It sounds good only when the bias towards Wikipedia and the US is intentional.

Please quote correctly; what I wrote was "bias towards Wikipedia and the US". For the rest you repeat after me.

On Tue, 20 Mar 2007, Gerard Meijssen wrote: He doesn't claim to be attempting to cover the breadth of Wikipedia. Implicit in the post is that it concerns Wikipedia usage in the United States. I find it is common for people to mistake limits placed on a study with a bias in the study. Let me give another example that might make this clearer. If a study concerns left handed people (one of the most commonly studied groups) then failing to include non-left handed people is not a bias in the study, it is a function of the limits of a study. Similarly if a study wanted only left handed US residents then the study would be about left handed US residents. This does not imply any bias in the study. Any study worthy of the name will go in to great detail when it comes to methodology of subject selection and any subsequent testing that is done. It will be great when smaller WP communities are studied too but this study clearly isn't doing that. Cheers, Rob

By golly I suppose you'd better just call Michigan State University and tell them the bad news. Surely, when they read this INTERNET CONVERSATION they will surely see their folly in conducting this study. Gerard I think you'd better contact Harvard, too. Keeping your great wisdom from them is inhumane. -S On 3/21/07, Gerard Meijssen <gerard.meijssen(a)gmail.com> wrote: > Hoi, > Given that the argument why only American people were included was the > cost of international telephony, your argument sucks. By restricting the > study to the United States it is explicitly about Wikipedia usage in the > United States. When you want to come to a conclusion on any subject with > respect to policies in the English language Wikipedia, the result will > not reflect how this project works. > > When you study left handed people, you will find only what is only true > to left handed people by comparing the results to right handed people. > Thanks, > GerardM > > Robert Brockway schreef: > >> On Tue, 20 Mar 2007, Gerard Meijssen wrote: >> >> >> >>> US Wikipedians will probably all be working on the English language >>> Wikipedia. It means that all the skills and experience associated with >>> small communities, working on an encyclopaedia that does not cover all >>> subject matter. Working on languages where the community it is done for >>> does not know what Wikipedia is, it is that experience that will be >>> missing. In that way it will hardly cover the breadth of Wikipedia. >>> >>> >> He doesn't claim to be attempting to cover the breadth of Wikipedia. >> Implicit in the post is that it concerns Wikipedia usage in the United >> States. >> >> I find it is common for people to mistake limits placed on a study with a >> bias in the study. Let me give another example that might make this >> clearer. If a study concerns left handed people (one of the most commonly >> studied groups) then failing to include non-left handed people is not a >> bias in the study, it is a function of the limits of a study. Similarly >> if a study wanted only left handed US residents then the study would be >> about left handed US residents. This does not imply any bias in the >> study. Any study worthy of the name will go in to great detail when it >> comes to methodology of subject selection and any subsequent testing that >> is done. >> >> It will be great when smaller WP communities are studied too but this >> study clearly isn't doing that. >> >> Cheers, >> >> Rob

Sure, Gerard. =) I don't think either of us know enough about the study to make long guesses/judgments about its methods. While I definitely think the points you make are valid and better than most, whether or not it applies to this study remains to be seen. While WP-l is a good place to get into lengthy debates, its all premature. Until of course we see the study cited in the Washington Post. Then we'll all get together and talk about what a bunch of lies it is. =) -Steve On 3/21/07, Gerard Meijssen <gerard.meijssen(a)gmail.com> wrote: the the the with will true people. with all for United with a commonly not a study. Similarly be it that this _______________________________________________ Wikipedia-l mailing list Wikipedia-l(a)lists.wikimedia.org http://lists.wikimedia.org/mailman/listinfo/wikipedia-l

Hoi, When this study hits the Washington Post and we look at it and consider it a bunch of lies, we are to blame when we did not speak up when we had the chance to do so. When you consider peer review, it is always done after the fact. It is much better to have input before a study is started. Those issues that are obvious can be addressed before time and money is wasted. It also leads to better science. If there is one study I would like to see done, is a wikipedia with a large ex-pat community and see how that affects the NPOV of the project. Thanks, Gerard

On 3/22/07, GerardM <gerard.meijssen(a)gmail.com> wrote: it a much to large Calling a study "a bunch of lies" before it's even been carried out is very bad faith. It's absolutely fine to question methodology - but we need to be friendly as well as critical - and I hope Benjamin has been inspired rather than turned off by the many comments and questions on these lists. I've been meaning to comment myself, but Phoebe's just said pretty much everything I was going to say. :-) Cheers, Cormac _______________________________________________ Wikipedia-l mailing list Wikipedia-l(a)lists.wikimedia.org http://lists.wikimedia.org/mailman/listinfo/wikipedia-l

Thanks Gerard. I absolutely understand your questions are intended as good faith ones - but I still don't think it's fair to suggest that this study might be (perceived as) a bunch of lies. The study will only aspire to be a study of American Wikipedians - and it could be a very interesting one within those imposed limitations. An essential part of research is to be open and honest about the decisions taken and the limitations these bring, rather than trying to cover them up. (I'm not saying here that human-subject research methodology can't ever be discussed and improved, but I *am* saying that it's never completely bias-free.) Cheers, Cormac On 3/22/07, GerardM <gerard.meijssen(a)gmail.com> wrote: _______________________________________________ Wikipedia-l mailing list Wikipedia-l(a)lists.wikimedia.org http://lists.wikimedia.org/mailman/listinfo/wikipedia-l

Gee did I just do that? Woops. -S On 3/20/07, Steve <subsume(a)gmail.com> wrote: _______________________________________________ Wikipedia-l mailing list Wikipedia-l(a)lists.wikimedia.org http://lists.wikimedia.org/mailman/listinfo/wikipedia-l

On Mar 20, 2007, at 9:11 AM, Benjamin Keith Johnson wrote: How would you like to verify my age? Oh, you want me to send credit card info? Billing zip code? CCV digits? Birth certificate? SSN? Studies like these are a nightmare to set up. -Bop _______________________________________________ Wikipedia-l mailing list Wikipedia-l(a)lists.wikimedia.org http://lists.wikimedia.org/mailman/listinfo/wikipedia-l