Welcome, Guest
Username: Password: Remember me

TOPIC: Adaptive procedure in Limesurvey

Adaptive procedure in Limesurvey 2 years 10 months ago #70651

  • chinglnc
  • chinglnc's Avatar
  • OFFLINE
  • Junior Lime
  • Posts: 29
  • Karma: 0
Hi

I wonder would it possible to create limesurvey questionnaire based on adaptive procedure.

For instance, if the user has answer 3 questions wrongly, he/she will be presented with certain question etc.

Here is an example of a two down one up procedure.

Q1- Correct
Q2- Wrong
Q3- Wrong
Back to Q2, Q3 which will keep repeating until he/she got 2 consecutive right answers or the program will terminate after xx attempts of wrong trials.

Can we do something like this in limesurvey?

Thanks
The administrator has disabled public write access.

Re: Adaptive procedure in Limesurvey 2 years 10 months ago #70669

  • Mazi
  • Mazi's Avatar
  • OFFLINE
  • LimeSurvey Team
  • Posts: 5333
  • Thank you received: 296
  • Karma: 249
At the current version of Limesurvey it is not possible to do looping or jumping backwards.
The new Expression Manager at the upcoming Limesurvey 2.0 version might be able to handle this.

Best regards/Beste Grüße,
Dr. Marcel Minke
(Limesurvey Head of Support)
Need Help? We offer professional Limesurvey support
Contact: marcel.minke(at)limesurvey.org'"
The administrator has disabled public write access.

Re: Adaptive procedure in Limesurvey 2 years 10 months ago #70700

  • TMSWhite
  • TMSWhite's Avatar
  • OFFLINE
  • LimeSurvey Team
  • Posts: 759
  • Thank you received: 82
  • Karma: 36
Interesting.

Creating a function in EM (Expression Manager) to jump back to a prior question would not be hard. The challenge is tracking how many times the question was answered wrong.

We have Equation question types, which are hidden equations that would let you check whether Q2 and/or Q3 were wrong. You could then conditionally jump back to Q2 if you did not have two consecutive questions correct. Keeping track of how many times you had to jump back would require that I get the answer assignment feature in EM working (e.g. so that you could have a variable like 'Q2andQ3wrong' and conditionally say {Q2andQ3wrong += 1} to change its value. Then, the rest of the survey would have Q2andQ3wrong as a condition so that you could skip the rest of the survey.

However, this is such a common style of surveying in Education and Psychology, that we should probably think about a generic solution where you could specify something like:
(1) terminate_if_more_than_x_questions_in_row_wrong
(2) jump_back_x_questions_if_more_than_x_questions_in_row_wrong
(3) max_times_to_jump_back_if_...
(4) add a correct_answer column for questions so that you don't need to write an Equation to test each one.

Even that may not be ideal. As you probably know, there is a booming field of computer adaptive testing which uses statistics to decide which question to visit next. It will now be easy to have LimeSurvey visit any desired question, but in order to implement such complex branching, we should probably see whether there is an open source computer-adaptive-testing engine that could support the navigation logic. Once we know which question can be visited next, LimeSurvey's EM can properly tailor the question based upon prior responses.
The administrator has disabled public write access.

Aw: Re: Adaptive procedure in Limesurvey 2 years 7 months ago #76893

  • flink
  • flink's Avatar
  • OFFLINE
  • Fresh Lemon
  • Posts: 19
  • Karma: 0
I would be VERY interested in adaptive testing (CAT) using Limesurvey. There are already some open source CAT-Algorithms. Do you think something like this could be implemented in the near future?
The administrator has disabled public write access.
Moderators: ITEd
Time to create page: 0.128 seconds
Donation Image