Timeline for Mark recapture with no knowledge of marked individuals
Current License: CC BY-SA 4.0
19 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jul 9, 2024 at 6:18 | vote | accept | Josh9999 | ||
| Jun 14, 2024 at 13:27 | history | became hot network question | |||
| Jun 14, 2024 at 9:41 | answer | added | Jarle Tufto | timeline score: 4 | |
| Jun 14, 2024 at 9:05 | history | edited | Jarle Tufto | CC BY-SA 4.0 | Fixed spelling mistakes |
| Jun 14, 2024 at 8:32 | history | edited | Josh9999 | CC BY-SA 4.0 | added 387 characters in body |
| Jun 14, 2024 at 8:18 | comment | added | Glen_b | @Josh It might be worth mentioning your notation for $k$ in the question. | |
| Jun 14, 2024 at 7:38 | comment | added | Josh9999 | @PBulls This is the main issue, I believe this is not enough to estimate $N$. I need to prove that this estimator converges to the sample size, not to the population size. | |
| Jun 14, 2024 at 7:33 | comment | added | Josh9999 | @Glen_b The population sizes are $N$ (total, unkown), $K$ (marked, unkown), and the sample sizes are $n$ (total sampled, known), $k$ (marked, known). | |
| Jun 14, 2024 at 7:15 | comment | added | Glen_b | @Josh The data they begin with is the number sampled (n) and the number of those that were marked (m, say)? | |
| Jun 14, 2024 at 6:19 | comment | added | user2974951 | If the individuals were sampled with replacement (without keeping track of already marked individuals) then a more appropriate distribution would be the binomial distribution. | |
| Jun 14, 2024 at 6:09 | comment | added | PBulls | Do you at least know if you've recaptured the same specimen, i.e. is the mark unique? If not you might be able to estimate $p$ but I don't see how that tells you anything about $N$. | |
| Jun 14, 2024 at 5:32 | comment | added | Josh9999 | Better said, they made 1 single (large) observation (size n) and they sub-sample (bootstrap or jackknifes, not entirely sure) this vector to create the histogram. | |
| Jun 14, 2024 at 5:27 | comment | added | Josh9999 | They create a histogram of the observed proportions (they make many observations) and they fit a hypergeometric distribution trough minimum squares. Thanks Glen_b. | |
| Jun 14, 2024 at 5:23 | comment | added | Glen_b | +1 Nice clear explanation. Do you know what estimators they're using to estimate N and p? | |
| Jun 14, 2024 at 5:20 | history | edited | Josh9999 | CC BY-SA 4.0 | Fixed typos; added 32 characters in body |
| Jun 14, 2024 at 5:19 | history | edited | Josh9999 | CC BY-SA 4.0 | added 34 characters in body |
| Jun 14, 2024 at 5:06 | history | edited | Josh9999 | CC BY-SA 4.0 | added 1 character in body |
| S Jun 14, 2024 at 5:04 | review | First questions | |||
| Jun 14, 2024 at 7:29 | |||||
| S Jun 14, 2024 at 5:04 | history | asked | Josh9999 | CC BY-SA 4.0 |