is 9.81, so the systematic error is (9.81-9)/9.81*100 = 8.25%. This is also called the experimental error.

But how do you get the uncertainties for D and H? Suppose that you have done the experiments and have the raw data for both of the variables. Then you can use either of these methods to get the uncertainties for D and H:

1. Use range/2 method.
2. Use the uncertainty from the equipment. This is usually labeled on the equipment itself. If you use an analogue equipment, like a ruler or an analogue ammeter, then the uncertainty is usually half of the smallest interval.
3. Make a reasonable estimate. This method is probably only used if you try to calculate time uncertainty of a stop watch. The estimate here is to account for the human errors and human reaction time when clicking the stop button.
4. Use the standard deviation of the data.

As mentioned, method 3 is rarely used. Method 4 is also rarely used in IB, at least in physics and chemistry (in biology though, it seems to be used a lot). The only two methods left are 1 and 2. In most situations, I recommend you to choose the method that gives the highest uncertainty. If method 1 gives a higher uncertainty for D, then use range/2 for D. If method 2 gives a higher uncertainty for V, then use the uncertainty from equipment for V.