-1
$\begingroup$

Suppose $X_1, \ldots, X_N$ are NOT identically distributed nor independent. Now, suppose I wanted to compute expectations, such as:

$$ E(X|Z) $$

where Z is another variable. Does this change anything than what we are used to?

$\endgroup$
  • 2
    $\begingroup$ What exactly is $X$ ? And also how do you solve this problem when they are iid ? $\endgroup$ – P. Quinton Feb 11 at 13:44
  • $\begingroup$ $X$ are any arbitrary covariate specification. Under iid I would assume that the iterated expectations would follow, and hence I could do iterated expectations on the above. But, with the very action of iterated expectations, does it implicitly require iid? $\endgroup$ – user321627 Feb 11 at 13:50
  • 2
    $\begingroup$ This is unclear to me, can you provide a short example ? $\endgroup$ – P. Quinton Feb 11 at 13:54
  • $\begingroup$ I have the impression that $X$ is a random variable while $X_n$ is a sample of $N$ draws. $\endgroup$ – Bertrand Feb 11 at 13:58
  • 2
    $\begingroup$ I don't think this is answerable until you say what the relation between the $X_i$ and $X$ is. Usually when people use this kind of notation the $X_i$ represent a random sample from a variable $X$. You say that your $X_i$ can have a different distribution to $X$, but then what are you assuming about the distributions they have? Clearly without some assumption these $X_i$ can tell you nothing. $\endgroup$ – Matthew Towers Feb 11 at 14:40

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Browse other questions tagged or ask your own question.