userfriendlyscience.com

Scientific research made accessible

Research tools | Explanations | R packages

Welcome to userfriendlyscience.com. In the future, this domain will strive to provide overviews
of free or cheap instruments for research, as well as explanations of how to use those tools.
For now, it only informs people about the userfriendlyscience R package, which you can install with:
Research tools | Explanations | R packages

install.packages('userfriendlyscience');

At this point, R will probably ask you to choose a mirror, and will then proceed to
download packages that are required by the userfriendlyscience package, and the packages
required by those packages, etc. You can also install the most recent (in progress) version, using:
install.packages('userfriendlyscience',
contriburl='http://userfriendlyscience.com/src/contrib',
type='win.binary', dependencies=TRUE);

R doesn't always manage to download the dependencies from CRAN when you use this latest version, so you might want to install the regular CRAN version of userfriendlyscience first, and then run the command above so that most required packages are loaded already. You can replace the type argument with "type='source'" to download the source package and compile it in R.Once it's done, you can load the package with:

require('userfriendlyscience');

You can get an overview of the userfriendlyscience package, with a link to the
index with a list of the functions in the package, with:
?userfriendlyscience;

Or you can go ahead and testdrive some function. Then, first create a simple dataset:
dat <- data.frame(x1 = factor(rep(c(0,1), 20)),
x2 = factor(c(rep(0, 20), rep(1, 20))),
y=rep(c(4,5), 20) + rnorm(40));

Or simply load your own dataset from SPSS. This command displays a simple file dialog
that allows you to select your datafile, which is then stored as 'dat' in R:
dat <- getData();

And then compare two means (replace 'x1' and 'y' with your own variable names):
meanDiff(x=dat$x1, y=dat$y);

Or for both predictors at the same time:
meanDiff.multi(dat, x=c("x1", "x2"), y=c("y"));

Or you can create a dot-line-violin plot:
dlvPlot(dat, x='x1', y='y', posDodge=.3);

Or a more complicated version with a moderator:
dlvPlot(dat, x='x1', y='y', z='x2', posDodge=.3);

Note that for some functionalities, such as itemInspection and scaleInspection,
you will need an installation of LaTeX. See the help for more information:
?rnwString

You can download a portable version for Windows at
http://miktex.org/portable
and for Mac OS X at http://tug.org/mactex/.
If you have any questions, you can contact me through http://behaviorchange.eu.