To: Mark A. Stang who wrote (13382 ) 8/15/1998 4:44:00 PM From: DaiS Read Replies (1) | Respond to of 23519
Mark, thanks...Look everyone, BigK will accuse me of fishing for sympathy next! The publication I posted had a J. Moberly with the affiliation Pfizer in 1995. This person might no longer work for Pfizer. Pfizer have 49,000 employees. I am sure that some of these post on SI and no doubt some might eventually drift to the Vivus thread to see what was happening. Provided that an employee gave no confidential information and just flung insults like Moberly, the maximum punishment they would get would be a ticking off - in the unlikely evident that they were caught. Stupid not to use an alias. The likelihood that our J. Moberly is the Pfizer J. Moberly is related to how unusual is the name J. Moberly I think. It would be much less remarkable if the name were J. Smith. Here is some data on the number of scientific publications 1995-98 by different names, Moberly: Pfizer - 3 Rest of World - 12 P/(P+R) 0.250 Smith: Pfizer - 50 Rest of World - 28116 P/(P+R) 0.0018 Jones: Pfizer - 32 Rest of World - 16000 P/(R+R) 0.0020 The proportions on the right can be treated as probabilities. The Smith and Jones are not significantly different, but Moberly vs Smith and Moberly vs Jones are highly significantly different. It's a bit like DNA fingerprinting, someone comes to the board and leaves their name behind as evidence instead of blood. In this case the blood is also found at Pfizer. I would emphasize that this is a very simplistic example of what could be done, for example relative name frequencies could be obtained in many different ways and there are many complications and considerations. I recall that there were no Moberlys in my database for Glaxo, Zeneca, Merck etc. What do you think Mark? DaiS