search

kaitlyngaynor / hopland-hunting

Exploration of Hopland hunter GPS data
Apache License 2.0
1 stars 0 forks source link

Can we augment sample size of successful harvest? #37

Closed kaitlyngaynor closed 1 year ago

kaitlyngaynor commented 2 years ago

I have been taking a closer look into the characteristics of successful hunters (particularly in the hour leading up to the kill) and am wondering if we should reconsider some tracks that were previously excluded.

We are currently using data from 24 successful hunters (2 of which are missing harvest times). We have recovered iGotU data from an additional 11 successful hunters (3 of which are missing harvest times) but have not included them in analysis. I think it's worth revisiting why we decided to exclude them—have a look in issues & see if it was because they overlapped too significantly with another hunter? If the latter, we should use them over their companion. If there's some other reason, then maybe some reconsideration is in order.

These 8 hunters have known harvest times, but were deemed unworthy of inclusion in the model:

These two hunters have data with behavioral states now associated, but we don't know the harvest time. I wonder if this is something we could figure out by investigating the tracks?

And these 3 hunters were excluded from analysis AND lack harvest times. So we'd have to investigate tracks to find them, if we wanted to use them in the "hour before harvest" analysis.

@amcinturff just FYI! I'm going to have a closer look when I get a chance, but if you have those old plots and want to dig in on your end, since you had finalized the "to include (Y/N)" column, feel free

kaitlyngaynor commented 2 years ago

@amcinturff I had a look at some of these tracks, and am not sure why you had marked them as "N" when you filled out the 'to include' column—would you mind taking a second look at those in the list? hopefully it's not too hard to dig up that PDF of plots I sent you back in Dec or Jan.

Also, do you think it would be possible to ascertain the time of the kill from manual investigation of the tracks for those 2-5 hunters? Not sure...

just trying to augment the sample size if we want to look at 'success'

kaitlyngaynor commented 2 years ago

@amcinturff Flagging this again, not sure if this emails you or not but this is something we can chat about in person! This remains the big 'to do' on the analysis front before we tie a bow on it

kaitlyngaynor commented 1 year ago

Decided to revisit our plan of subjectively removing hunters that traveled together based on visual inspection, and instead use tracks from all hunters that yielded usable data (even if partial).

kaitlyngaynor commented 1 year ago

So we are now including all of these:

These two hunters have data with behavioral states now associated, but we don't know the harvest time. I wonder if this is something we could figure out by investigating the tracks?

And these 3 hunters were excluded from analysis AND lack harvest times. So we'd have to investigate tracks to find them, if we wanted to use them in the "hour before harvest" analysis.

kaitlyngaynor commented 1 year ago