Hey LimeSurvey Time,

I have read through previous threads on Multiple Choice functions of LimeSurvey, but I am really wondering how well LimeSurvey can serve as Question Bank for Multiple Choice Quizzes.

To integrate this in higher academia, I believe the application should be able to show the correct answer immediately after answering a single question.

So, for example, if there are 20 questions and the first one is:

What does ADHD stand for?
1) all dummies hate dummies
2) random wrong answer
3) random wrong answer
4) attention hyperactivity deficit disorder

The user then submits a wrong or correct answer and the application would return:

Attention hyperactivity deficit disorder is correct. ADHD is the most frequent psychological disorder in children and bla bla bla.

Thus the information would immediately be corrected or reinforced.

If LimeSurvey is not ideal for this, any recommendations?

Thank you very much, LimeSurvey Team - I really appreciate open source because it makes the world go round :side:.

Kind Regards!

Hello,

with LimeSurvey 1.92 (which was released about a week ago) there are many new possibilities using the ExpressionManager to set relevance of a question, validate an answer or even do some micro tailoring.

If you want to show the correct answer immediately you can for example
1. Add your ADHD question
2. Add a new question (question type: Text display) and enter your information/solution text.
3. Use the question relevance to show the text only if the participant has answered the ADHD question (e.g. "Edit current question" -> Insert "!is_empty(ADHD)" into "Relevance" -> "Save")

If you want to do some micro tailoring you can add something like telling the participant if he/she chooses the right or wrong answer.

If you want to try it out I'll add a question group to this post so you can create a new survey, import the question group and have a look how it works.

valkener-

Steve's approach does provide immediate feedback, but there is the risk that the person would use that information to change their answer before clicking Next. So, you would get the reinforcement, but might not capture their initial answer.

Also, if the person was crafty (and patient), they could look the HTML source of the code to figure out what the correct answer is, since the equation to do the tailoring is embedded in the JavaScript (although it is hard to find and read).

If that is a concern, there are two possible options:
(1) Put the feedback on the next page - so you capture their answer, and the next page shows whether they are correct or not, plus the reinforcement. That way they can't game the system. This can be done now in version 1.92.

(2) Wait for (or help us with) one of our planned enhancements. I want to augment the data model to capture the full history of answers and timings, so if you pick one answer, but then change to another (either while still on the same page, or by backtracking to a prior page and changing the answer there), you would get that full history of changes. With such an approach, you can provide instant feedback, but still be sure that you can capture the first submitted answer (and also know whether the person tried to change the answer).

/Tom

While, it's great that there is a way to get the immediate results, is there an easier way to set it up?
Writing the above equation is error prone and tedious and something a regular survey creator can't easily do.
Is there an easier way? maybe a template you could point to where it's already setup. I could just import the question and edit it.