Meaning of conservative variance estimate

[mkenglish]

I anticipated a high number of differentially abundant taxa in the dataset that I used ancombc on, so I set conserve = FALSE, as the manual says to do. Where exactly does this parameter come into play? In the tests that ancombc runs under the hood, what does a conservative variance estimate mean?
Thank you for any and all help!

[FrederickHuangLin]

Hi @mkenglish,

Thanks for your inquiry! The conserve argument specifies whether we would like to take into account the variance of the bias term in calculating the total variance of the estimated differential abundance.

Specifically, when conserve = FALSE, we are using equation (33) in the paper; while when conserve = TRUE, we are using equation (35) in the paper.

Best,
Huang

[mkenglish]

Thank you so much!