• The conclusion that I eventually left out in an attempt to avoid quarreling with colleagues (hello Marco Marsella) (though sometimes quarelling is a holy duty): It clearly means that an average instructor cannot, in the current development stage of the specifications, without an intensive support, implement confidence degrees in a unit of learning. Despite all these nice tools and the many engineers around, all heavily funded by European citizens, the following basic requirement still cannot be met: any assessment tool of the next generation should, for each question, provide an easy possibility to get three answers from the student a) the answer itself b) confidence degree expressed in percentages of chances c) comment or justification (a few sentences).
• At last, I am back to my study, after hard (and probably useless) work for icoper WP5 and 6.
• No Dutch today, very very bad. But too tired.