I am trying to program a psychological test, and have about 32 answers, but then in order to give people a result, a number of calculations need to be made. This of course can be done with the expression manager, but each of those is producing a 'question' count. And people will completely freak out if they start on a questionnaire with 120 questions, even though it is just 32.

Is there any way to exclude calculated fields from the question count ?

Easy way : disable this in Survey option, and show your own count.

Yes, not using a certain feature is of course one way.

I was more wondering whether in the Display options or so there might be an option. Like there is 'Hide Question' and then there could be an option for these to also exclude from count if Hidden.

No, that option is not available.

Personally, I have no use for the question count or question numbering "features" as they are totally misleading in surveys with conditional questions (which most have).

1. Don't show it.
2. In case you are willing to give a hint, insert it in the welcome message.
"This survey has 32 questions"

As Tony has said, the question count will be missleading in many, many cases, it is almost impossible for LS to create a correct question count for everyone. I would go with Joffm's suggestion: Take the count out and either don't show it at all (because i also can create a survey with 10 very simple single answer questions or I can create 10 very long array questions and the time to fill the survey will be significantly different, even with the same amount of questions) or put your own count at the beginning manually.
Sorry, no such feature and I doubt that this will ever come, because there is no real solution to the problem, so why tackle it? I'd rather have the developers other issues, with a higher chance of success.

In 2.73 and before : you can set such question to hidden .

Then it was not included in count.

Else : my opinion : we must didn't include :
- Equation question type
- Text display question type

We still can not remove unrelkevant question, but better than add whole.

Maybe a bug report can be done ?

holch wrote: Sorry, no such feature and I doubt that this will ever come, because there is no real solution to the problem, so why tackle it?

A good progress bar would rely on such counts too.

Other survey tools sometimes allow different indication modes. You often know what percentage a respondent will see (2 questions out of 10 in that group etc.).

The calculation routine for progress-bars is then based on
a) Questions shown (to calculate that it helps, to have an additional field where you can make a guess what percentage of questions a user will see)
or
b) Question position in the survey

I mentioned that in a feature request here: bugs.limesurvey.org/view.php?id=9677

Yes, I know, Jelo. But to be honest I think we have more important issues in Limesurvey than a perfect progress bar and I guess you agree.

Hi,
I think the main question was about "There are X questions in the survey".
This can be included in the welcome text with an individual wording.

The progress bar is something different.
Even if you have some options that affect the calculation, it never will be correct.
Here an example of a commercial tool



In each question you may set either the total percentage or a step that is added to the previous value.
But it usually will fail if there are loops or different routings and extensive filters.

Example:
Before the loop you finished 50% of the survey.
After the loop 70%.
The maximum number of loops is 10, So each loop will add 2% to the progress bar.
But the respondent only is asked two loops. So the progress bar will jump from 54% to 70%.

Or if you completed 20 questions of 50. So the progress bar is at 40%.
Now there is a question that makes you jump over the next 25 questions. The progress bar will jump to 90%.

In my opinion there is no satisfying solution if the decision whether a question is displayed or not is made during the survey.
But questions with an initially setting "Don't show this question" should be considered.


Therefore you will read very frequently "Do not use progress bars if you want to design a good survey"

Joffm

Being able to exclude hidden equation question types would already do the trick. Thanks !

I just created a survey in version 4.1.4.
1 question (radio)
2 questions (equation)
6 questions (text display)



and the question count seems to be exactly what you asked for in your first post.