www.phidot.org

I am trying to figure out how I add year and study area group effects to a staggered entry design encounter history. I have two different study areas and two separate years. The animals in the first study area were tagged from September - February and the animals in the second study area were tagged from November - December. Anyone have any hints or suggestions?

cjacobs wrote:I am trying to figure out how I add year and study area group effects to a staggered entry design encounter history. I have two different study areas and two separate years. The animals in the first study area were tagged from September - February and the animals in the second study area were tagged from November - December. Anyone have any hints or suggestions?

cjacobs,
First hint I would give is you should read the MARK manual and look at the help files in MARK. The answer to your question above (unless there is some complexity you are not telling us about) is discussed, in significant detail with examples, in chapter's 2 and 16 as well as in the 'Encounter Histories Help File' provided in MARK.

bret

Hi Bret...thanks for your post. I have read both chapters in the Mark Manual and just looked at the Help File, but am still unclear. My problem is that for staggered entry design, I have the following format:

10101010101010101010101010101010101010101010101010101010101010 2 0;
10000000000000000000000000000000000000000000000000000000000000 1 0;
00101010101010101010101010101010101010101010101010101010101010 0 5;
11000000000000000000000000000000000000000000000000000000000000 0 1;

This is just a small piece of my dataset. As you can see it contains 31 occassions ( is LDLD format for Known Fate) and four individuals, two of which were tagged during the first week, and two of which were tagged during the second week. (The MARK manual says that I need a frequency column for each group). My complete data set includes 21 groups, so I have 21 frequency columns. So where would I include the information about whether the groups is from Study Area 1 or Study Area 2? I understand how to do this for regular entry ( not staggered)...For staggered enrty would I simply make the first freq. column ( or first two depending on how I want to code it) code for the Study Area...and then I can just add the staggered entry after? I dont see how this would work...Any more hints?

Caitlin

I somehow misintepreted my data when I posted it....this is what I meant to put ( only slight changes)

10101010101010101010101010101010101010101010101010101010101010 2 0;
10000000000000000000000000000000000000000000000000000000000000 1 0;
00101010101010101010101010101010101010101010101010101010101010 0 5;
00110000000000000000000000000000000000000000000000000000000000 0 1;

As you can see it contains 31 occasions (in LDLD format for Known Fate) and 9 individuals, 3 of which were tagged during the first week, and 6 of which were tagged during the second week. (The MARK manual says that I need a frequency column for each group). My complete data set includes 21 groups, so I have 21 frequency columns. While all of the individuals listed above are part of Study Area 1, I am unsure where I would include the information to differentiate Study Area 2 once I add it to the encounter history (I was going to just add 0's up to the week when these Study Area 2 individuals were tagged in November). Would I simply make the first freq. column (or first two depending on how I want to code it) code for the Study Area...and then just add the staggered entry after? I dont see how this would work...

cjacobs wrote:I somehow misintepreted my data when I posted it....this is what I meant to put ( only slight changes)

10101010101010101010101010101010101010101010101010101010101010 2 0;
10000000000000000000000000000000000000000000000000000000000000 1 0;
00101010101010101010101010101010101010101010101010101010101010 0 5;
00110000000000000000000000000000000000000000000000000000000000 0 1;

As you can see it contains 31 occasions (in LDLD format for Known Fate) and 9 individuals, 3 of which were tagged during the first week, and 6 of which were tagged during the second week. (The MARK manual says that I need a frequency column for each group). My complete data set includes 21 groups, so I have 21 frequency columns. While all of the individuals listed above are part of Study Area 1, I am unsure where I would include the information to differentiate Study Area 2 once I add it to the encounter history (I was going to just add 0's up to the week when these Study Area 2 individuals were tagged in November). Would I simply make the first freq. column (or first two depending on how I want to code it) code for the Study Area...and then just add the staggered entry after? I dont see how this would work...

Caitlin,
Your on the correct track with the structure for the different groups (21, which I assume are some sort of sex/age class designation?). As for including study area, if you follow the current .inp format you are using, you could just expand the 'groups' to 42 frequency columns giving you one unique column for each combination of group and area. Renumbering and labeling these groups in MARK will allow you to manipulate each PIM (representing a single group-area combination) independently, which I think is what you are after.

But, with 42 unique group*area combinations, will you have enough data to fit the models you are interested in? If you have 31 sampling occasions, you could, in theory, be talking about a lot of estimable parameters to deal with if you have group*time models that are of interest. So, be judicious in your model development.

Bret

PS: Sorry for being slow on response, was not anything to do with your question, I was on vacation.

Hi Bret,

Thanks so much for your reply...I was beginning to think I was so far off that you couldn't even begin to answer my question! I'm glad you were just on vacation. Anyway, something you said made a lightbulb go off in my head and a bunch of things that had been unclear to me, all of a sudden made sense (after going back through the MARK manual). I was making everything much more complicated than I needed to. I have now modified my encounter history so that there are only two groups (the two study areas) and although I am still using staggered entry, I am not separating out each cohort. Once again, your reponse helped me so much and I really apprecaite it!

Caitlin