I am reading an internal paper that says:

Let $\sigma^2 V_1$ equal the variance of $\sum_{m\in M}Z_m - Z_0$ and $\sigma^2V_2$ equal the covariance of $||M||^{-1}\sum_{m\in M}Z_m - Z_0$ and $Z_m - Z_0$.

I am unfamiliar with the notation of the $V$s. Can anyone explain what this is referring to or point me in the right direction?


1 Answer 1


The V's are used for the variance-covariance matrix e.g. in longitudinal data where there is no longer a single variance term, but instead, a matrix representing the variance and covariance between measurements.

The example you describe appears to refer to each matrix separately depending on the subscript 1 or 2, which I don't recall seeing before.

  • $\begingroup$ That certainly helps. But what, then, is the significant of the $\sigma^2$ in the notation? I am familiar with building out covariance matrix, usually denoted $\Sigma$, with time series data. Unless the $\sigma^2$ is just effectively factored out? $\endgroup$ Apr 15, 2014 at 6:08
  • $\begingroup$ I don't have a definitive answer on this, nor the expertise... but I believe the σ2 is essentially factored out because the diagonal variances in the matrix might differ in some predictable way and the covariances in the off-diagonals may also have some pattern. That way, the matrix contains the information for the pattern only. I am not so sure this notation works for a completely unstructured matrix, which would require a different variance and covariance parameter in each cell of the matrix that might not be directly relatable to a single σ2. $\endgroup$
    – Moose
    Apr 16, 2014 at 18:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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