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 some measured datapoints. Each of the points has attached an error, depending on how precise the measurement was done.

Now I'm using the squared weighted mean with the inverse of the errors as weights so that I emphasize more on the preciser values and less on the values where the error is large.

What I'm wondering is what can I say about the precision of my squared weighted mean. Is there a way to calculate the expected error range for it?

share|improve this question
Are these observations independent? Do you know the mean and variance of each individual observation? – Macro Aug 30 '12 at 20:15
This gets interesting if the individual point errors were themselves estimated, such as from repeated measurements ;-). Might that be the case? – whuber Aug 30 '12 at 20:38
As it turns out the expected errors for each measured value are the variances from the multiple measurements for that value. We therefore have $w_i=\frac{1}{\sigma_i^2}$, $w_i$ being the weights for each value. – cdecker Aug 30 '12 at 21:06
@cdecker I think using 1/standard deviation is a more appropriate weighting than dividing by the variance. – Michael Chernick Sep 4 '12 at 14:47
up vote 1 down vote accepted

Suppose you have k independent random variables X$_i$ i=1,2,..,k each having variance σ$_i$$^2$ and let X$_w$= Σ c$_i$ X$_i$. The variance of X$_w$= Σ c$_i$$^2$ σ$_i$$^2$. You can use that variance to get the standard error of that estimate.

share|improve this answer

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.