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

{| width="800"
 * style="background-color: #EEE"|[[Image:owwnotebook_icon.png|128px]] Project name
 * style="background-color: #F2F2F2" align="center"|  |Main project page
 * style="background-color: #F2F2F2" align="center"|  |Main project page


 * colspan="2"|
 * colspan="2"|

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) : [[Image:DataCiteSimpleVar Plot.pdf]] and [[Image: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: [git://gist.github.com/487090.git]


 * }