Biologically Meaningful SE's

[Jordan Green]

Good Afternoon,

First of all, many thanks to those maintaining unmarked and generating discussion within the group.

The first chapter of my MSc thesis is focused on the the optimization and design of monitoring networks to achieve various objectives associated with bird monitoring. One such objective may be to use the data to build occupancy models to predict occurrence to advise management activities, direct conservation efforts...etc. 

When designing the monitoring network, metrics of interest may be to maximize the probability of detecting a species at least once during K sampling events (p*) and minimizing the SE of estimates. This information can then guide how many sites and repeat visits would be advised depending on the detection/occupancy probability of the species. 

Considering any arbitrary threshold for SE could be used to design the network,  I am curious as to what is a biologically realistic/meaningful threshold to use? I am looking to generate discussion so any feedback is appreciated!

Cheers,

Jordan Green

[Jeffrey Royle]

hi Jordan,

 you often see people identify specific coefficient of variation (CV) targets such as 20%, reflecting the fact that the standard error of a quantity often depends on the magnitude of the parameter to be estimated.

 In a design context it is more common I think to try and optimize a design (minimize the CV or SE) while holding total effort fixed. Thus for a given project budget you can vary spatial and temporal replicates (or other controllable quantities) to minimize SE.

regards

andy

--
*** Three hierarchical modeling email lists ***
(1) unmarked (this list): for questions specific to the R package unmarked
(2) SCR: for design and Bayesian or non-bayesian analysis of spatial capture-recapture
(3) HMecology: for everything else, especially material covered in the books by Royle & Dorazio (2008), Kéry & Schaub (2012), Kéry & Royle (2016, 2021) and Schaub & Kéry (2022)
---
You received this message because you are subscribed to the Google Groups "unmarked" group.
To unsubscribe from this group and stop receiving emails from it, send an email to unmarked+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/unmarked/d520d787-aa46-4366-a498-199e96790a6dn%40googlegroups.com.

[Murray Efford]

Since Jordan asked for discussion...

You still need to know whether the design is going to do the job, i.e. detect variation you might consider significant.

The "traditional" target of 20% relative standard error (RSE, aka CV) has always seemed weak to me, and claims like  'good enough for management purposes' don't help. You might like to check Fig. 1 and the first paragraph of the Discussion in our 2019 paper DOI: 10.1111/2041-210X.13239. 

Murray

---

From: unma...@googlegroups.com <unma...@googlegroups.com> on behalf of Jeffrey Royle <jar...@gmail.com>
Sent: Thursday, 29 May 2025 06:54
To: unma...@googlegroups.com <unma...@googlegroups.com>
Subject: Re: [unmarked] Biologically Meaningful SE's

 

To view this discussion visit https://groups.google.com/d/msgid/unmarked/CAE%2BwgF0wjmgBFw-RdtzhMTAen8oVYMiUpVZ%2B5V-UnW-geG-h8w%40mail.gmail.com.

[Jordan Green]

Thank you both for your comments, I will go ahead and test different thresholds for CV to see how parameter estimate vary. Are there any other indicators that come to mind when measuring how well a given design works? I'm am trying to include a wide perspective of objectives/indicators in my discussion. 

Jordan