Talk:Delphi effect

From Wikipedia, the free encyclopedia

from VfD:

  • Keep. I wrote the article in question, and it's not a neologism to me. It has been quoted at me many times over the past few years, especially in the context of management training, and I was surprised to find only two references on a Google search. In the manner in which it was used, I expected to find a corpus of evidence at least as good as Myers-Briggs and Belbin. Documenting its lack of existence in a widely used reference work such as Wikipedia may help prevent management 'gurus' from treating it as fact and hence stop them from promulgating near, if not complete, falsehoods in this regard.
Eric Raymond has not claimed authorship of this concept, simply reporting what he heard, so I believe the concept to have had long and separate existence from his work "The Cathedral and the Bazaar" in at least both the USA and UK.
What would be good is if someone could reference a peer-reviewed study of the effect that shows it to be real - somehow I don't think one will be found.
Remember that 'Wikipedia is not paper', so it's not taking up needed space. I believe that documenting that something commonly believed to be true 'the Delphi Effect' is actually false helps improve the quality of the sum of human knowledge.
194.6.81.93 15:19, 5 Oct 2004 (UTC)
- More to the above, I've finally found what may be a reference here:

http://www.gov.on.ca/OMAFRA/english/landuse/facts/delphi.htm#Delphi%20Expert%20Opinion%20Method

- see the section labelled "Delphi Expert Opinion Methodology", which gives a set of references which may give someone familiar in the field enough to go on. My reading of it is that it is subtly different to the putative "Delphi Effect", but may be the kind of thing that gives (possibly incorrectly) credibility to the "Delphi Effect". 194.6.81.93 18:36, 5 Oct 2004 (UTC)
Your having finally found a reference does not inspire any confidence in me as to notability. Delete --Improv 19:40, 6 Oct 2004 (UTC)
Sorry, I did not make myself clear - I had found a link to a plethora of references, which seemed ,to me at least, adequately serious to merit notability - by following these I found the Linstone/Turoff book - then see discussion below, apparently including the original author of the Cathedral and the Bazaar. Would you agree to a forward/redirect to Delphi method with possibly short explanatory text explaining why the two are actually the same thing? 194.6.81.93 10:05, 7 Oct 2004 (UTC)
  • Delete, how this is supposed to differ from standard statistical practice is quite beyond me. A sample of size n reduces the deviation by sqrt(n). Big samples good, small samples bad. See standard error (statistics). GWO 15:39, 5 Oct 2004 (UTC)
- I think you may have missed the point (or perhaps I have). The effect is not about minimising deviation, but getting the 'best' advice or reaching the 'best' decision. It is not immediately evident that a larger group of experts will give better or more accurate advice than a small group or even a single expert. The 'Delphi effect' asserts that this is the case (larger is better), and many people accept this without challenging the basis of the statement. Having challenged the basis of the statement, I can find no evidence for it. 194.6.81.93 18:03, 5 Oct 2004 (UTC)
  • Delete. Don't put things on wikipedia for political purposes. Not notable. --Improv 17:05, 5 Oct 2004 (UTC)
    • What's up with me voting twice recently?! Argh. Sorry, folks. --Improv 18:45, 9 Oct 2004 (UTC)
- Sorry, I don't understand what is political about this - hopefully I'm not trying to proselytise, simply to point out that an apparently accepted (by at least some 'influential' people, like management trainers etc.) statement has no basis in research. Please explain what is political about this - I need educating! 194.6.81.93 18:03, 5 Oct 2004 (UTC)
  • Keep. My google search [1] found 233 references for it; the phrase does seem to have some currency. --Tagishsimon
  • Keep A consensus is more reliable if the group's members have varied experience and are not slave of peer pressure. --Pgreenfinch 18:39, 5 Oct 2004 (UTC)
  • Keep. Concept is in use, article is not pushing POV. — Gwalla | Talk 19:46, 5 Oct 2004 (UTC)
  • Article updated with references to RAND work and online book of work on the Delphi Method, which is referenced in other online histories and has extensive bibliography. I suggest the Delphi effect is a loose, informal, reformulation of the Delphi method/technique. Perhaps the deletion vote should be reconsidered/restarted in this light? WLD 21:04, 5 Oct 2004 (UTC)
  • Just found that Delphi Method has a Wikipedia entry. I suggest that the Delphi Effect article is still retained, even if only a pointer to Delphi Method. Describing it as a looser formulation...etc. would seem to me to add value as it provides a link for those researching the Delphi Effect to find the Delphi Method work. WLD 21:16, 5 Oct 2004 (UTC)
  • Keep. It may be dubious, but the concept is attested in serious work -- whether proven useful is another matter -- and might be fertile. Bill 22:40, 5 Oct 2004 (UTC)
  • Keep for reasons given above, or merge appropriately. The bulk of the info belongs on wikipeida. —siroχo 04:39, Oct 6, 2004 (UTC)
  • Comment I emailed Eric S. Raymond about this, inviting him to reply here rather than to me if he preferred. He replied as follows (his reply was dated 2004/10/06 Wed AM 06:36:13 CDT) [[User:Dpbsmith|Dpbsmith (talk)]] 16:25, 6 Oct 2004 (UTC)
I'd do so, but I can't figure out how to edit in a mode that generates the discussion timestamp.
I see that a source has been found. Congratulations, this is better than I managed! Having read the page and the discussion, I'm going to change "Delphi effect" to "Delphi method" in the upcoming 4th edition of CatB.
I do not think mine is a "looser formulation" or has dropped references to forecasting. I did use the word "prediction".
"The Shockwave Rider" came out in 1975, the same year as the Linstone/Turoff book. This is probably not a coincidence. I'd bet John Brunner read the book.
  • Comment dbpbsmith has quoted me correctly. I suggest that "Delphi effect" should forward to "Delphi method".
For the record: I myself first learned of the Delphi phenomenon from the Brunner novel. A few years later I ran across a one-paragraph reference to it which described the context of the original research at RAND (a detail that I don't believe Brunner mentioned) but did not actually explain it. When I wrote CatB, I debated with myself about whether to include a concept that I had acquired mostly from an SF novel with only thin confirmation from a single nonfiction source. I decided that since I was using it essentially as color to make a more interesting narrative about an argument well supported by other facts and cites, I could afford to take the chance - but ever since I have considered the absence of a good nonfiction cite for it in CatB to be perhaps the most serious defect in the paper. Thanks to Wikipedia I have now been able to correct this: CatB 4/e will include the URL of the Linstone/Turoff book. I cannot recall exactly (it was more than twenty years ago!) but I now suspect that scant one paragraph was itself a review or cite of Linstone/Turoff. Eric S. Raymond 16:42, 6 Oct 2004 (UTC)
  • Comment Taking Eric S. Raymond's comments on board, I'll agree that a forward to Delphi method is the right way to go. I'll merge in the references to the Linstone/Turoff book (and others) to improve that article. Unfortunately, I don't think I can achieve this within the VfD timescale - is it possible to stop/reset the clock on this? 194.6.81.93 10:05, 7 Oct 2004 (UTC)
    • Just go ahead. It shouldn't really matter. First, Delphi method isn't going to go away, so you don't need to worry about that. Second, Delphi method needs to be edited in light of this discussion, regardless. Third, Delphi effect probably won't get deleted at this point, because given the discussion, I don't think a sysop will judge that there's consensus to delete. You probably should create an account and cast a Keep vote, by the way. Finally, if by any chance Delphi effect were to be deleted, you or someone else somebody could always re-create it, entering #REDIRECT [[Delphi method]] as the initial content, with a brief explanation in the edit summary. Theoretically re-creating a deleted article is just cause for speedy deletion, but it almost certainly wouldn't happen in this case. (If it did, the appropriate thing to do would be to shrug and give up, not much time having been wasted in creating the direct). [[User:Dpbsmith|Dpbsmith (talk)]] 12:14, 7 Oct 2004 (UTC)

end moved discussion