Cross Validated is a question and answer site for people interested in statistics, machine learning, data analysis, data mining, and data visualization. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I have two sets of experimental data for exponential decays of isotopes and I calculated the ratio of the two isotopes as a function of time. The theoretical model and the fit of the experimental data are respectively :



The two curves are near of each other (I'm only interested in comparing the [0-110] x-range) but is there a way to calculate an error between the two curves (like a standard deviation)? I need to determine if the data are normally distributed so I thought getting the standard deviation from the theory/fit would be a correct way... Any ideas?

share|improve this question
If you have good reasons to believe the theoretical model, can you look at the errors in the actual data minus the corresponding data from the theoretical model? Is there a pattern in the errors, evidence of additional variables, etc? – bill_080 Nov 20 '12 at 1:18

You could use bootstrapping to estimate the distribution of your decay parameter. Then you test whether the theoretical parameter (+- region of practical equivalence) falls within 95% of that distribution.

share|improve this answer
How would this bootstrap work, exactly? How do you resample from, say, a set of $(t,x)$ values? (Although there are some well-known ways, one has to make choices and assumptions. I believe that addressing my question will force you to consider assumptions about the errors in the data, which may have some relevance to the second part of the question about a normal distribution.) – whuber Nov 19 '12 at 21:21

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.