Talk:Non-paternity event

Re-write
I've completely rewritten the article to make it more encyclopaedic. The old article was written like a piece of advice for someone considering DNA testing. I have also added many more references and deleted single reference from the old article that was really a commerical endorsement.--Gak (talk) 02:32, 13 February 2009 (UTC)

Merge with Misattributed paternity
I merged this article with Misattributed paternity because after I removed dubious, unsourced/poorly sourced material and statements that misrepresented their sources from both articles, they were both stubs with almost identical content and sources, practically verbatim, and there was no meaningful difference between them. It didn't seem to me that they even are 2 distinct topics since the sources use the phrases misattributed paternity and non-paternal interchangeably. PermStrump (talk) 20:29, 21 March 2016 (UTC)

Table with non-paternity frequencies
Hi Permstrump. Here is the table. Please comment if there is anything you disagree with.


 * I collapsed the table and references here b/c it was taking up a lot of space. I removed this table from the article because it's plagiarism and a copyright violation in addition to being original research by synthesizing data from multiple sources. The column names and most of the data come from a single source and we don't have permission from the author to reproduce it (see WP:COPYVIO). Data from additional studies were added to that table. This is a problem because it's combining data from multiple sources (see WP:SYNTH). It's also a problem because in order to make the data from the other sources fit with the column headings, it required personal interpretation of the data by the wikipedia editor since the data weren't reported the same way in the other studies (see WP:OR). This table existed in a previous version of this article and I looked up all of the sources a few weeks ago before I removed it the first time. I can't remember off the top of my head which one the bulk of this table came from. I'm not sure if you've even cited it. I'm going through the research I did before, so when I find it, I'll post an update. Regardless, the bottom line is that it's illegal and original research and needs to be removed. PermStrump (talk)  14:51, 9 April 2016 (UTC)
 * Ok, the bulk of the data come from Bellis et al. (2005) and is a copyright violation. Everything from Sykes und Irven 2000 and below is OR. PermStrump (talk) 15:08, 9 April 2016 (UTC)
 * FYI I posted a comment about this on WT:Copyright problems] PermStrump (talk) 15:35, 9 April 2016 (UTC)
 * Thanks for answering so quickly and clearly, Permstrump. I have now checked the journal's copyright requirements (BMJ Publishing) and you are right: although they will provide free permission for reproducing small amounts of content and figures, they do require notification, which we have not (yet) done. So I have deleted the first part of the table which is from Bellis 2005. As for your second point on original research, this hardly applies. No conclusion is drawn in the table which is not already in the paragraph (which you yourself wrote...) - the table simply updates the Wikipedia reader with more recent research than Bellis 2005. Are you in agreement with this solution of a shortened table? 86.154.101.88 (talk) 17:53, 9 April 2016 (UTC)
 * : Thank you for removing the copyrighted material. The table is OR because the data weren't reported this way in the studies cited so the editor who created it had to interpret the findings in order to extract numbers. Most of the rows contain major errors or fabricated information that isn't in the source at all.
 * Forster et al. (2015): The sample size for both rows is "<24097." How is the reader supposed to know what <24097 means? Does it mean close to 24097? Half of 24097? We don't know because the authors never told us. The study says they analyzed DNA from 24,097 families from Europe, the Middle East and Africa, but it doesn't give a breakdown of the total number/percentage from each region. This is original research because the editor thought it was ok to just the total sample size and indicate that "well, each ethnicity had to have been less than that."
 * King and Jobling (2009): Couple of problems. To name a few, (1) the table lists the bias as, "Founder postulated(+)." What the heck does that mean? Those words don't come from the study. It suggests to me that whoever put that in the table thought this study involved paternity tests, which it doesn't. (2) The sample size for all of the rows is 1! Presumably the person who created this table was basing that off of page 1097 where it says the haplotype clusters for those 5 last names could potentially "reflect foundation by a single man." This demonstrates a complete lack of understanding of the entire study and is a perfect example of why taking data reported in one way and entering it into a table in a different way is original research and also why original research doesn't belong on WP.
 * Sykes and Irven (2000): Similar mistakes as with King and Jobling, nothing remotely similar to "Founder postulated(+)" appears in the text. The word "bias" doesn't even appear in the text. Similar to King and Jobling, Sykes and Irven weren't writing about paternity tests. So putting the sample size from those 2 studies in the same column as the sample size from the other studies that involved paternity testing is like comparing apples and oranges.
 * Edwards (1957): It's from 1957. Even if it were more recent, they didn't use DNA testing like the other studies cited, once again it's comparing apples and oranges and is WP:SYNTH.
 * PermStrump (talk) 01:25, 10 April 2016 (UTC)


 * I agree with most of what you say and have removed the editor's original research ("sample size" and "bias") from the table and simply kept the published values. I am less persuaded by your argument that the classic Edwards 1957 study is inappropriate (yes, it is apples and oranges, but precisely that comparison would be informative for the reader, and it is the first such study), but I have removed him for the moment because you seem to have a strong view on this. If you change your mind on Edwards 1957, it would be much appreciated.81.131.173.86 (talk) 07:08, 10 April 2016 (UTC)
 * The Forster study needs to go. First of all, it's not at all representative of any kind of normal population sample. Secondly, it's inaccurate. All of the people from the Middle East weren't Kurds, just most of them. Likewise, all of the people from Africa were Nigerian. Then the only thing left is the study by King and Jobling which as already been discussed in the narrative. WP:WHENTABLE says it prefers narrative or lists over tables anyway and tables are supposed to be a last resort for data that can't be clearly expressed in the narrative or in a list. PermStrump (talk) 06:40, 11 April 2016 (UTC)
 * I am afraid your argumentation against the table is becoming subjective now: omitting Sykes and Irven (whom you forget in your reasoning) would mean deleting the first ever Y-surname misattribution study. And omitting the Forster et al. immigration study would mean omitting the first immigration testing results. Bellis 2005 is an excellent compilation, but it is now 11 years out of date, and it overlooked the Y-surname work. Hence the need for a table displaying newer results from new research. Imagine someone now conducts a study on non-paternities among a sample of homosexual couples (only recently legalised), or in a sample of Catholics versus Protestants. Such newer research could then conveniently be added to the table in the future. The Wikipedia reader then has all the facts at his/her fingertips in the table.
 * And I have already advised against omitting the first paternity testing study by Edwards 1957 - I will not insist, but my reasoning here is that the early studies in each of the four categories (academic paternity tests, disease screening, surname studies, immigration testing) should be less influenced by public awareness of the power of DNA than the later studies in each category. Consider for example that a pregnant woman nowadays might think twice about disease screening if she thought that the result might go onto a database and possibly endanger her relationship with her current partner at some point in the future. This may partly explain why the Swiss value is much lower than the original English study. I am not saying this aspect should be mentioned in the Wikipedia article now, but please keep an open mind on Wikipedia when this issue starts to come up in the research.86.154.101.114 (talk) 07:00, 12 April 2016 (UTC)

Dr. Guha's comment on this article
Dr. Guha has reviewed this Wikipedia page, and provided us with the following comments to improve its quality:

"I suggest adding a sub-section on "Non-paternity before genetic testing". This could go as follows. "Anthropologists classify different societies in terms of "paternity confidence" (the likelihood that a woman's children were fathered by the official long term partner) based on social norms towards premarital and extramarital sex in these societies, as well as on the extent of practices like wife-swapping. Guha (2016) discusses some of the anthropological literature on this issue. Broadly, cultures where inheritance was in the male line were more likely to have restrictive sexual norms, implying a relatively low probability of non-paternity in these cultures." Please cite Guha, Brishti (2016): Grandparents as Guards: A Game-Theoretic Analysis of Inheritance and Post-Marital Residence in a world of Uncertain Paternity, MPRA Working Paper no 70954."

We hope Wikipedians on this talk page can take advantage of these comments and improve the quality of the article accordingly.

Dr. Guha has published scholarly research which seems to be relevant to this Wikipedia article:


 * Reference : Brishti Guha, 2012. "Grandparents as Guards: A Game Theoretic Analysis of Inheritance and Post Marital Residence in a World of Uncertain Paternity," Working Papers 37-2012, Singapore Management University, School of Economics.

ExpertIdeasBot (talk) 16:34, 19 May 2016 (UTC)

Racial bias in Permstrump's editing?
Hi Permstrump. I have been observing your editing activities for a while, and it appears to me that you are systematically deleting demographic parameters of non-white populations from this (and possibly other) Wikipedia pages. You give various reasons for your deletions from time to time, (copyright infringement etc.), but the final outcome of your edits is always the same: non-white demographic values are deleted. I am notifying Doc James and am considering to report you for racial bias to a Wikipedia administrator, but would first like to invite you to comment.86.154.102.134 (talk) 10:48, 4 June 2016 (UTC)
 * What has been going on here? I looked at the talkpage because from the introduction it was immediately clear that the article had been "sanitised" with weasel wording. Apparently it has not recovered from the March 2016 event? Why was the deleted material not restored? The rate estimates are the essential core of this article.
 * Clearly, rates are going to vary substantially between cultures, in general, pre-agricultural and post-agricultural societies are much more promiscuous, so that we expect low rates in agricultural societies, and high rates everywhere else (to the point where, in some non-agricultural societies, there had never been any attempt to identify fathers, the entire concept presupposes Western-style genealogy where there is a tradition of paternity to be verified in the first place). --dab (𒁳) 08:50, 13 July 2017 (UTC)