Constructing the design matrix

[scoster]

Hi,

I'm a new user to MARK, have read the Gentle Introduction, and have searched the forum, but can't find the answer to a question. I'm trying to understand how to construct the full design matrix in order to later construct additive models (in closed-population models using heterogeneity and behavior) , and in reviewing the full design matrix shown in the Gentle Introduction (14.6, pg.14-24) I don't understand why the interaction between heterogeneity and encounter rate was left out of the full model?

This question is nagging at me because in trying to construct a full model from the design matrix using my own data set (with variables group=sex, heterogeneity, behavior, and time), I'm not sure which of the interactions get left out of the full model.

Any insight would be greatly appreciated.

Thank you,
Stephanie

[cooch]

Hi,

I'm a new user to MARK, have read the Gentle Introduction, and have searched the forum, but can't find the answer to a question. I'm trying to understand how to construct the full design matrix in order to later construct additive models (in closed-population models using heterogeneity and behavior) , and in reviewing the full design matrix shown in the Gentle Introduction (14.6, pg.14-24) I don't understand why the interaction between heterogeneity and encounter rate was left out of the full model?

This question is nagging at me because in trying to construct a full model from the design matrix using my own data set (with variables group=sex, heterogeneity, behavior, and time), I'm not sure which of the interactions get left out of the full model.

Any insight would be greatly appreciated.

Thank you,
Stephanie

Because you can't, in fact, create a full DM for the model you're describing - a lot of parameters would not be identifiable. This is noted in the text. Also, don't forget that Pi is a constant over time - as such, there is no interaction between Pi and encounter probability, since that would imply time varying Pi - this is also noted in the text.

[scoster]

Thanks for your insight, that makes sense.

Stephanie