DataONE:Notebook/Data Citation and Sharing Policy/2010/07/22

Project name

Main project page Previous entry      Next entry

Entry title Nic Weber 10:21, 22 July 2010 (EDT):So far I've done two CI Plots (not sure that is the correct name) : File:DataCiteSimpleVar Plot.pdf and File:ReqShareSimpleVar CIPlot.pdf In the first plot, the simple variable represents those journals that give instructions how to cite data. It doesn't seem to me there is much to explore here-- The one interesting note, but I don't know if it's exactly "evidence" is that OA subscription models seem to be more likely to have citation plans than either Mix or Sub journals. With OA the confidence interval = > (100*binconf(22,307)) PointEst Lower Upper 7.166124 4.779716 10.61124 In the second plot, the Impact Factor steadily rises with a journals request / requirement for data sharing. Evolutionary Biology also looks promising as a category, but Heather pointed out the confounding factor is that EvoBio journals might simply be Higher Impact journals EvoBio confidence interval = > (100*binconf(40,307)) PointEst Lower Upper 13.02932 9.71623 17.25619 Below are some of the CI plots that I did this afternoon. The titles describe the variable tested. Nic Weber 02:26, 23 July 2010 (EDT): I played around with the LogIt Model on the UCLA site. It was fairly over my head...I used the field " Instructions How to Cite data" for a response variable, and included "Impact Factor" and the coded field "Policy Requests / Requires Data Sharing" for variables as predictors. I put my code here: [1]