Delta method and model averaging

[constant survivor]

Hello everyone,
again it's me. This time really looking forward to your thoughts on concrete calculation problems.
With the link at the end you have access to an excel sheet (read and write). With the calculations in there I hope I have found a way to solve the following problem:

I have two models which I want to average (model averaging output in the xcel sheet).
One model is with constant phi and the other is with two estimates for phi. The latter model has 5/29 intervals for phi1 and 24/29 intervals for phi2 (this is their weight in column B).

As the top model is the one with constant phi, I wanted to "join" the two weighted average estimates according to their weight (column B) so that in the end I have again only one single estimate for the phi(.) model. Of course then I have to recalculate SE and CI and all this stuff is related to Delta method I guess.

So please, I am thankful for comments on the calculations in the xcel sheet. I hope it is sufficient to answer my question. The column headers show a description of the formula used and the parameter that is calculated. The calculations are done with xcel operations. As I am from Germany just note: "WURZEL" is german for "squareroot" and "SUMME" is german for... you guess it: "SUM". The rest should be esperanto-like.

The numbers that I hope to represent my overall result are within column J-O.

Oh man, I hope this is comprehensible for anybody out there. Be aware of the fact please, that I am a rooky in mathematics and statistics. The table is not my work alone by the way...

https://www.dropbox.com/scl/fi/8f9irq3z1onz25yj065q8/phidotforum.xlsx?dl=0&rlkey=bdfju500hvoischm8vch32fjt

Best
Hannes

[constant survivor]

404 error on above link...

maybe this:

https://www.dropbox.com/scl/fi/aalc5z0iso0j7w41nnxgr/phidotforum.xlsx?dl=0&rlkey=2my9opy5irgkdreji3s4w7cv8