The second pilot is to be interpreted in the aggregate. We used scenarios from our first pilot (and the scenarios thus differ per site, but the priming method does not).
Analyses per site:
1. Bamberg (http://rpubs.com/ivanropovik/256812)
2. Fort Lewis (http://rpubs.com/ivanropovik/256811)
3. Kassel (http://rpubs.com/ivanropovik/256897)
4. Presov (http://rpubs.com/ivanropovik/259185)
Together with the original authors we thus decided (after collecting 5 out of our 9 sites) to change the priming method.