Wikipedia talk:Wikipedia Signpost/2015-03-25/Recent research


 * The Pulmonary Medicine review is informative and, to my layman mind, persuasive that the "study" in question is misguided. Or, to use more words less precisely, a bungled hatchet job. However, the point about accuracy stuck in my mind as possibly valid.  Did the chosen WP articles actually say many things that are not true?  That's my impression of what "inaccurate" ought to mean.  Or rather, did they say things that are contrary to conventional opinion among practitioners in the relevant field?  Or is "accurate" defined according to other criteria?  Jim.henderson (talk) 21:25, 28 March 2015 (UTC)


 * Hello, Jim. Thank you for reading my review, and for your insightful questions.


 * The paper's author claims to use the modified DISCERN tool to assess accuracy. Yet most of the DISCERN questions are unrelated to accuracy. As far as I can tell, the researchers assigned a score between 1 (serious/extensive shortcomings) and 5 (minimal shortcomings) for each article's accuracy. This was determined by comparing the articles to medical textbooks. The lowest rated article was "(Nail) clubbing". As far as I can tell, the accuracy of the article looks just fine.


 * This paper (and other assessments of Wikipedia's medical content) seem to have a tacit assumption that the researchers know how to assess these articles. Unfortunately this is not the case. Axl ¤ [Talk] 20:40, 29 March 2015 (UTC)


 * Wild. Seems like the most obfuscatory depths of, say, social science. The author says he's checking whether Wikipedia is a good tool for learning, but apparently is only interested in whether it can substitute adequately for a textbook.  Well, duh!  I worked 41 years as a telephone switching technician.  Becoming good at it required attentive time in classrooms.  Also reading a bunch of books, journals and other things.  And seeking guidance from old practitioners.  Were someone to ask today how to become a good technician, I would answer, do as I did, and since we've got Wikipedia, use that, too.  But as a substitute?  What?  Who would expect to need statistical analysis to put down that idea?  We're not trying to write a textbook, or a series of textbooks, for practitioners.  We're trying to write an encyclopedia so, for example, doctors can learn something about how telephones connect to each other, or engineers can get a rough idea of the main ways lungs get sick.  When we fail at that, as we often do, we should be knocked for it.  Not when a switching technician remains clueless after trying to use Wikipedia to set a NT6X18A line card for Ground start.  Jim.henderson (talk) 13:34, 31 March 2015 (UTC)