Talk:PanelWhiz
From Wikipedia, the free encyclopedia
Dear Fellow-Editors, if you make changes, it would helpful if the original informational content were left intact. the last several edits, make the informational content meaningless. The descriptions of data structure are necessary to give enough information to social scientists to understand the functionality. Unfortunately, the last edits rendered the description meaningless. The description contained is focussed completely on a factual technical representation of the software. There is no advertising. -DeNew 14:58, 3 April 2007 (UTC)
Contents |
[edit] PanelWhiz (Link on Stata Page)
I've removed the links, both external and wiki to PanelWhiz. At the time of this comment, PanelWhiz does not appear to be notable enough to merit an article. I also see no reason for WP:EL#Advertising and conflicts of interest to have been disregarded in this situation. Finally, since Stata has been around for over 20 years, and PanelWhiz has been around for less than one year, I find it hard to believe that PanelWhiz is so integral to Stata that it deserves a mention in the Stata article. I'm willing to be convinced otherwise, and that convincing would start with the establishment of the notability of PanelWhiz via multiple non-trivial articles about PanelWhiz in reliable sources. —The preceding unsigned comment was added by Aagtbdfoua (talk • contribs) 13:00, 7 April 2007 (UTC).
- PanelWhiz - A Panel Data Front-End for Extracting Panel Data in Stata. See also PanelWhiz.
- Panel Data - Known Panel and Repeated Cross-Section Data Providers
The removed links are above. I'm also baffled why the same link was listed twice. - Aagtbdfoua 13:03, 7 April 2007 (UTC)
[edit] Support for PanelWhiz
Dear Aagtbdfoua,
As far as I know the two largest packages EVER written for Stata are GLLAMM and PanelWhiz. GLLAMM refers to a class of econometric estimators and PanelWhiz (in its previous version: SOEP Menu) refers to a methodology of retrieving data. There are NO other packages written for Stata in the world that are comparable in scope. Both packages are regularly presented at official Stata User's Meetings in the United Kingdom and Germany. In both instances, Statacorp has changed its source code in its executable to incorporate features introduced in the packages (in PanelWhiz's case: including multi-language support for variable and value labels). Might I suggest contacting the president of Statacorp (William Gould) and asking him his opinion whether PanelWhiz should be linked on the Stata wiki-page? He knows both packages very well. Might he not be a reputable person, having intimate knowledge of the functionality? An important part of Stata is that it is expanable! Users can write programs themselves. PanelWhiz contains something like 30,000 lines of Stata code. There are SEVERAL HUNDRED additional commands for Stata in the package PanelWhiz. It is the reason why PanelWhiz is also linked to Stata: PanelWhiz Link on Stata homepage. Users have been using SOEP Menu (basically PanelWhiz only for the German Panel SOEP) since 2003.
The Data Providers GSOEP and HILDA link PanelWhiz directly on their homepages:
The advantage of PanelWhiz linking to Stata and allowing a PanelWhiz wiki-article, is that you will allow users to find the package easily and help them in their research. Often the users are Masters and PhD students and other researchers. so much research is flawed, due to incorrect data work. PanelWhiz reduces these errors dramatically, as complex data sets contain typically many many files. The German household panel contains more than 200 data files. PanelWhiz handles this internally and CORRECTLY. It was presented recently at the 2007 German Stata users' meeting but already in 2003 at the 2003 UK Stata users' meeting. It has been around several years.
On page A-13 of the official scientific review of the ENTIRE institute RWI-Essen, where i work, the precursor to PanelWhiz, SOEP Menu is explicitly reviewed: "So wird am Institut beispielsweise das von einem Institutsmitarbeiter programmierte SOEPMENU, ein in das Softwarepaket STATA eingebundenes Modul zur Extrahierung von Daten aus dem SOEP, erweitert." [Link]
I hope you will think over your decision. everything in the PanelWhiz wiki-link is factually written in an attempt to help researchers use data sets easily and improve their research. several data set providers, ship their data already prepared for PanelWhiz (Australian HILDA since 2006, German SOEP starting 2007). Please ask some empirical social scientists using Stata, before you remove the PanelWhiz page.
Sincerely, DeNew 14:36, 7 April 2007 (UTC).
[edit] RFC comments
Comment via Wikipedia:Requests for comment/Maths, science, and technology#Mathematics
OK, it's very specialist, but it looks to have similar notability to many other articles in this territory: see Category:Statistical software. On that basis, I'd say it's notable enough both for inclusion and a link from the Stata page (also on the grounds above that Stata mention it on their own website).
However, there is an associated issue that needs dealing with. Is User:DeNew John P. Haisken-DeNew, the developer of PanelWhiz? If so, this steps right into the territory of Wikipedia:Conflict of interest, and that user shouldn't be editing the article directly. I notice IP edits 194.94.98.161 (talk • contribs) and 80.142.92.162 (talk • contribs) are both from the same geographical location (NRW) as the PanelWhiz company. Tearlach 06:50, 8 April 2007 (UTC)
[edit] RFC Response
I can see the point there about a *potential* for "conflict of interest". Indeed, as the obvious user-name suggests, User:DeNew, is Dr. John P. Haisken-DeNew. I am that person. I have never tried to hide my identity. My last name is "Haisken-DeNew" - i don't think was a legitimate wiki-user name, thus i chose my (shorter) birth name. I don't know of the other IP edits or what they mean. I was not aware that my editing this directly would somehow taint the text. i just happend to know a lot about panelwhiz. there have been several edits from a user named "Clappingsimon", which have been very helpful, which i do appreciate. "Clappingsimon" made several points more directly, than I could have. if the consensus is that the currently entered text is somehow tainted simply because much of it came from me, the developer, then i would remove the entire text from wikipedia and that would be the end of it. I simply think there would be loss to the social science community. i do have specialized knowledge that i would like to share with the scientific community and this is altruistically motivated.
I see the loss on the Stata page already. I have been working with Stata for 15 years. the unique strength of Stata indeed is its panel data capability. yet, this is almost completely missing in its write-up. some of the links (with a panel data dimension) were even removed. these editorial decisions are made by persons, i suspect, not familiar with Stata or its capabilities. perhaps i am wrong, but then i would understand the edits even less.
it boils down to one issue i think: if my previously contributed text is somehow "tainted" bacause it came from me (regardless of the factual and informational content), then i would remove it myself (and save you the trouble) and withdraw from this endevour completely. Please let me know, one way or another.
Sincerely, Dr. John P. Haisken-DeNew 11:52, 8 April 2007 (UTC).
p.s. a PanelWhiz company does not exist.
- if the consensus is that the currently entered text is somehow tainted simply because much of it came from me, the developer, then i would remove the entire text from wikipedia and that would be the end of it.
- There's no need for anything as drastic as that. The usual advice per WP:COI is that editors with a close connection are invaluable for their expert knowledge of sources, but should collaborate via the Talk page. It's more about the edit process being seen to be free of potential conflict of interest, rather than any implied judgement of automatic bias.
- Best move is to see what responses come via the RFC. (The decision to remove it or not is anyway collective now, since you posted it under the GFDL). Tearlach 13:06, 8 April 2007 (UTC)
thanks, this sounds quite reasonable indeed. Sincerely, DeNew 16:08, 8 April 2007 (UTC)
