Talk:Rorschach test#An informal survey for the inclusion of the 10 inkblots

I've been using Google's Gemini 2.5 Pro Experimental large language model to create summaries for the most popular articles with {{tl|Technical}} templates. This article, Rorschach test, has such a template in the "Population norms" section. Here is the paragraph summary at grade 5 reading level which Gemini 2.5 Pro suggested for that section:

:People disagree about what's considered a "normal" score on the Rorschach inkblot test. Some studies found the test seemed to say too many healthy people had problems, like confused thinking or trouble getting along with others. This made some people worry the test made normal people look like they had problems when they didn't, while others thought the test might be right. However, a later big study looked at test results from many different countries. It found that the number of people with high scores for things like stress or thinking problems was about what doctors would expect to see in healthy people, suggesting the test might not be finding too many problems after all.

While I have read and may have made some modifications to that summary, I am not going to add it to the section because I want other editors to review, revise if appropriate, and add it instead. This is an experiment with a few dozen articles initially to see how these suggestions are received, and after a week or two, I will decide how to proceed. Thank you for your consideration. Cramulator (talk) 12:21, 2 April 2025 (UTC)

:I thought there was a policy, across the encyclopaedia, that LLMs were not to be used to add content? Getting other editors to add material you have produced is also frowned upon? Martinevans123 (talk) 12:37, 2 April 2025 (UTC)

::I've looked through WP:LLM and I believe this is an acceptable approach. "LLMs can be used to copyedit or expand existing text and to generate ideas for new or existing articles." I am open to alternative ideas for how to accomplish the goal of helping readers understand overly technical articles. Cramulator (talk) 13:27, 2 April 2025 (UTC)

:::I believe technical articles need technical language. The sentence "{{tq|However, a later big study looked at test results from many different countries}}" doesn't fill me with confidence. But I will let others comment. Thanks. Martinevans123 (talk) 13:42, 2 April 2025 (UTC)

::::I agree with {{user|Martinevans123}}. This is an effort to "fix" a problem that doesn't exist. We don't need to dumb down scientific and medical articles so fifth graders can understand it. It risks creating inadequate (and possibly incorrect) information. There are numerous links in the article for readers who seek more information. This article has an ugly history of nonexperts trying to rewrite it. We don't need to add AI into the mix. As for use of LLM in general, this needs to be discussed by the entire Wikipedia community, not just those interested in one article. You can start at WP:Village pump. Sundayclose (talk) 15:19, 2 April 2025 (UTC)

:::::I just regenerated all 68 summary suggestions for ninth grade reading level here, because I've learned that's the target reading level for STEM articles, and I was planning to fold them in to these talk page posts before the fifth grade level summaries. I'll definitely start a Village pump discussion. Cramulator (talk) 15:43, 2 April 2025 (UTC)

::::::Aaaaaargh, have you added this on all your proposed pages? See Martinevans123 comment above. I would personally ask you to STOPPITT! - Roxy the dog 17:20, 2 April 2025 (UTC)

I am retracting this and the other LLM-generated suggestions due to clear negative consensus [https://en.wikipedia.org/w/index.php?title=Wikipedia:Village_pump_(miscellaneous)&oldid=1283927858#I_boldly_put_LLM-generated_summary_suggestions_on_the_talk_pages_of_the_68_most_popular_articles_with_Technical_templates at the Village Pump]. I will be posting a thorough postmortem report in mid-April to the source code release page. Thanks to all who commented on the suggestions both negatively and positively, and especially to those editors who have manually addressed the overly technical cleanup issue on six, so far, of the 68 articles where suggestions were posted. Cramulator (talk) 22:17, 4 April 2025 (UTC)

{{edit semi-protected|Rorschach test|answered=yes}}

I am a clinical psychologist. I request removal of Rorschach card images and attendant descriptions of what they 'are'. On this page and in the Wiki Common: https://en.wikipedia.org/wiki/Rorschach_test

The publication of these images and the analysis of the blots affects the validity of the test. The premise is that the presentation is new to the subject. So publishing these images and the attendant descriptions can ruin the usefulness of the test.

The APA, along with the psychological community, emphasizes that even though they are in the public domain, the Rorschach test images should not be published publicly or circulated outside of a clinical setting to maintain their effectiveness as a diagnostic tool. The goal is to preserve the test's validity and utility for clinical assessments. 23KMG (talk) 17:41, 18 May 2025 (UTC)

:I trust you are aware of what happened on 2009? See e.g. Talk:Rorschach test/images and Talk:Rorschach test/2009 consensus review. Thanks. Martinevans123 (talk) 18:22, 18 May 2025 (UTC)

::{{not done}}: per the above response. Day Creature (talk) 22:31, 18 May 2025 (UTC)

:::This is not to say that the question should not be discussed again. For example, since 2009 it may be that the American Psychological Association might have new or more evidence to bear. It's still not clear, however, that Wikipedia would ever agree that article content should be governed by advice from the APA. Martinevans123 (talk) 09:40, 5 June 2025 (UTC)

Beyond the discussion around public domain and image issues (which have been discussed before here), I have to say I am a bit surprised by the decision to include all of the test inkblots in the article, along with frequent responses from three different theorists for each image, and commentary for all of them as well. Given that the goal of the article is to offer a broader, more encyclopedic overview, it’s hard to see the reason in presenting the material in such a detailed, almost instructional way. It almost feels less like a panoramic review and more like a step-by-step mini-guide. Tam01 (talk) 18:58, 18 June 2025 (UTC)

:Tend to agree. The issues do greatly overlap. I believe the original APA concerns were particularly over the detailed interpretations, although mere pre-exposure to the images themselves was also a source of contention. I don't think the articles on other subjective projective tests insist on showing all of the test material. Martinevans123 (talk) 19:16, 18 June 2025 (UTC)

My two cents: The images are included in the article to say "I told you we could do it" by image proponents to those who opposed the images. And apart from the ethics of Wikipedians' blatantly destroying test validity by publishing the images, what those same Wikipedians don't seem to realize is that the "frequent responses" actually do harm to those who take the test. To the naive test-taker, giving the "frequent responses" when administered the test doesn't make the test-taker look psychologically healthy (as might be the motivation for a parent in a custody evaluation); it makes them look very pathological. All of this reflects a profound ignorance of how the test works and the vast amount of research underlying its interpretation. But as I said, that's my two cents, and I have no desire to fight a battle over the images or the listing of "frequent responses". And to the image zealots, relax. I have no intention to change anything regarding the images. You'll be wasting your time if you want to pick a fight with me about the science. I learned long ago that in some corners of Wikipedia, expertise has no meaning, even if the expert can point to the research backing his expert opinions. I'll simply ignore the comments that reflect utter ignorance and move on to more important matters. That's not a slam on Wikipedia in general, just this particular part of it. The quality of other aspects of Wikipedia, including most medical articles, is good. That's because the medical experts here worked together to create very beneficial guidelines, such as WP:MEDRS. Sundayclose (talk) 19:28, 18 June 2025 (UTC)

:Ouch. What a very cynical and bitter analysis. But my exact argument in the RfC 16 years ago. And still now, alas. Martinevans123 (talk) 19:52, 18 June 2025 (UTC)

::Given all this, how do we go from here? I'm not suggesting we remove all the inkblots or anything that drastic (as it will get reverted ASAP). But I do believe a measured reduction in that kind of detail would be a constructive next step. It might improve neutrality and better reflect how similar psychological tests are handled on Wikipedia. Curious to hear your thoughts.Tam01 (talk) 21:04, 18 June 2025 (UTC)

:::If any of the blot proponents are still around, making even a slight change on this matter could, and likely will, result in a bloodbath. It will get very ugly. You are certainly entitled to try to make changes, but I don't want any part of it. I'll quietly disappear. Sundayclose (talk) 21:27, 18 June 2025 (UTC)

:::And editors who are still watching this article and who still support inclusion of all the ink blots should comment in this thread with their arguments for keeping them. I suspect that some may comment with "where's the evidence of any harm from the APA (or BPS)?" But, as I think I tried to argue in the original RfD, this is the sort of evidence that is practically and ethically impossible to construct. Martinevans123 (talk) 21:38, 18 June 2025 (UTC)

::::I am more on the line of questioning how including all of the inkblots, along with such extensive commentary on their interpretation, aligns with an encyclopedic approach at all. It’s less about proving demonstrable harm (which, as noted, is nearly impossible to measure in this context) and more about asking whether the current presentation meets Wikipedia's encyclopedic standards. Tam01 (talk) 21:53, 18 June 2025 (UTC)

:::::Yes, that seems a perfectly valid argument and perhaps a stronger one. Martinevans123 (talk) 21:55, 18 June 2025 (UTC)

::::::For the moment, I have added Template:Excessive examples to the inkblots session. Will work on it later on. Tam01 (talk) 22:07, 18 June 2025 (UTC)

The inkblots are the test, they are not "examples" of some concept described in the article. The template therefore does not make any sense, thus I removed it. As for the excessive detail, a concise description of ten inkblots does not seem to me excessive at all; if anything, it seems quite right. But well, one can still think about splitting the description and make it more thorough in a separate article, if only that was the concern. I find it funny that y'all above are explicitly debating bad faith ways to sneak past consensus on the inkblot images. are candidly debating in the open what type of argument would stick to obtain the single end result of removing the images, judging how they might be weaker or stronger, to undermine a long standing consensus. edited --cyclopia^speak! 18:52, 19 June 2025 (UTC) --cyclopia^speak! 14:37, 19 June 2025 (UTC)

:I have edited here in perfectly good faith. Kindly amend or retract that accusation. Many thanks. Martinevans123 (talk) 14:52, 19 June 2025 (UTC)

::WP:SPADE. It's 21 years I edit here, it's not like I am born yesterday. The discussion above is (admirably) clear: it is about reducing the amount of inkblots and/or information related directly to them. As acknowledged above, the harm arguments etc. have not been accepted by consensus so far, so it seems editors above are trying to find another rhetorical line that might stick with the same end result (removing at least some of the inkblots). This exchange in particular: "It’s less about proving demonstrable harm (which, as noted, is nearly impossible to measure in this context) and more about asking whether the current presentation meets Wikipedia's encyclopedic standards." "Yes, that seems a perfectly valid argument and perhaps a stronger one." - stronger with respect to what? why comparing the two arguments in strength if not about how they manage to obtain the same desired end result? cyclopia^speak! 17:37, 19 June 2025 (UTC)

:::Kindly retract and/or strike your accusation or I will request Admin intervention. Regards. Martinevans123 (talk) 17:57, 19 June 2025 (UTC)

::::You are welcome to explain the comment exchange above. If there is a satisfactory explanation, I'll be glad to retract and apologize. cyclopia^speak! 18:27, 19 June 2025 (UTC)

:::::I am not interested in any "rhetorical line". My view is still that the images should not appear, with explanations, as they currently do, in this article. If this is not possible, the article should at least match other articles for psychological tests and simply show examples of the test materials. I believe the argument for consistency as stronger than the argument for preventing harm, as it is a boarder argument that applies across all of Wikipedia, and does not rely on the assessment of any external evidence. I do not regard expressing this logic as "{{tq|explicitly debating bad faith ways to sneak past consensus}}". Again, kindly strike or retract your accusation. I am not requesting any apology. This is the third time I have now politely asked you. If you do not, I will report you at WP:AN/I. Martinevans123 (talk) 18:48, 19 June 2025 (UTC)

::::::Thanks, I will amend the comment above. cyclopia^speak! 18:50, 19 June 2025 (UTC)

:::::::Sorry, but your "amendment" seems to have simply involved a re-write, with additional words, in very similar terms, which still ends with "{{tq|to undermine a long standing consensus}}". That is not my understanding of how consensus works and how it should be respected. Oh, and you still "find it funny"? Martinevans123 (talk) 18:59, 19 June 2025 (UTC)

::::::::I am confused. I struck the bad faith accusation. I did it sincerely, since your comment indeed made me change my mind. [https://en.wikipedia.org/w/index.php?title=Talk%3ARorschach_test&diff=1296391531&oldid=1296390799 This] indicates that you actually believe, in good faith, that what you debated above is a reasonable way of conduct. Therefore I apologize for having written and thought you were in bad faith; I was wrong. Yes, I find what you stated above and reiterated in your explanation problematic and somehow amusing in its candor and good faith; I believe we should not throw arguments and evaluate them as strategies to obtain a given end result that would also, in this case, overthrow a long standing consensus (yes, I know that consensus can change; it's the strategy I find problematic). I do not believe this is a personal attack and again, I sincerely understand you are in good faith. cyclopia^speak! 19:07, 19 June 2025 (UTC)

:::::::::Thank you for that clarification. I think it might help matters here if we all try to avoid wild personal accusations and focus more on the content arguments. I don't see very much wrong with "{{tq|candor and good faith.}}" Martinevans123 (talk) 19:11, 19 June 2025 (UTC)

::::::::::Agreed, and apologies. cyclopia^speak! 19:19, 19 June 2025 (UTC)

Thus the confusion from decades ago returns. Each blot is, in fact, an example of a test item. The original MMPI had 567 items. If we wanted to include the entire MMPI we would list all 567 items. If we wanted examples we would list only a few. Rorschach has ten example items, all of which are now in the article. Note that I'm not stating a position about whether all items (i.e., blots) should be included, nor do I intend to endlessly and futily debate that point. I'm simply clarifying what an example of a test item is. In any event, three editors now support the template. It should not be removed without consensus. Sundayclose (talk) 14:56, 19 June 2025 (UTC)

:And I also find the false accusation of bad faith unacceptable. A major problem in the debates decades ago was the false accusations of bad faith which muddled the relevant issues about inclusion of the blots. I'll add my request that the false accusation be retracted and never be done again. An apology would also be appropriate, although I wouldn't bet a penny that it will actually be offered. Sundayclose (talk) 15:07, 19 June 2025 (UTC)

::Apologize for the quite insulting sentence «The images are included in the article to say "I told you we could do it" by image proponents to those who opposed the images», and then I can begin to consider rethinking about my statements. cyclopia^speak! 17:44, 19 June 2025 (UTC)

:Including 567 items would be a bit too much. Including 10 items is quite balanced. In Solar System we discuss all the planets and other major items, but we don't list all minor planets, obviously. cyclopia^speak! 17:41, 19 June 2025 (UTC)

::How is including 10 items "balanced", when those 10 items constitute the entire set of the test material? I really don't see much similarity between the Solar System and a subjective projective psychological test. Martinevans123 (talk) 18:01, 19 June 2025 (UTC)

:::Balanced as in, it doesn't skew the article (including 567 items would make >90% of the article about the test items). Given that the items are few and can be comfortably included in a section, we can afford to cover all of them. cyclopia^speak! 18:26, 19 June 2025 (UTC)

::::I'm unconvinced that Wikipedia rules on article size should determine whether a psychological test that has a smaller number of test items should be "wholly exposed" in an article. The article has more the appearance of a practioner's guide than a selective encyclopaedic summary. I don't see why the essential concept that ambiguous random ink blots can be interpreted differently by different people can't be adequately conveyed by showing just a few examples of the test items, rather than the entire test. Martinevans123 (talk) 19:55, 19 June 2025 (UTC)

::No, there is no balance between including 10 percent of test items compared to 100 percent of test items. It's not balanced. It's overkill, to put it mildly. To my knowledge, there is no other article about a psychological test on Wikipedia that includes all test items. And that has nothing to do with copyright or ethics. It's about acceptable and poor writing style. Again, I'm not arguing about whether any or all test items should be included. I'm clarifying what exactly the words "balanced" and "example" mean to try to stem the rising confusion from discussions years ago. Sundayclose (talk) 18:04, 19 June 2025 (UTC)

:::WP:OTHERCONTENT; I am not going to check all articles about psychological tests, but most probably almost all such tests contain much more than 10 items. 10 items are a reasonable amount of material to include in full in an article; the threshold is very much subjective but let's say that if they were 15 or 20 I would find your argument much more reasonable. As it is now, the discussion of the blots in my opinion is not out of proportion given the size and breadth of the article. cyclopia^speak! 18:35, 19 June 2025 (UTC)

::::Just for the sake of a counterexample, Sexual Compulsivity Scale contains all items; they are also ten. Not that it matters much, because such a decision depends on several contextual factors, but well, it's not like this article is a lone outlier. cyclopia^speak! 18:41, 19 June 2025 (UTC)

:::::Regarding "contextual factors" I don't see that particular psychometric scale as a subjective projective psychological test that requires expert interpretation. I was thinking more of something like the Blacky pictures test which currently has no examples. Perhaps that one is "an outlier"? Martinevans123 (talk) 19:13, 19 June 2025 (UTC)

::::::Thematic Apperception Test, another widely used projective test involving visual images, has no examples. Sundayclose (talk) 21:19, 19 June 2025 (UTC)

:::::::Oh gosh. I wish I'd thought of that one (see below)... Martinevans123 (talk) 21:27, 19 June 2025 (UTC)

::::::The point is, it is not very helpful to look at other articles for guidance, especially in this case. Very few other psychological tests are like the Rorschach, AFAIK, conflating a small number of items, massive notability of the test and the way individual test items have been discussed in depth by several sources. Given that, I find the arguments about excessive detail/examples unconvincing. The inkblot paragraph now makes up less than 10% of the article text, if I measured it right - excluding references, notes etc. cyclopia^speak! 19:29, 19 June 2025 (UTC)

:::::::Perhaps the Blacky pictures test was a poor example, as it was rooted far more firmly in Freudian theories of psychosexual development and was aimed at children, both of which things don't really apply to the Rorschach. In fact, I've just noticed that this article is given :Category:Freudian psychology even though there's not one single mention of the good Professor. The Thematic Apperception Test has no examples, whereas showing just one example might be really useful. I'm guessing this may be partly because it is still protected by copyright, but that's not entirely clear at the article. Martinevans123 (talk) 19:39, 19 June 2025 (UTC)