I hope this message finds you well. I am reaching out today with a question regarding LimeSurvey plugins.

I am interested in finding a plugin that is capable of measuring the time spent on each individual survey question. To clarify, I would like to know the duration spent by a participant on Question 1, Question 2, and so forth during their survey completion. 

If there is a plugin that can address this need, I would greatly appreciate any recommendations. 

Thank you in advance for your assistance.

Best regards,

Hi,
this is a built-in feature.
Read the manual about "activating a survey"

Joffm 

Thank you for your prompt response.I did find a built-in feature that records the response time, as you pointed out. However, this feature seems to track the time on a per-page basis. The functionality I am seeking is to track time on a per-question basis, not per page.It seems I may have to create this functionality myself.Thank you again for your help and guidance.Best regards,

Just like Joffm, this is built in.

However, to know how much time they spent on a specific question, you need to use the question by question mode. If you use the group by group mode, you will only know how much time they spent on each "group" / page (if you use group by group but only put one question per group, then of course the group time is also the question time). For all-in-one this will of course only give you the time spent on the questionnaire as a whole.

How do you know how much time someone spends on a question if there is more than one question on a page?

You could measure the time from opening the page until the first question is answered, then from there until the second question is answered, etc.

However, this does not guarantee that this is really the time spend on each question, if you have more than one question on a page. Because it is quite common that people read all questions on a page, before they start to answer.

So I think if measuring the time spent PER question is important, they only way to get usable data is present only one question per page. Anything else is guess work.

Thank you for your suggestion.Indeed, preparing one question per page would solve the issue in terms of tracking time per question. However, I worry that if there are a lot of questions, this approach might increase the burden on the respondents.I appreciate your thoughtful input.

However, I worry that if there are a lot of questions, this approach might increase the burden on the respondents.

But the number of questions doesn't change, it is still the same number of questions, just now one per page.

Yes, there might be one or the other additional click, but generally what we see in practice is that one question per page works better than when you have a lot of questions per page. The standard should always be question per question with a few exceptions, where you have either very, very short questions (e.g. demographics, age, gender, etc.) that can be put together on one page, or because the logic of the survey requires it (rare).

Also, keep in mind that more and more people respond via a smartphone / mobile, where question per question is a lot more comfortable for respondents and basically a must.

Thank you for your insightful suggestion.Indeed, for a survey with a smaller number of questions, your method could be quite effective. However, the survey I am currently planning consists of around 60 questions, which would require 60 page transitions - I fear this may place an undue burden on the respondents.On the other hand, I am also considering reducing the number of questions. If that's the case, your suggestion might work well.Thank you again for your thoughtful proposal.Best regards,

However, the survey I am currently planning consists of around 60 questions, which would require 60 page transitions - I fear this may place an undue burden on the respondents.

As @holch,
the recommendation is always "one question per page", except short questions that belong together thematically like "gender", age"
And it is much easier for the respondent to click the "Next" button than to scroll and scroll.

And this way you have more opotions to give a feedback to your respondents, encourage them, offer a break...
Do not just fire questions, but try to set up a conversation!
Lead the respondent through the questionnaire by text display question that anounce and explain the next part!
You see, 60 questions is a lot. Try to reduce, so it is possible to complete the questionnairewithin 15 minutes!

Here something to read about it




 

However, the survey I am currently planning consists of around 60 questions, which would require 60 page transitions - I fear this may place an undue burden on the respondents.

The undue burden are the 60 questions, not the 60 page transitions, in my opinion.

I don't think that having less page transitions will make respondents find a survey with 60 questions less cumbersome.

But that is just my opinion.

With more than one question per page, as I said, you could measure (via Javascript) the times people answer a specific question and then calculate the differences between the questions.

However, there are a lot of downsides to this that might make the measurements useless, in my opinion.

1. There is no way that you can guarantee that you are really measuring the time for a specific question. respondents could scroll, check other questions, go back to a previous question after the have seen the next question, etc.

2. For some question types it will be very difficult to determine when to stop measurement.
- Multiple response question: How do you know the respondent has finished? How do you know if the respondent isn't already reading the next question?
- Array type questions: When do you stop the counter for an array? When they answered all subquestions?
- What if they change their mind and go back to a previous subquestion or even a previous question to change their answer?

So, if the measurement of the time spent per question is really important to your study design (and from all the effort you put in it seems that way), in my opinion, the only clean way to measure these times is when you present each question separately.

Because if you need these times for your analysis, but can't control all these issues I mentioned above, what is the data worth that you collected? Nothing! Because you can't say that the time you received is actually the time spent on a specific question.