Question 1

Dans un didacticiel, si un élève commet une erreur, il est préférable de:

1. Informer l'élève de son erreur et lui donner la réponse correcte
2. Informer l'élève de son erreur et lui fournir un indice qui l'oriente vers la bonne réponse
3. Afficher une icône qui clignote, l'élève peut cliquer sur cette icône pour demander l'aide du tuteur
4. Laisser à l'élève un certain temps de réflexion pour qu'il trouve son erreur par lui-même

I would like to start with some studies that have been done in the area. The most famous is the behavioral model of instruction which is an application of Skinner's principles of operant conditioning. For Skinner, behavour is contigent on the relationship between a response and it's consequence, which is called reinforcement. He strongly believes that "we learn when what we do has reinforcing consequences".

According to Skinner, reinforcement must be immediate. One of the key ideas in this model, that I'm going to refer to, is feedback.

A behavioral model is appropriate for CAI. Because computer is a good tool for implementing behavioral principles. Care must be taken, however, to adopt behavioral principles so that instruction is in harmony with the computer as an instructional medium. Because, defining an error is difficult, particularly in CAI.

McKeanie though concluded that the value of feedback probably lies on the information or the motivation it provides rather than the stamping in of stimulus-response connections.

That is, that negative reinforcement (information that a response is incorrect) may decrease the probability repeat the error, but it doesn't necessarily help him learn the correct one.

Moreover, reinforcement need not necessarily be immediate. Several studies showed that immediate feedback may not be as instructionally effective as delayed feedback.
If it's not the role of feedback to strengthen responses then there is no reason to ask whether feedback must be immediate or under what conditions is essential.

Anderson, says that on errors feedback must be immediate but the reason is not negative reinforcement. After a delay, a student forgets the responses he gave himself on the initial test.

Also, students who get delayed feedback take more time to study it than those that get it immediately. Another point is, that it is difficult for a student to find and correct his errors. So, he then continues with a series of steps based on that error.

This can make students overwhelmed by information to recover errors, may not actually learn the correct response, delayed feedback may learn students incorrect procedures and of course causes frustration. So feedback nature depends on subject matters and nature of the tasks been given.

Another theory, constructivism proves that student should be left alone to discover it's own errors. But this is proven to work better in complex problems.

Having in mind all of the above, that conclude years of studies and experiments on the area, we can now analyse the four possible answers on question 1.

• The first answer follows the simple principle of Skinner's : immediate feedback.
  
• The second answer goes one step further and involves also the research.
  
• The third answer promotes the freedom of the student to choose whether he wants help or not.
  
• The fourth answer doesn't follow most of the above principles. First, it is not immediate, so it doesn't follow the behavioral model of Skinner's. Second, according to Anderson it is difficult for a student to find out his own errors, this takes a lot of maturity. But according to constructivism theory it is a good model for complex problems.

Personally, I tend to believe after all the reading I've done, that there is no answer than can be considered the right one. There is no golden rule to follow.

Comparing the first two answers, I prefer the second one that goes one step further and motivates the student to find the correct answer and work on it.

But if the student has a certain maturity (educationally) third answer seems most appropriate because leaves the freedom to the student to decide whether he needs some help and review his answer. Of course, this has the risk of adopting wrong answers.

If the task is a problem solving then definately I would choose the fourth answer because the task is much more complicated even though, I strongly believe that is difficult for a student to realise that he has commited an error so as to recover it.

So, I distinguish 2 categories of answers :

1. immediate feedback, answers 1,2,3 and
2. delayed feedback, answer 4

It seems to me that the best solution would be a combination of both categories in a way that the student left to discover his errors and think about, but a feedback should be stand by at any time considered to be appropriate in case that the student has taken a wrong path or persists on the same mistakes.

Back to "Travaux" page