search for: _case_

...cess to a Windows mahine, or virtual machines running Windows as a guest OS. If the need is only to cross-build packages, it would probably be easier to do so within the Unix package install system, effectively as another sub-architecture. So this posting is an opportunity for users to make a _case_ (including offering resources) for continued support for cross-building either of R itself or of packages. Either on this list or to R-Windows at r-project.org (which goes only to the Windows' developers). (Cross-building is currently broken in the development version of R.) -- Brian D....

...not interested in individual differences among subjects, which I assume are due to differences in use of the scales. I have selected the 20 cases haphazardly. I assume that each measure (familiarity and emotion) contains two kinds of kinds of effects, each with its own error, one specific to the _case_ (hence "common" to all subjects) and one that is "idiosyncratic" to the particular subject-case combination. ------ I. One wrong way to analyze these data is to compute the correlation within each subject and then do (e.g.) a t-test on the correlations. This is wrong because...