Research talk:Newsletter/2013/October

Academics

Thanks Adler.fa for the summary. Without venturing into the details of the paper like how the sample was chosen, I see two fundamental issues with its premises.

1. The first is that the "quality and scientific impact of researchers" is assumed to be the "ground truth" about their notability. This is not correct, because Wikipedia is an encyclopedia and not a researchers almanac: researchers of lesser impact are often more "notable" for other reasons, for instance it's not uncommon for rectors, deans etc. (in countries where they're elected between and by researchers) to have lower research "scores", but they get notability by other means.
2. The second is that Wikipedia is about verifiability, not truth. In addition to what above, even when a researcher is "more important", this doesn't matter if nobody in the world knows (yet). As a practical example, open access gives research higher impact in "popular culture", including Wikipedia: if all important researchers happened to be hid in ivory paywalls towers, Wikipedia would probably cover few of them.

Statistics would be more useful if corrected/integrated by checking what above. Additionally, there are two common deficiencies/nitpicks which pave the way for even more research:

• only the English Wikipedia, which is the most inclusionist and chaotic language edition, is considered;
• the mere existence of an article is taken as index of "importance", which is a) silly from a practical perspective, because only pageviews and the like can measure the actual impact of an article; b) ingenuous from a wiki perspective, compared to e.g. the number of incoming links/"internal pagerank"; c) invalid from a policy perspective, because the only true test of the notability criteria compliance is surviving a deletion discussion.

--Nemo 20:08, 1 January 2014 (UTC)