Best Method of Analyzing Experimental Percent Error

[Jefffff]

Homework Statement

I have performed a chemistry experiment where I analyzed how the diluting of two reactants with the same volume and initial concentration impacts the percent yield of a precipitate product. Thus my IV was concentration and my DV was percent yield. My question is what is the correct/most effective method of analyzing the percent error of my experiment?

Homework Equations

percent yield = experimental yield / theoretical yield * 100%

percent error = (|theoretical value - experimental value|) / theoretical value * 100%

The Attempt at a Solution

I have considered two prospects:

1 . I plot concentration of reactants versus experimental yield and linearize if necessary, finding the slope of the graph. I then plot concentration versus theoretical yield and repeat the same process. I use the experimental slope as my experimental value and the theoretical slope as my theoretical value in the percent error calculation. The problem with this is that there is no consideration of the dependent variable, which is percent yield.

2. I find the averages of all experimental masses for every IV, (averages of averages), getting one number representing the average of all my experimental data, and compare this to the averages of all my theoretical masses (calculated through stoich) in the percent yield equation.

Without examining the root mathematical causes of the variance in percent yield due to changes in concentration, I don't think I can reasonably calculate the percent of error. If this is the case, I was wondering if anyone could point me in the right direction.

Any help would be greatly appreciated!

 

[mfb]

It depends on what you want to do with that number.

Do you expect that all experimental values are too low/high in the same (relative) way?
Do you expect yield/concentration to be linear?

If both answers are "yes", method 1 is probably a good idea.
"yes&no": 2 is probably a good approach, but a weighted average could be better, where the weight takes into account the precision of the measurement.
If the answer to question 1 is "no", then what do you want to quantify with a single number?

 

[Jefffff]

My results when I graphed molarity against percent yield were a weak positive correlation. The slope was 51 exactly however the correlation was 0.8978. I will consider both prospects as I think the percent yield would increase linearly with molarity increases but I haven't been able to find any information online that supports this claim.