# Problem with variance decomposition results

**URL:** <https://forum.dynare.org/t/problem-with-variance-decomposition-results/17613>\
**Category:** ML/Bayesian estimation\
**Created:** [9 March 2021 16:28 UTC](https://forum.dynare.org/t/problem-with-variance-decomposition-results/17613 "2021-03-09T16:28:12Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![lssp1](https://forum.dynare.org/letter_avatar_proxy/v4/letter/l/a587f6/32.png) [@lssp1](https://forum.dynare.org/u/lssp1)\
**Post date:** [9 March 2021 16:28 UTC](https://forum.dynare.org/t/problem-with-variance-decomposition-results/17613/1 "2021-03-09T16:28:12Z")

</div>

Dear Professor Pfeifer,

I was doing the bayesian estimation, and I found that the variance decomposition results after estimation were quite different when I changed one of the observation data series. For instance, the variance decomposition results I get when using Y, C, I, R, Pie as observation series are so different from the results using Y, C, R, Pie, Hours or some other combinations as observation series.

In this case, how should I judge which one is better or which variance decomposition result is correct?

Looking forward to your reply and thank you for your time.

---

<div class="post-metadata">

**Author:** ![jpfeifer](https://forum.dynare.org/user_avatar/forum.dynare.org/jpfeifer/32/5044_2.png) [@jpfeifer](https://forum.dynare.org/u/jpfeifer)\
**Post date:** [12 March 2021 08:30 UTC](https://forum.dynare.org/t/problem-with-variance-decomposition-results/17613/2 "2021-03-12T08:30:06Z")

</div>

There is no easy answer. See [Problems with the results of Bayesian estimation?](https://forum.dynare.org/t/problems-with-the-results-of-bayesian-estimation/16067)