Generally speaking, if the tasks imply the subject's response, load can be measured by assessing the number of correct responses (with respect to the number of responses expected) and the reaction times (time interval between a task request and the subject's response). It is known that reaction time is generally a more sensitive measure with respect to correct responses, as it begins to show modulatory effects due to increasing cognitive load (usually, a delay in response), well before a decrease in correct responses rate happens. I.e. the subject gradually deploys larger and larger effort, still managing to sustain optimal performance in terms of correct responses, up to a point in which a breakdown of performance is observed through a decrease of correct responses rate. I suggest literature by Susanne Jaeggi about cognitive and working memory load.

Dear Erika

Thanks you very much. Your answer helps a lot.

Paul Chandler publishes widely in this field - refer to links below.

http://ro.uow.edu.au/do/search/?q=author%3A%28%20paul%20chandler%20%29&start=0&context=119687&sort=score

http://ro.uow.edu.au/edupapers/763/

http://socialsciences.uow.edu.au/education/contacts/UOW173850.html

 Thank you to Egon and the rest . I am also in the midst of evaluating the effectiveness of Augmented Reality in TVET  learning .still finding  suitable measuring methodology and paramaters