Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. 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 a sequence of $n$ random variables, each drawn from a different distribution $X_1\sim A_1, X_2\sim A_2, \ldots, X_n\sim A_n$. The distributions $A_1, A_2, \ldots, A_n$ have nice properties: finite mean, variance, as well as third and fourth moments; all of them are defined over the support $\mathbb{R}$. Unfortunately, besides the moments and support, I am not given any more information about $A_1,A_2,\ldots,A_n$.

I am interested in studying the convergence of the normalized sum of these random variables:


in particular, I would like to know if knowledge of moments of $X_i$'s is sufficient to show so that the sequence satisfies Lindeberg's condition. If it satisfies the condition, then CLT applies to the summation. I think that I understand the formula for the condition:

$$\lim_{n\rightarrow\infty}\frac{1}{s_n^2}\sum_{k=1}^n\int_{\{|X_k-\mu_k|>\epsilon s_n\}}(x-\mu_k)^2f_k(x)dx=0$$

where $s_n^2=\sum_{k=1}^n\sigma_k^2$.

(I wrote the formula as a Riemann integral, please let me know if I made a mistake -- the wikipedia article uses Lebesgue integral but I am not as comfortable with those.)

If moments are not sufficient than what additional information about $X_i$'s do I need? If they are sufficient, how do I use their properties to prove that condition is satisfied? I think I understand the intuition behind the condition: no single random variable in the sequence can "overwhelm" the variance of the sum -- is that correct intuition?

share|cite|improve this question

If I understand your question correctly then perhaps Lyapounov's condition might help. Take a look at pages 359-362 of Probability and Measure by Patrick Billingsley. Basically if your random variables are independent and some conditions on $|X_{n}|^{2+\delta}$ for $\delta>0$ and $E[|X_{n}|^{2+\delta}]$ are satisfied then the Lindeberg condition is implied. This is not exactly (as far as I can see) about the moments of the $X_{n}$'s but it is concerned with properties of your random variables.

share|cite|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.