[Rcicr-users] partial data

[Dotsch, R. (Ron)]

Dear Thomas,

Thanks, I’m glad you like them. The libraries you mention should be easily installed on Windows using install.packages. Let me know if you had to use something different for that.

About your question: In principle scenario A should work fine with the package. In contrast to earlier scripts I wrote in python or matlab, this package does not assume that every stimulus that has been generated has been judged. It only uses those stimuli that are referenced to in the stimuli argument of generateCI2IFC (this also works with batchGenerateCI2IFC). So both scenario A and B should work without making any adjustments to the package (but please let me know if you run into problems here, if you have data and it doesn’t work as advertised, I can debug).

Whether you want to go for scenario A or B depends on what you’re going for: scenario B ensures that every participant saw the same stimuli, reducing error when comparing the CIs of different stimuli sets. But if you want to have as many stimuli judged as possible by a group of participants, then scenario A would work too.

Anyhow, for the rcicr package, A or B are both easy to handle.

Best,

Ron
________________________________
Dr. Ron Dotsch

Utrecht University
Social and Organizational Psychology (Room E2.22)

Website: http://ron.dotsch.org


On Nov 14, 2014, at 14:49, Thomas Schubert <thomas.wolfgang.schubert at gmail.com<mailto:thomas.wolfgang.schubert at gmail.com>> wrote:

Hi Ron (and the list),

congrats on the new R version of the scripts, it worked nicely.

I noticed that a newly installed R on a windows machine also needed the libraries jpeg, matlab and aspace.

I have a question about missing or partial data. How does the script deal with them at the moment? In particular, what would be the better strategy:

a) generate a lot of target images (say, 770), randomly allocate only around 150 to each participant, but sample more participants per condition, or

b) generate 300 target images, but get every participant of a smaller sample to answer every image pair?

would the software even tolerate the partial data coming out of a) ?

t


--
thomas schubert
department of psychology, university of oslo
schubert at igroup.org<mailto:schubert at igroup.org>
http://www.igroup.org/schubert/

_______________________________________________
Rcicr-users mailing list
Rcicr-users at lists.r-forge.r-project.org
http://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/rcicr-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.r-forge.r-project.org/pipermail/rcicr-users/attachments/20141114/cc3538ca/attachment.html>