Talk:Same-sex parenting: Difference between revisions
TheArmadillo (talk | contribs) |
|||
Line 532: | Line 532: | ||
I think the Legal Issues section had some important linkages to related topics. Its [https://en.wikipedia.org/w/index.php?title=LGBT_parenting&diff=580326298&oldid=580326211 deletion] wasn't discussed and may have been accidental. Would anyone object to restoring it?--[[User:Trystan|Trystan]] ([[User talk:Trystan|talk]]) 22:28, 5 November 2013 (UTC) |
I think the Legal Issues section had some important linkages to related topics. Its [https://en.wikipedia.org/w/index.php?title=LGBT_parenting&diff=580326298&oldid=580326211 deletion] wasn't discussed and may have been accidental. Would anyone object to restoring it?--[[User:Trystan|Trystan]] ([[User talk:Trystan|talk]]) 22:28, 5 November 2013 (UTC) |
||
:I think it might have been a mistake. {{ping|TheArmadillo}}? –[[User:Roscelese|Roscelese]] ([[User talk:Roscelese|talk]] ⋅ [[Special:Contributions/Roscelese|contribs]]) 00:53, 6 November 2013 (UTC) |
:I think it might have been a mistake. {{ping|TheArmadillo}}? –[[User:Roscelese|Roscelese]] ([[User talk:Roscelese|talk]] ⋅ [[Special:Contributions/Roscelese|contribs]]) 00:53, 6 November 2013 (UTC) |
||
::I accidentally duplicated a section somehow, and I think when the duplicate was deleted (by someone else) they accidentally took out the Legal Issues section as well. [[User:TheArmadillo|TheArmadillo]] ([[User talk:TheArmadillo|talk]]) 03:15, 6 November 2013 (UTC) |
Revision as of 03:15, 6 November 2013
This is the talk page for discussing improvements to the Same-sex parenting article. This is not a forum for general discussion of the article's subject. |
Article policies
|
Find sources: Google (books · news · scholar · free images · WP refs) · FENS · JSTOR · TWL |
Archives: Index, 1, 2, 3, 4, 5, 6, 7, 8Auto-archiving period: 90 days ![]() |
This is the talk page for discussing improvements to the Same-sex parenting article. This is not a forum for general discussion of the article's subject. |
Article policies
|
Find sources: Google (books · news · scholar · free images · WP refs) · FENS · JSTOR · TWL |
Archives: Index, 1, 2, 3, 4, 5, 6, 7, 8Auto-archiving period: 90 days ![]() |
![]() | This article is of interest to the following WikiProjects: | |||||||||||||||||||||||
Please add the quality rating to the {{WikiProject banner shell}} template instead of this project banner. See WP:PIQA for details.
Please add the quality rating to the {{WikiProject banner shell}} template instead of this project banner. See WP:PIQA for details.
|
New Study on Education Outcomes
Douglas Allen with Simon Fraser University has released a new study on high school graduation rates by family type. He controls for a large number of variables, including variables which have not been included in previous models, usually because of a lack of data. He uses an enormous Canadian dataset, from the 2006 census. It has the advantage of using self-reported data for same-sex couples, and it is able to distinguish children who are actually the sons and daughters of same-sex couples from children who live in the same household as same-sex couples, which has been a problem with US data. I've read the study, and the methodology is good (they use a logit multiple regression of high school graduation for 17-22 year olds on twenty or so variables). The only notable drawback is one the authors mentioned; the census did not distinguish same-sex married couples and same-sex common-law couples with children, so they weren't able to separate them. However, many of the conclusions do not depend on this distinction. This is the best analysis I have seen of the best dataset.
Someone who hadn't read the study deleted it, but I reverted the deletion; you would think you would have to be quite familiar with a study to either post it or delete it.
- Allen, Douglas (December 2013). "High school graduation rates among children of same sex households" (PDF). Review of Economics of the Household. 11 (4). Springer Science and Business Media: 635–658. doi:10.1007/s11150-013-9220-y. Retrieved 26 Oct 2013.
{{cite journal}}
: CS1 maint: year (link)
Update: It was deleted again. Let's have this debate, I suppose. 68.144.160.191 (talk) 06:20, 27 October 2013 (UTC) Update 2: I noticed that the Rosenfeld paper from 2010 has not been removed, even though the author removed sample points from the dataset for the reason that the families had recently moved. He might have just controlled for the variable. Apparently it suited his conclusions better to remove the offending data. I truly hope that bias isn't clouding anyone's judgment here. May we fairly represent the literature without promoting a particular point of view for personal reasons. 68.144.160.191 (talk) 06:25, 27 October 2013 (UTC)
- Actually, Rosenfeld has talked about this in a response, also published in Demography, to another paper of Allen's where some of the same methodological flaws were present. Rosenfeld deliberately controlled for family upheavals as best he could, because if you don't control for that sort of relevant factor, you really aren't testing whether or not the sex of parents affects child outcomes at all. I can quote from the reply if anyone wants, but basically, as we probably all know, same-sex families are more likely to have the children be adopted (incl. later in life) or fostered, and lumping those children in with children who have grown up with the parents since birth or a very young age means that you're attributing to the same-sex parents the factors that led them to end up with the same-sex parents, even if their progress afterwards improved. –Roscelese (talk ⋅ contribs) 15:07, 27 October 2013 (UTC)
- Rosenfeld did not "control" for family upheavals. Controlling in a regression means that you include the variable in the model. Rosenfeld actually removed observations from the dataset. This is not a good practice in linear modeling. Allen controlled for the variable by actually including it. In my opinion, which I won't enforce in this case, this could be a methodological flaw serious enough to justify removing the reference to Rosenfeld's paper.136.159.142.129 (talk) 18:18, 28 October 2013 (UTC)
- Actually, Rosenfeld has talked about this in a response, also published in Demography, to another paper of Allen's where some of the same methodological flaws were present. Rosenfeld deliberately controlled for family upheavals as best he could, because if you don't control for that sort of relevant factor, you really aren't testing whether or not the sex of parents affects child outcomes at all. I can quote from the reply if anyone wants, but basically, as we probably all know, same-sex families are more likely to have the children be adopted (incl. later in life) or fostered, and lumping those children in with children who have grown up with the parents since birth or a very young age means that you're attributing to the same-sex parents the factors that led them to end up with the same-sex parents, even if their progress afterwards improved. –Roscelese (talk ⋅ contribs) 15:07, 27 October 2013 (UTC)
- I'm not going to touch on the methodology, I'll leave that to others, but I just wanted to comment on the denial here of reliability concerns regarding the source/author. "There are no sources saying that Allen works for NOM..."? Erm...http://www.ruthinstitute.org/pages/boardMembers.html ChiZeroOne (talk) 09:50, 27 October 2013 (UTC)
- So what if he does? a lot of the other authors are in same-sex relationships (Gartrell, e.g.) jj (talk) 13:54, 27 October 2013 (UTC)
- Uh, you are going down a road you don't want to go down by stating that someone's sexual orientation makes them an unreliable source, as opposed to (from what it appears) literally being paid to draw certain conclusions. –Roscelese (talk ⋅ contribs) 15:07, 27 October 2013 (UTC)
- Let us judge on the merits, not on who says them. I was not suggesting ones sexual orientation makes her or him an unreliable source. jj (talk) 23:15, 27 October 2013 (UTC)
I'm the original IP. I dug up my old account so I could post with an identity on here. It looks like no consensus has been reached, but it doesn't seem to be in dispute that the Allen study has the best dataset. I have not seen any particular criticisms of Allen's methodology (other than my own, above). His paper appears to have been peer-reviewed for inclusion in the journal. The points against Allen's paper are that it is not endorsed by the usual secondary sources, and that the author is an associated expert for an organization associated with NOM (I stand corrected on that point; I made a good faith effort to find a source for the claim but didn't find one at first). Are those who oppose my position saying that methodology and data collection are not really relevant, and that we generally only accept studies endorsed by secondary sources? If that's the case, should we review *all* of the papers referred to in this article? I don't have a vested interest in this topic; I endorse the legality of gay marriage and child adoption, but as someone with experience in social science regression methods, I feel compelled to defend a paper whose methodology and data appear superior to other papers in the field. TheArmadillo (talk) 22:19, 28 October 2013 (UTC)
- Methodology is very relevant, and that's one of the things the study has been criticized for! If there are other studies in the article that diverge from scientific consensus due to poor methodology, bring it up and we might well see fit to remove those as well. –Roscelese (talk ⋅ contribs) 22:36, 28 October 2013 (UTC)
- Why don't you bring up a list of the methodological problems; I haven't seen any cited besides the one I mentioned above. You mentioned sampling but I'm not aware of any sampling problems with Allen's paper. I've already referred to a major problem with the Rosenfeld paper (dropping observations instead of controlling for variables). TheArmadillo (talk) 22:58, 28 October 2013 (UTC)
- Agree with TheArmadillo, especially since the relevant section on school outcomes has only primary sources, so we should show all of them fairly. jj (talk) 14:29, 29 October 2013 (UTC)
- Why don't you bring up a list of the methodological problems; I haven't seen any cited besides the one I mentioned above. You mentioned sampling but I'm not aware of any sampling problems with Allen's paper. I've already referred to a major problem with the Rosenfeld paper (dropping observations instead of controlling for variables). TheArmadillo (talk) 22:58, 28 October 2013 (UTC)
Proposition: we include both the Rosenfeld paper and the Allen paper in the article, but make the Allen section I wrote briefer to reduce emphasis. We also include the main criticism of each paper: for Rosenfeld, he dropped data instead of controlling for variables; for Allen, the census he used did not distinguish married and common-law same-sex couples. We say that there is little literature on the subject and no clear consensus has yet been reached. I think this is fair, and perhaps even generous to Rosenfeld considering his methodology. To some, it may be generous to Allen based on his affiliation. Overall, I think it is a balanced representation of the current literature. Let's try to reach a consensus by the time the page protection is lifted. TheArmadillo (talk) 16:44, 29 October 2013 (UTC)
- Support In an area of the article with no secondary sources, this is fair. jj (talk) 17:32, 29 October 2013 (UTC)
- Oppose; NPOV does not require "balance". Primary-source studies that contradict scientific consensus due to obviously poor methodology by authors with an obvious financial and political agenda do not belong in our article. If you think you can demonstrate similar problems with Rosenfeld, you might gain consensus to remove it, but Allen should not be added. –Roscelese (talk ⋅ contribs) 14:52, 30 October 2013 (UTC)
- Question: Where do you get the "contradict scientific consensus" portion of your rule? There are a lot of papers that are cited here that have "obviously poor methodology by authors with an obvious financial and political agenda," For example:
Gartrell and Bos's 25-year longitudinal study, published 2010, was limited to mothers who sought donor insemination and who may have been more motivated than mothers in other circumstances.[42] Gartrell and Bos note that the study's limitations included utilizing a non-random sample, and the lesbian group and control group were not matched for race or area of residence. The study was supported by grants from the Gill Foundation, the Lesbian Health Fund of the Gay and Lesbian Medical Association, Horizons Foundation, and the Roy Scrivner Fund of the American Psychological Foundation.[35]
- To the methodology issue: you have done nothing to prove Allen's methodology was flawed since The Armadillo's request for a statement on that at 22:58 10/28 UTC (above). Indeed, the "obviously poor methodology" seems solely the lack of being picky enough about which children are included, but that has been disputed by Armadillo by saying that the proper method is to control for variables (as Allen did), not dropping data (as Rosenfeld did). That statement needs a response.jj (talk) 18:02, 30 October 2013 (UTC)
- Perhaps removing both is an option, but there is no scientific consensus on educational outcomes as far as I am aware. Even if there were a consensus, and the Allen paper contradicted it, you have not given any evidence that it contradicts the prevailing view because of "obviously poor methodology". It isn't obvious; in fact by comparison the Rosenfeld paper it has superior methodology, as I have said already. Perhaps you've just forgotten to mention your methodological issues with the paper. I would be very open to hearing them. TheArmadillo (talk) 18:45, 30 October 2013 (UTC)
- Update: Based on what JasonJack said, I must have misunderstood what the concerns with the methodology are. I didn't realize that Roscelese was actually criticizing the fact that Allen controlled for mobility instead of dropping data. I'm very confident, to reiterate my position, that Allen's approach is the correct one. Another approach would be to run two separate regressions on each population of children, but this would introduce new problems. Rosenfeld did not even do this. Controlling for variables is exactly what regression is designed to do. Dropping data biases our estimates and makes standard errors larger, leading to less significant results. This is not a good practice. While I would respect disagreement, I really think this is quite a simple matter; there is no problem with Allen's method in this respect. TheArmadillo (talk) 21:52, 30 October 2013 (UTC)
- I'm no longer entirely sure what you're arguing, but I've already explained, above, one of the problems with Allen's methodology is this: He failed to control for a pretty central concern, that is, family stability, and when you don't account for that - when totally by coincidence you have one group that's made up of more unstable families and one group that's made up of more stable families - you're obviously going to get skewed results because those results are going to be based more on stability than on sexes of parents. Allen claims that this improves his study, but all it means is that he's sampling from groups that are disparate in more ways than the variable he's studying. Quite backwards. As I also stated above, Rosenfeld pointed out these flaws in Demography. One relevant excerpt: "In their revision of my analysis, Allen et al. preferred to analyze the outcomes of all children, regardless of how long they had lived with their current families. Allen et al. therefore attributed to the current family (at the time of the census) child outcomes that may have been produced years before the current family was formed. Allen et al. violated a fundamental rule of causal order, which is that later characteristics ought not be used to predict earlier events." Also, per my mention above of adopted and foster children, "Allen et al. reached the conclusion that children in same-sex-couple families fare worse in school by including all children regardless of how long the child has lived with the family (see their Models 2 and 4) and by including adopted and foster children along with the head of household’s own children (their Models 3 and 4). Allen et al.’s finding of worse school performance by children living with same-sex couples is due to their conflating the initial disadvantage of children who come into same-sex couple families (a disadvantage that appears to be substantial) with the progress children experience during the time when they are actually being raised by same-sex couples (progress that is excellent)." –Roscelese (talk ⋅ contribs) 01:07, 31 October 2013 (UTC)
- What you quoted of Rosenfeld from 2012 was in response to an earlier Allen study, not the one in question (2013), so maybe that's where your mistake is. This most recent paper was very recently released, and took those criticisms into account. For one thing, the new Canadian dataset directly links children to parents. For another, Allen explicitly controlled for family stability and marital history. While it is certainly *possible* that there are other omitted variables for which we don't have data (meaning there should be further study in this field), as is often unavoidable in linear models, it is most certainly not justified to discard approximately half the data, as Rosenfeld did. Allen's approach is a very reasonable one given the datasets we have. TheArmadillo (talk) 01:21, 31 October 2013 (UTC)
- Well, not quite; Allen checks whether the child has moved but not whether they moved with the parents. That's part of the family stability thing. (Allen cites his own earlier study in the new one.) I may as well link you a better organized analysis of Allen, since why bother to copy everything here: [1] –Roscelese (talk ⋅ contribs) 01:53, 31 October 2013 (UTC)
- Fair enough, but lacking data on whether children moved with their parents is a shortcoming of the data, and not a methodological flaw (I suppose, however, that if he were really good, he would find some inventive way to get around the problem). It still isn't justified to just drop half the relevant data. To be fair, after reading that blog post, Allen's method of determining graduation rates is a bigger methodological problem (I don't remember how Rosenfeld did it), although we don't have a prior reason to think that his method would be biased against either LGBT or non-LGBT families. Overall, the literature on this topic is really inconclusive. Allen brings the best dataset and has the most controls, but we need more data or a creative way of measuring educational attainment to come to anything solid. If we keep the section on school outcomes, I think we have to characterize the question as underdeveloped. TheArmadillo (talk) 02:18, 31 October 2013 (UTC)
- Well, not quite; Allen checks whether the child has moved but not whether they moved with the parents. That's part of the family stability thing. (Allen cites his own earlier study in the new one.) I may as well link you a better organized analysis of Allen, since why bother to copy everything here: [1] –Roscelese (talk ⋅ contribs) 01:53, 31 October 2013 (UTC)
- What you quoted of Rosenfeld from 2012 was in response to an earlier Allen study, not the one in question (2013), so maybe that's where your mistake is. This most recent paper was very recently released, and took those criticisms into account. For one thing, the new Canadian dataset directly links children to parents. For another, Allen explicitly controlled for family stability and marital history. While it is certainly *possible* that there are other omitted variables for which we don't have data (meaning there should be further study in this field), as is often unavoidable in linear models, it is most certainly not justified to discard approximately half the data, as Rosenfeld did. Allen's approach is a very reasonable one given the datasets we have. TheArmadillo (talk) 01:21, 31 October 2013 (UTC)
- I'm no longer entirely sure what you're arguing, but I've already explained, above, one of the problems with Allen's methodology is this: He failed to control for a pretty central concern, that is, family stability, and when you don't account for that - when totally by coincidence you have one group that's made up of more unstable families and one group that's made up of more stable families - you're obviously going to get skewed results because those results are going to be based more on stability than on sexes of parents. Allen claims that this improves his study, but all it means is that he's sampling from groups that are disparate in more ways than the variable he's studying. Quite backwards. As I also stated above, Rosenfeld pointed out these flaws in Demography. One relevant excerpt: "In their revision of my analysis, Allen et al. preferred to analyze the outcomes of all children, regardless of how long they had lived with their current families. Allen et al. therefore attributed to the current family (at the time of the census) child outcomes that may have been produced years before the current family was formed. Allen et al. violated a fundamental rule of causal order, which is that later characteristics ought not be used to predict earlier events." Also, per my mention above of adopted and foster children, "Allen et al. reached the conclusion that children in same-sex-couple families fare worse in school by including all children regardless of how long the child has lived with the family (see their Models 2 and 4) and by including adopted and foster children along with the head of household’s own children (their Models 3 and 4). Allen et al.’s finding of worse school performance by children living with same-sex couples is due to their conflating the initial disadvantage of children who come into same-sex couple families (a disadvantage that appears to be substantial) with the progress children experience during the time when they are actually being raised by same-sex couples (progress that is excellent)." –Roscelese (talk ⋅ contribs) 01:07, 31 October 2013 (UTC)
- Update: Based on what JasonJack said, I must have misunderstood what the concerns with the methodology are. I didn't realize that Roscelese was actually criticizing the fact that Allen controlled for mobility instead of dropping data. I'm very confident, to reiterate my position, that Allen's approach is the correct one. Another approach would be to run two separate regressions on each population of children, but this would introduce new problems. Rosenfeld did not even do this. Controlling for variables is exactly what regression is designed to do. Dropping data biases our estimates and makes standard errors larger, leading to less significant results. This is not a good practice. While I would respect disagreement, I really think this is quite a simple matter; there is no problem with Allen's method in this respect. TheArmadillo (talk) 21:52, 30 October 2013 (UTC)
- Roscelese, I hope we haven't mischaracterized your methodology concerns. It seems that the issue is excluding certain children entirely or controlling for possible factors (such as stability). It seems like either is plausible. Frankly, when you compare it to the sample size issues in the other papers, this is a much smaller issue (as a small sample size makes it less likely to find statistical significance). jj (talk) 20:25, 30 October 2013 (UTC)
- Question: Where do you get the "contradict scientific consensus" portion of your rule? There are a lot of papers that are cited here that have "obviously poor methodology by authors with an obvious financial and political agenda," For example:
Alternate Proposal
Proposition in the alternative: to remove all references to Rosenfeld's paper because adding in one but not another primary source on the same topic is unfair.jj (talk) 18:02, 30 October 2013 (UTC)
- Comment: This would mean excluding three of the only studies that had a random sample (Rosenfeld and both Allen studies), but it is what the rules articulated above would lead. jj (talk) 18:10, 30 October 2013 (UTC)
- Conditional Support: I would support this as the next-best option if we cannot come to agreement to include both papers while mentioning their flaws. However, I think this would be a very unfortunate arrangement, and would deprive readers of the opportunity to read the authors' relevant findings. TheArmadillo (talk) 18:45, 30 October 2013 (UTC)
Consensus
The debate has stalled, and apparently we haven't got a consensus. Roscelese, please follow up on the points made above a few days ago. If you would rather not take part, let others come to their own consensus and respect that on the article page. TheArmadillo (talk) 16:43, 5 November 2013 (UTC)
- And please comment; we don't want to be reverted, but if there aren't dissent there's an illusion of consensus and then we are reverted again. No one objected to removing Rosenfeld and yet it was reverted. jj (talk) 16:50, 5 November 2013 (UTC)
Question
Why is there so much faith put in the APA and ASA in order to exclude other points of view? It seems to me that we now have four peer-reviewed papers in the last 18 months that contradict the consensus. It seems like we need to be especially careful to acknowledge that this is part of a larger political debate and, as such, the conclusions either side takes are politically motivated. jj (talk) 23:25, 27 October 2013 (UTC)
- I understand fully that the rules prefer secondary sources. My point is since politics is intermingled, the secondary sources are more likely to ignore sources against their point of view, so we should re-consider how we approach using the APA. jj (talk) 23:30, 27 October 2013 (UTC)
- The two Allen studies have looked at more children of same-sex couples than all the other studies combined have. Further, the Regnerus study also provides insight into what (unstable or otherwise) same-sex couples look like, which is very important. jj (talk) 23:54, 27 October 2013 (UTC)
The Regnerus study provides nothing. The Regnerus study did not specifically examine children raised by same-sex couples, and provides no support for the conclusions that same-sex parents are inferior parents or that the children of same-sex parents experience worse outcomes. So long as an adult child believed that he or she had had a parent who had a relationship with someone of the same sex, then he or she was counted by Regnerus as having been “raised by” a parent in a same-sex relationship.
Regnerus’s first published analysis of his research data failed to consider whether the children lived with, or were raised by, the parent who was, at some point, apparently involved in “a romantic relationship with someone of the same sex” and that same-sex partner.
Regnerus categorized children as raised by a parent in a same-sex romantic relationship regardless of whether they were in fact raised by the parent and the parent’s same-sex romantic partner and regardless of the amount of time that they spent under the parent’s care. When the Regnerus study compared the children of parents who at one point had a “same-sex romantic relationship,” most of whom had experienced a family dissolution or single motherhood, to children raised by two biological, married opposite-sex parents, the study stripped away all divorced, single, and stepparent families from the opposite-sex group, leaving only stable, married, opposite-sex families as the comparison. (the comparison group consisted of individuals who "lived in intact biological families (with mother and father) from 0 to 18, and parents are still married at present”). Thus, it was hardly surprising that the opposite-sex group had better outcomes given that stability is a key predictor of positive child wellbeing. By so doing, the Regnerus study makes inappropriate apples-to-oranges comparisons. More here: http://www.asanet.org/documents/ASA/pdfs/12-144_307_Amicus_%20%28C_%20Gottlieb%29_ASA_Same-Sex_Marriage.pdf (American Sociological Association) http://pediatrics.aappublications.org/content/early/2013/03/18/peds.2013-0376.full.pdf+html (The American Academy of Pediatrics) Anyway, his study shows flaws that even a layperson can spot. — Preceding unsigned comment added by 188.167.186.203 (talk) 07:30, 28 October 2013 (UTC)
- And that's a critique that demonstrates exactly what I acknowledged when I said (un-stable or otherwise). It does not answer my huge issue with not including Allen and pretending that the critiques of the majority are "misrepresentations" when Marks and Rosenfeld both are peer-reviewed studies that say such critiques are genuine. jj (talk) 13:18, 28 October 2013 (UTC)
Your problem is you dont understand point of this study. Regnerus claims that study is about same-sex COUPLES and PARENTING but he doesnt have information about same-sex couples. His data: Respondents who: Had a parent in a same-sex relationship - "Lesbian" mothers: 175; "Gay" fathers: 73 Lived with parent’s same-sex partner more than 3 years - "Lesbian" mothers: 40; "Gay" fathers: 1 Came from “planned” gay families : "Lesbin" mothers: 30 – 45; "Gay" fathers: less than 1 And he categorized children as raised by a parent in a same-sex romantic relationship regardless of whether they were in fact raised by the parent and the parent’s same-sex romantic partner (meaning: my divorced bio father had same sex relationship and I lived with my hetero mother and hetero step-father or single hetero mother but I am in group of "gay fathers"! ) But in this study, only 57% said they had lived with their mother and her partner for at least four months before the age of 18, and only 23% reported living with their father and his partner for the same length of time. Only 23% of LM children and 2% of GF children reported living with their parents and their parents’ same-sex partners for three years or more! And when looking at the outcomes of those children, we are being led to believe that those outcomes are in some way related to the short amounts of time that those children spent with their gay or lesbian parents while in a same-sex relationship, and not the fifteen-plus years the vast majority of them spent outside of that dynamic. The illogic behind this comparison is mind-boggling. And to "unstable" "same-sex couples": He enlarged his LM and GF groups by lumping together a mishmash of overlapping characteristics into two messy samples(single parent, divorced parents...)The other six categories are relatively homogenous for straight people, but Regnerus’s enlargement of LM and GF groups makes them deliberately heterogeneous. And now having done this, he’s about to compare two deliberately heterogeneous categories (LM and GF) to a deliberately homogeneous category. And Marks, Allen... - Roscelese says: "... as opposed to (from what it appears) literally being paid to draw certain conclusions." Touché! But APA, ASA, AAP are not lobby groups.... — Preceding unsigned comment added by 188.167.186.203 (talk) 14:52, 28 October 2013 (UTC)
- You say the ASA is not a lobby group but you also cite a amicus brief by the ASA designed to overturn traditional marriage laws. Exactly my problem. And Allen's studies went through peer review. jj (talk) 16:36, 28 October 2013 (UTC)
- Filing an amicus brief doesn't by any stretch of the imagination make an organization a lobby group. In fact, there are lots of organizations that engage in actual lobbying but aren't lobby groups. The APA and ASA are wholly mainstream professional scientific organizations whose published views reflect consensus among scholars in their respective fields. For our purposes, they are top-notch secondary sources. We can debate the merits and flaws of a given study till we're blue in the face, and that gets us nowhere. That's why we prefer secondary sources (such as APA and ASA) which have taken the time and the trouble to sift through the various studies and arrive at conclusions based not on the work of any one researcher or one study but rather on the overall state of research in the field. Rivertorch (talk) 17:30, 28 October 2013 (UTC)
As the IP who was edit warring said, that Allen study has perhaps the best data set in the article-- the only one comparable to it is the one by Rosenfeld, which is included. I would think we should either eliminate the school outcome section altogether or include the two allen papers and the two Rosenfeld papers--right not we only have one Rosenfeld.
And when you say to the IP, "push your POV on Conservapedia"-- What are you looking for? There are no secondary sources in the school outcomes section, so we should either gut that section or add Allen. WP:NPOV. jj (talk) 20:10, 28 October 2013 (UTC)
- Unfortunately, with the block in place, we can neither remove primary sources if we're shifting to secondary nor add Allen. I'm not saying the earth is flat, but the reaction to the IP's edits seem to be as if I was. jj (talk) 20:19, 28 October 2013 (UTC)
- I'm trying to be both civil and advocate my position well. jj (talk) 20:21, 28 October 2013 (UTC)
- IP 188 has already explained some of the methodological issues, but I'm also going to respond to one of the other claims in the first comment in the thread: JasonJack, if you don't think the APA and ASA have any more authority on this issue, you need to not be editing medical and social science articles. –Roscelese (talk ⋅ contribs) 14:55, 30 October 2013 (UTC)
- Maybe I am misunderstanding, but it appears as if IP 188 only criticized the methodology of Regnerus, and not Allen. With Allen IP 188 only referred to the potential conflict of interest. TheArmadillo (talk) 18:54, 30 October 2013 (UTC)
Up-to-date reviews and meta-analyses produced by the largest and most influential medical organizations like the APA are exactly the kinds of sources WP:MEDRS says we should prefer. There are a lot of primary sources that are peer-reviewed and published in journals but yet are still garbage. I'm saying this is true generally, not at all limited to the primary sources here. At some point in recent history there was agreement that we shouldn't be using primary sources in this article at all because we have high-quality authoritative secondary sources. My view is that we should follow through on that and remove all the primary sources. Zad68
15:02, 30 October 2013 (UTC)
- This is an emerging field of study within sociology. Similar articles often use primary sources where there is a gap in the established secondary literature. With respect to educational outcomes, I think the best policy is to refer to the most prominent primary sources (Rosenfeld, Allen) in the absence of secondary sources. However, I would tentatively support removal of all primary sources for the sake of consistency (which is necessary to keep bias in check), even though it might hurt the article. TheArmadillo (talk) 18:54, 30 October 2013 (UTC)
- Suggest we focus this on the two vote/discussions above. jj (talk) 20:30, 30 October 2013 (UTC)
Difference between boys and girls
The Allen paper that was recently uploaded is fascinating; I went to the trouble of reading it, as well as the Rosenfeld paper it criticized and the previous Allen paper that was less thorough. I realize there is controversy here, but the paper makes an extraordinary distinction between the daughters and sons of same-sex couples. It doesn't seem as if the rest of the article addresses this. Is there any other literature examining the difference between boys and girls raised by same-sex couples? 136.159.142.129 (talk) 18:46, 28 October 2013 (UTC)
- No, there isn't, unless someone proves my knowledge wrong. Allen's survey demonstrates that. We should include the Allen paper on the grounds the IP listed alone. jj (talk) 20:12, 28 October 2013 (UTC)
Edit request
Could you change the map to File:LGBT adoption in the world.png, as in our other articles? I recently updated the map here, as it was several years out of date, but it appears that only the other map is being maintained by anyone else. — kwami (talk) 20:21, 28 October 2013 (UTC)
Also, since it looks like this article is getting some attention, it would be a good idea for someone else to review the map. I have no idea what I'm doing, but I removed Cambodia and the Philippines because I couldn't verify them. If those were errors, then they sat undetected for years, and there could be more; if they were correct, then our sources are horrible and I probably made other mistakes. — kwami (talk) 20:24, 28 October 2013 (UTC)
Allen
Why are people including Rosenfeld but not Allen? Both have methodological flaws and the only distinguishing characteristics between them are original research. jj (talk) 18:16, 4 November 2013 (UTC)
- No one ever responded to my last edits in the debate above a few days ago. I sort of thought we had settled on either keeping both or removing both. Surprised that certain users are insisting on keeping Rosenfeld and shutting out Allen without justification. Don't we all agree that both have fairly significant flaws, and that we prefer secondary sources in any case? And that we should be consistent with respect to primary sources? TheArmadillo (talk) 05:45, 5 November 2013 (UTC)
- According to the discussion above, you explicitly failed to gain consensus to add Allen. There also doesn't seem to be consensus to remove Rosenfeld at the moment, though that may change with Z's review of the primary sources. Do not edit war, and do not insert inferior content to make a WP:POINT. This is disruptive. –Roscelese (talk ⋅ contribs) 16:51, 5 November 2013 (UTC)
- You didn't respond for five days; the discussion was left at agreement between me and another user. You can understand why we thought there was some sort of consensus. Nevertheless, sure, let's wait on this review of primary sources. TheArmadillo (talk) 20:08, 5 November 2013 (UTC)
- I'd already stated my position. Did you think that my failure to re-state it meant that I was leaning back from my keyboard in awe at your brilliance, unable to type? No. –Roscelese (talk ⋅ contribs) 20:10, 5 November 2013 (UTC)
- Working towards consensus should have meant you participated in the Rosenfeld removal vote, IMHO. jj (talk) 20:21, 5 November 2013 (UTC)
- I'll ignore the snarkiness. Your position did not already take into account everything that was brought up in the discussion. So yes, it seemed as if you were unable to properly respond. You still haven't properly responded. TheArmadillo (talk) 21:40, 5 November 2013 (UTC)
- Your complaining is not productive. I have made my position clear. –Roscelese (talk ⋅ contribs) 00:52, 6 November 2013 (UTC)
- I'd already stated my position. Did you think that my failure to re-state it meant that I was leaning back from my keyboard in awe at your brilliance, unable to type? No. –Roscelese (talk ⋅ contribs) 20:10, 5 November 2013 (UTC)
- You didn't respond for five days; the discussion was left at agreement between me and another user. You can understand why we thought there was some sort of consensus. Nevertheless, sure, let's wait on this review of primary sources. TheArmadillo (talk) 20:08, 5 November 2013 (UTC)
- According to the discussion above, you explicitly failed to gain consensus to add Allen. There also doesn't seem to be consensus to remove Rosenfeld at the moment, though that may change with Z's review of the primary sources. Do not edit war, and do not insert inferior content to make a WP:POINT. This is disruptive. –Roscelese (talk ⋅ contribs) 16:51, 5 November 2013 (UTC)
Sourcing review - are any primary sources needed?
Off and on for some months now (or for at least the few months I've been involved here...) there's been a lack of consistency in the approach to the use of primary sources for this article. I thought we had agreement to remove the primary sources, but maybe not. I'd like to see if we can develop that consistency now. From my experience in working in other areas, I've come to find the use of most primary sources problematic. Generally primary sources are of widely varying quality, and it is not always easy to tell whether the findings of a primary source are noteworthy or even actually relevant to a particular article topic. Therefore, whenever they are available, secondary sources produced by independent authoritative bodies should be used in preference to and in displacement of any primary sources.
Based on this, I pulled all the sources from the article for review, using my GA tool, which is really designed for WP:MED content. The results are below, and where a source has a PMID, the source type is listed. A number of the sources here have PMIDs but some don't; someone with access to the other sources will need to review them and fill the table in. The goal is to have a review of the sourcing, identify where primary sources are used inappropriately, and update the article to use an authoritative secondary source instead, or otherwise eliminate its use. There was one comment that the studies in this area are too new to depend on secondary sources only, but I don't think that's true... it appears there have been enough published in this area that the use of a primary source should be the unusual exception. Zad68
21:07, 4 November 2013 (UTC)
- If everyone is more or less agreed that we should be selective about primary sources, I'll be supportive. I do have access to most of the sources in the table, so if I have time I'll begin to review them. TheArmadillo (talk) 05:50, 5 November 2013 (UTC)
- Sorry, a question about formatting ... I've reviewed the first source in the table and found that it meets WP:RS and is used appropriately, but I'm unsure how to reflect that in the table. TheArmadillo (talk) 20:24, 5 November 2013 (UTC)
Sourcing
Sources table
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
In this table:
|
Misrepresentation of research by opponents of LGBT rights
This insistence on fabricating entirely new claims that depart from the sources is really very tiresome. This is not "criticism of research," and it does not come from "critics of research." The ACP, for instance, was specifically founded to oppose LGBT rights. Much of the research being misrepresented has nothing to do with same-sex parenting at all, and actually, if you will read the section (shocker!) you'll see that that's exactly the problem. –Roscelese (talk ⋅ contribs) 16:54, 5 November 2013 (UTC)
- Why do we have this section? Could it be merged with the methodology section? It seems like we can keep all the info; it's just the heading that's a concern. And the critique/misrepresentation is a methodology issue. Also, Roscelese, we're awaiting your comment above. jj (talk) 17:04, 5 November 2013 (UTC)
- You got my comment above. –Roscelese (talk ⋅ contribs) 18:54, 5 November 2013 (UTC)
- Agree with removing this section entirely. It reads as arguing a position, rather than plainly describing what reliable secondary sources reviewing the topic have to say, in a non-partisan manner. It also seems like it's giving the viewpoints of organizations like NARTH far too much weight. It certainly should not be a subhead under Children's outcomes. If it were up to me, I'd have a separate new section covering the viewpoint, put the content in this subsection there, but boil it down to about a paragraph as a part of that new section.
Zad68
17:12, 5 November 2013 (UTC) - Propose we move the entire contents to the methodology section and start editing it there. Anyone opposed?jj (talk) 17:17, 5 November 2013 (UTC)
- Maybe my vision for this article is just entirely not in sync with everyone else's, but... I more or less oppose because I wouldn't have a Methodology section at all. An encyclopedia article itself shouldn't go over the minutiae of this and that study, so-and-so's methodology and this other researcher's view on that. It should just state the results. If two equally authoritative sources come to conflicting results, it should say "This, or that." If an authoritative source says one thing but a relatively minor researcher says something else, per WP:FRINGE I'm disinclined to even have the article mention the minor researcher. The more I look at it the more I think WP:TNT applies to large sections of this article. If most other editors here like the way the article is developing, great, I'll leave y'all in peace with it, but it's not what I think this article should look like.
Zad68
17:29, 5 November 2013 (UTC)
- Maybe my vision for this article is just entirely not in sync with everyone else's, but... I more or less oppose because I wouldn't have a Methodology section at all. An encyclopedia article itself shouldn't go over the minutiae of this and that study, so-and-so's methodology and this other researcher's view on that. It should just state the results. If two equally authoritative sources come to conflicting results, it should say "This, or that." If an authoritative source says one thing but a relatively minor researcher says something else, per WP:FRINGE I'm disinclined to even have the article mention the minor researcher. The more I look at it the more I think WP:TNT applies to large sections of this article. If most other editors here like the way the article is developing, great, I'll leave y'all in peace with it, but it's not what I think this article should look like.
- I think trimming the misrepresentation section could be a good idea, but various professional observations have observed (= it's not our original or POV observation) that there's misrepresentation going on of studies that don't have to do with same-sex parenting in an attempt to discredit same-sex parenting. We could sum it up in a few sentences to the effect that opponents of LGBT [rights/parenting/marriage/etc.] have taken various studies on parenting and used them to claim that same-sex couples are worse, when in fact, insert sources pointing this out here, the studies did not research same-sex parents at all and must not be used to make those claims. –Roscelese (talk ⋅ contribs) 18:54, 5 November 2013 (UTC)
- I agree with Zad68 that the article could use a good nuking and fresh pavement. The problems Roscelese points out in the Misrepresentation section are a prime example of the types of sourcing issues that plague the current version. — ArtifexMayhem (talk) 19:46, 5 November 2013 (UTC)
- Sorry, which problems I point out? –Roscelese (talk ⋅ contribs) 19:49, 5 November 2013 (UTC)
- The general problem of presenting sources that have no bearing on the topic. E.g., the quotes from the Child Trends study in that section have no bearing on LGBT parenting. Nor are they important to the point that the study has been misrepresented. However, I was reading over that section when you first posted and it looks like I transferred some of my thinking to your post. Apologies. — ArtifexMayhem (talk) 20:42, 5 November 2013 (UTC)
- Okay, I see what you're saying. Do you think we should still retain the material about parenting studies being presented as having to do with same-sex parenting, and just reduce the weight? –Roscelese (talk ⋅ contribs) 00:54, 6 November 2013 (UTC)
Protection
I've just fully protected this article due to the ongoing edit warring. Please discuss the matter here instead of continually reverting. Mark Arsten (talk) 18:41, 5 November 2013 (UTC)
Deletion of Legal Issues section
I think the Legal Issues section had some important linkages to related topics. Its deletion wasn't discussed and may have been accidental. Would anyone object to restoring it?--Trystan (talk) 22:28, 5 November 2013 (UTC)
- I think it might have been a mistake. @TheArmadillo:? –Roscelese (talk ⋅ contribs) 00:53, 6 November 2013 (UTC)
- I accidentally duplicated a section somehow, and I think when the duplicate was deleted (by someone else) they accidentally took out the Legal Issues section as well. TheArmadillo (talk) 03:15, 6 November 2013 (UTC)