yarvin-norman@cs.yale.edu (Norman Yarvin) (11/09/90)
dsstodol@daimi.aau.dk (David S. Stodolsky) writes: > A statistical >procedure is used to identify the most >knowledgeable representative of each >consensus position... Then your method is complete crap. -- Norman Yarvin yarvin-norman@cs.yale.edu "What a hell of a heaven it will be, when they get all these hypocrites assembled there!" -- Mark Twain
cox@stpstn.UUCP (Brad Cox) (11/16/90)
In article >27172@cs.yale.edu> yarvin-norman@cs.yale.edu (Norman Yarvin) writes: <dsstodol@daimi.aau.dk (David S. Stodolsky) writes: <> A statistical <>procedure is used to identify the most <>knowledgeable representative of each <>consensus position... < <Then your method is complete crap. Does anyone else find this retort as offensive as I do? -- Brad Cox; cox@stepstone.com; CI$ 71230,647; 203 426 1875 The Stepstone Corporation; 75 Glen Road; Sandy Hook CT 06482
shea@edson.East.Sun.COM (Tim Shea - Sun BOS Software) (11/17/90)
[stuff deleted...]
<dsstodol@daimi.aau.dk (David S. Stodolsky) writes:
<> A statistical
<>procedure is used to identify the most
<>knowledgeable representative of each
<>consensus position...
<
<Then your method is complete crap.
Does anyone else find this retort as offensive as I do?
Yes. It was beyond what was necessary to express an opposing
point of view.
--
Tim Shea Sun Microsystems (508) 671-0430
Boston Development Center shea@East.Sun.COM
Two Federal Street
Billerica, MA 01821
dbc@bushido.uucp (Dave Caswell) (11/19/90)
.<> A statistical .<>procedure is used to identify the most .<>knowledgeable representative of each .<>consensus position... .< .<Then your method is complete crap. . .Does anyone else find this retort as offensive as I do? . Certainly someone does. On Usenet you'll find most all opinions. On the other hand it isn't hard to see what prompted the comment. At least the retort applied to the original comment and not the person who made the comment. -- David Caswell dbc%bushido.uucp@umich.edu
xanthian@zorch.SF-Bay.ORG (Kent Paul Dolan) (11/20/90)
cox@stpstn.UUCP (Brad Cox) writes: yarvin-norman@cs.yale.edu (Norman Yarvin) writes: ><dsstodol@daimi.aau.dk (David S. Stodolsky) writes: ><> A statistical ><>procedure is used to identify the most ><>knowledgeable representative of each ><>consensus position... ><Then your method is complete crap. >Does anyone else find this retort as offensive as I do? It's more than that. Essentially starting with one person's spleen venting over USENet voting rules, this has turned from a group where collegial discussions of the political and technical challenges of designing a cooperatively maintained calendar were conducted courteously among people with strong opinions, to a hotbed of bickering. Sorry gang, I haven't seen anything worth reading here in a month; it's the Big "u" for comp.groupware. I'll check back in a few months and see if the rowdies have moved on. Kent, the man from xanth. <xanthian@Zorch.SF-Bay.ORG> <xanthian@well.sf.ca.us>
jim@ultima.socs.uts.edu.au (Jim Underwood) (11/20/90)
cox@stpstn.UUCP (Brad Cox) writes: >In article >27172@cs.yale.edu> yarvin-norman@cs.yale.edu (Norman Yarvin) writes: ><dsstodol@daimi.aau.dk (David S. Stodolsky) writes: ><> A statistical ><>procedure is used to identify the most ><>knowledgeable representative of each ><>consensus position... >< ><Then your method is complete crap. >Does anyone else find this retort as offensive as I do? Sorry Brad! It seems fair enough comment to me. "Consensus" means "agreement of opinion on the part of all concerned" (OED) How do you expect to represent this statistically ??????? ______________________________________________________________________________ Jim Underwood University of Technology, Sydney (jim@ultima.socs.uts.edu.au) PO Box 123, BROADWAY Systems Theorist N.S.W. 2007 School of Computing Sciences AUSTRALIA ______________________________________________________________________________
pease@amarna.gsfc.nasa.gov (Pease) (11/21/90)
>In article >27172@cs.yale.edu> yarvin-norman@cs.yale.edu (Norman Yarvin) writes: ><dsstodol@daimi.aau.dk (David S. Stodolsky) writes: ><> A statistical ><>procedure is used to identify the most ><>knowledgeable representative of each ><>consensus position... >< ><Then your method is complete crap. I've jumped into the middle of this thread, and so maybe the following remarks are inappropriate to what is being discussed; so please forgive... The use of what is termed the DELPHI technique for reaching a consensus was originally to have participants vote on specified set of options with a numeric value (1-5 or 1-10 being used) and the DELPHI program would use some statistical techniques for ariving at the the most likely option. The technique was then modified to have a second round of voting where participants could change their vote after seeing the results of individual votes from the first round. In this variation, for the first round the voters would also provide a value for ones expertise in the topic being considered. Thus for the second round of voting the participants might decide to modify their vote based on knowledge of how those persons who considered themselves experts on the topic voted. There was not attempt for some person(s) to decide who was expert. It was up to the individual to decide for themselves as to their own level of knowledge of the subject topic being considered. Phil Pease My witty disclaimer - everything I perceive, through either sensory or extrasensory means, has been filtered to such an extent that you had better not attempt to attribute anything I say to anyone else.
gl8f@astsun7.astro.Virginia.EDU (Greg Lindahl) (11/21/90)
In article <18637@ultima.socs.uts.edu.au> jim@ultima.socs.uts.edu.au (Jim Underwood) writes: >><Then your method is complete crap. > >>Does anyone else find this retort as offensive as I do? > >Sorry Brad! It seems fair enough comment to me. >"Consensus" means "agreement of opinion on the part of all concerned" (OED) >How do you expect to represent this statistically ??????? This is known as a "dictionary flame". The person who was talking about statistical concensus obviously has a different definition. It's constructive to ask him what it is, but it isn't constructive to call it crap. Fair has nothing to do with being constructive. Why do Usenet posters sometimes find it necessary to be twits?
jim@ultima.socs.uts.edu.au (Jim Underwood) (11/23/90)
gl8f@astsun7.astro.Virginia.EDU (Greg Lindahl) writes: ># Stuff about virtual journals and refereeing by statistical consensus - ># Followed by various strong comments and discussions as to whether ># said comments are offensive ># >This is known as a "dictionary flame". The person who was talking >about statistical concensus obviously has a different definition. It's >constructive to ask him what it is, but it isn't constructive to call >it crap. >Fair has nothing to do with being constructive. Why do Usenet posters >sometimes find it necessary to be twits? Sorry, I'm new to this. I was trying to be brief and assumed people would get the point rather than carrying on about what was or wasn't polite. I wasn't meaning to "flame" but was expressing STRONG disagreement. Is that constructive? Should it be? I interpret the argument as going something like this: "We can use the net to 'publish' in a way that avoids the delays and the restrictions and biases of god-editors of conventional journals. But much discussion in present news groups is random, undirected and of low standard. So how do we referee the virtual journal? By collecting comments from the group and somehow statistically analysing the results to define 'group knowledge'" This seems to be an example of trying to solve a political or social by defining it away - by finding some measure or technique which we call "objective" then using that measure because it is there - even although it may have nothing to do with our original values. In the case of a journal it might be "establishing truth through market research". (why not?) It seems to me that groupware will be more useful if it supports quite a different model of discussion. Draft articles would be criticised, supported, provisionally amended and otherwise discussed by named persons. (Named so that regular contributors would get an idea of each other's views and values). Eventually a small subgroup would form which would "run" with the article and beat it into its final form. Other subgroups might produce opposing articles. From time to time the discussion group would split and completely independent groups form. Other groups might merge. Of course it might be hard for authors to "own" ideas, but that's another problem. The trouble with statistical or any other automated form of moderation (the expert system editor?) is that they objectify the relationships among the group members, and these relationships are exactly what makes the group work. Bulletin boards already seem to have most of the technology to support this type of interaction but a quick browse through the newsgroups shows that it is not happening. Is it lack of interest or lack of attitude? This newsgroup seems an excellent place to continue this discussion and maybe develop some guidelines to encourage the process. It just seems to me that hey need to be developed by trial and error, not imposition. P.S. I use the dictionary because it usually describes things in about a tenth (or in this case a hundredth) of the words I can. ______________________________________________________________________________ Jim Underwood University of Technology, Sydney (jim@ultima.socs.uts.edu.au) PO Box 123, BROADWAY Systems Theorist N.S.W. 2007 School of Computing Sciences AUSTRALIA ______________________________________________________________________________
gl8f@astsun7.astro.Virginia.EDU (Greg Lindahl) (11/28/90)
In article <18657@ultima.socs.uts.edu.au> jim@ultima.socs.uts.edu.au (Jim Underwood) writes: >It seems to me that groupware will be more useful if it supports quite a >different model of discussion. Draft articles would be criticised, >supported, provisionally amended and otherwise discussed by named persons. >(Named so that regular contributors would get an idea of each other's >views and values). Eventually a small subgroup would form which would >"run" with the article and beat it into its final form. Other subgroups >might produce opposing articles. From time to time the discussion group >would split and completely independent groups form. Other groups might >merge. Of course it might be hard for authors to "own" ideas, but that's >another problem. Well, this is fine if you're looking at something as static as a typical journal. It would take quite a while for an article to see the light of day. Most usenet newsgroups, on the other hand, work on quite quick turn-around and produce some quite interesting information. >The trouble with statistical or any other automated form of moderation >(the expert system editor?) is that they objectify the relationships >among the group members, and these relationships are exactly what makes >the group work. However, if the object is to produce something like a well-moderated comp.unix.wizards newsgroup, such methods might provide a way do get the articles out quickly. Your idea of groupware would never work for such a group. Different problems have different solutions. Too bad nobody has a prototype for each.
dsstodol@daimi.aau.dk (David S. Stodolsky) (12/01/90)
jim@ultima.socs.uts.edu.au (Jim Underwood) writes: >This newsgroup seems an excellent place to continue this discussion >and maybe develop some guidelines to encourage the process. >It just seems to me that hey need to be developed by trial and error, >not imposition. I do not see where this comes from, since when does posting your ideas count as imposition? The original article was not followed up at all. This post proposes a totally new approach to the problem. I have yet to see a single good criticism. -- David S. Stodolsky Office: + 45 46 75 77 11 x 21 38 Department of Computer Science Home: + 45 31 55 53 50 Bldg. 20.2, Roskilde University Center Internet: david@ruc.dk Post Box 260, DK-4000 Roskilde, Denmark Fax: + 45 46 75 74 01