Have had one chat with one teacher that asked/pointed me about different evaluation methods that perhaps could be interesting in the future. Just annotating the thing here for your consideration / future reference.

Basically the idea implies to add one extra field "confidence degree" on each question response so, when attempting the activity, the student will have to fulfill both the response (true/false, multichoice... whatever) AND how confident he/she is with that response.

Later, when calculating points for each question, it is evaluated differently based in the confidence degree introduced. Obviously, the point given must encourage students to use "sincere" confidence degrees and so on. It seems that there are some studies about that (links below).

The system has two advantages IMO, better (clever) evaluation and interesting feedback for teachers. Note that both features (the introduction of confidence degrees and their use in the evaluation) can be connected or no (i.e. we can force the introduction and not use it in evaluation).

Original link explaining the concept (in Spanish):

http://www.quadernsdigitals.net/datos_web/hemeroteca/r_6/nr_79/a_906/906.html

Translated by Google (looks readable for one Spaniard like me):

http://translate.googleusercontent.com/translate_c?hl=en&ie=UTF-8&sl=es&tl=en&u=http://www.quadernsdigitals.net/datos_web/hemeroteca/r_6/nr_79/a_906/906.html

As said, just one idea...ciao