I am new to Statistics and trying to intuitively understand how a change in degrees of freedom affects the mean of a chi-square distribution.

Suppose, We have $n$ normal random variables such that $X_1 + \cdots + X_n = 0$

Now, for a chi-square distribution, the expected value is = degrees of freedom.

In this case, the degrees of freedom is $n-1$.

If we consider an example where $X_1 + X_2 = 0 $.

Then $X_1^2 + X_2^2 = 2 X_1^2$ although has $1$ degree of freedom yet it's mean value is $2$!

Where am I making a mistake?

Thanks a lot!

  • 2
    $\begingroup$ As @Xi'an points out in his answer, you probably mean $E[X_1+X_2]=0$, e.g. 1 and -1. $\endgroup$ – PaulG Feb 21 at 18:30
  • $\begingroup$ @PaulG Thank you very much. Could youi please have a look at this question as well. I have been thinking about this for long : stats.stackexchange.com/questions/510523/… $\endgroup$ – MathMan Feb 21 at 18:34

If $X_1+\ldots+X_n=0$ with probability one, the vector $(X_1,\ldots,X_n)$ is not a Normal vector, hence $X^2_1+\ldots+X^2_n$ is not a $\chi^2_n$ random variable.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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