Dynamic occupancy goodness of fit test

[jessiec...@gmail.com]

Hi there,

I am please needing some advice/clarification on how to determine whether a dataset is appropriate to use (i.e. not over- or under-dispersed) for dynamic occupancy modelling according the outcome of the mb.gof.test for the global occupancy model. I understand that the c-hat value needs to be as close to 1 as possible (extreme values >3/4 indicate lack of fit). In addition to an appropriate c-hat value, does the chi-square p-value need to be greater than 0.05 in order for the data to be appropriate or is the c-hat value of more importance?

Below are the outcomes of goodness-of-fit tests for two global occupancy models. Would both datasets be appropriate to use or only the dataset used in Model 1?

Model 1

Goodness-of-fit for dynamic occupancy model

Number of seasons:  10 

Chi-square statistic:

 Season 1  Season 2  Season 3  Season 4  Season 5  Season 6  Season 7  Season 8 

   2.8930    2.4356    2.4501    8.9530    4.1339    3.6981    1.2422    0.8141 

 Season 9 Season 10 

   7.2423    4.1897 

Total chi-square = 38.0521 

Number of bootstrap samples = 1000

P-value = 0.068

Quantiles of bootstrapped statistics:

  0%  25%  50%  75% 100% 

 9.1 21.4 25.8 30.9 55.5 

Estimate of c-hat = 1.44

Model 2

Goodness-of-fit for dynamic occupancy model

Number of seasons:  10 

Chi-square statistic:

 Season 1  Season 2  Season 3  Season 4  Season 5  Season 6  Season 7  Season 8 

   1.5827   21.4999    7.0277   11.1775    4.0588    2.2721    1.3821    6.3738 

 Season 9 Season 10 

   4.6023    4.9692 

Total chi-square = 64.9461 

Number of bootstrap samples = 1000

P-value = 0

Quantiles of bootstrapped statistics:

  0%  25%  50%  75% 100% 

  11   23   27   33   62 

Estimate of c-hat = 2.29 

All the best,

Jess

[Marc J. Mazerolle]

Hi Jess,

The chi-square suggests lack-of-fit, more so for model 2 than model 1 (marginal). The c-hat estimates > 1 indicate that the lack of fit would be due to overdispersion. You could then proceed with your analysis and adjust your inferences by the c-hat.

Best,

Marc

____________________
Marc J. Mazerolle

Professeur agrégé et directeur du bac. en environnements naturels et aménagés

Département des sciences du bois et de la forêt
2405 rue de la Terrasse
Université Laval
Québec, Québec G1V OA6, Canada
Tel: (418) 656-2131 ext. 407120
Email: marc.ma...@sbf.ulaval.ca

Veuillez noter que je suis en télétravail. Le meilleur moyen de me rejoindre est par courriel.

---

De : unma...@googlegroups.com <unma...@googlegroups.com> de la part de jessiec...@gmail.com <jessiec...@gmail.com>
Envoyé : 14 octobre 2020 04:21
À : unmarked
Objet : [unmarked] Dynamic occupancy goodness of fit test

 

[Externe UL*]

--
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 on the web visit https://groups.google.com/d/msgid/unmarked/15eb8593-a95b-4628-9e6a-4cfe50bdac1cn%40googlegroups.com.

*ATTENTION : L’émetteur de ce courriel est externe à l’Université Laval.

Évitez de cliquer sur un hyperlien, d’ouvrir une pièce jointe ou de transmettre des informations si vous ne connaissez pas l’expéditeur du courriel. En cas de doute, contactez l’équipe de soutien informatique de votre unité ou hamec...@ulaval.ca.

---

[jessiec...@gmail.com]

Hi Marc,

Thank you for your reply last month. Just to clarify, if I am following a step-wise approach to my dynamic occupancy models, whereby I determine probability of detection first (i.e keeping psi, col and ext constant) do I need to use QAIC (with the c-hat value of the global model) to determine which factors best describe detection to use in the subsequent models?

All the best,

Jess

[Marc J. Mazerolle]

Hi Jessie,

Yes, I would use QAIC using the c-hat of the global model to determine the most parsimonious model for detection probability.

Best,

Marc

____________________
Marc J. Mazerolle

Professeur agrégé et directeur du bac. en environnements naturels et aménagés

Département des sciences du bois et de la forêt
2405 rue de la Terrasse
Université Laval
Québec, Québec G1V OA6, Canada
Tel: (418) 656-2131 ext. 407120
Email: marc.ma...@sbf.ulaval.ca

Veuillez noter que je suis en télétravail. Le meilleur moyen de me rejoindre est par courriel.

De : unma...@googlegroups.com <unma...@googlegroups.com> de la part de jessiec...@gmail.com <jessiec...@gmail.com>

Envoyé : 13 novembre 2020 05:50
À : unmarked
Objet : Re: [unmarked] Dynamic occupancy goodness of fit test

 

[Externe UL*]

To view this discussion on the web visit https://groups.google.com/d/msgid/unmarked/73b47abb-745a-4691-999d-de10c9381754n%40googlegroups.com.

[jessica comley]

Perfect.

Thank you. 

All the best, 

Jess 

To view this discussion on the web visit https://groups.google.com/d/msgid/unmarked/1605271608695.13292%40sbf.ulaval.ca.