variance ratio

[mattansb]

No description provided.

[codecov]

Codecov Report

❌ Patch coverage is 96.22642% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 90.90%. Comparing base (8fdbfad) to head (2b8c634).
⚠️ Report is 60 commits behind head on main.

Files with missing lines	Patch %	Lines
R/vars_ratio.R	96.22%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #619      +/-   ##
==========================================
+ Coverage   90.74%   90.90%   +0.16%     
==========================================
  Files          57       58       +1     
  Lines        3564     3617      +53     
==========================================
+ Hits         3234     3288      +54     
+ Misses        330      329       -1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[bwiernik]

Let's a few options here for CI methods and computing SD ratio instead, pulling them from the metafor::escalc() code.

(SD ratios are often called variation/variance ratios in biology and ecology work.)

Maybe let's call the function variation_ratio()?

[mattansb]

For independents samples, metafor uses these (with the +1/(2 * (n1i - 1)) -1/(2 * (n2i - 1)) small sample bias correction):

yi <- log(sd1i/sd2i) + 1/(2 * (n1i - 1)) - 1/(2 * (n2i - 1))
vi <- 1/(2 * (n1i - 1)) + 1/(2 * (n2i - 1))

What should vi be here if the correction is not applied? 🤔

Maybe let's call the function variation_ratio()?

Sure.

[bwiernik]

vi is the same -- the adjustment is an additive constant so it doesn't affect the variance because of the calculus

[mattansb]

Oh right, duh 🤦‍♂️