Hi,
If you really want to confuse your respondents and make them angry, the solution is really easy (at first glance).
Something like
{if(randomnumber==1,"|identique|different","|different|identique")}
as subquestion text.
But what about the scale? This isn't changed at all.
In one case "different" gives a value of 0, sometimes "identique"
This means you have to create your random numbers before and store them.
As you have to recalculate the values you have to know the value of the random number.
Maybe like this
With a small javascript snippet create a string containing 24 random numbers "1" or "2". (QRandom)
Then you can change the above formula to
{if(substr(QRandom,0,1)==1,"|identique|different","|different|identique")} for the first question
{if(substr(QRandom,1,1)==1,"|identique|different","|different|identique")} for the second question
{if(substr(QRandom,2,1)==1,"|identique|different","|different|identique")} for the third question
...
And regarding the scale. I have to admit, I do not know your exact scale, so here I use 0 - 100.
In an expression at the end of the survey you can recalculate like
{Q1_1=if(substr(QRandom,0,1,)==2,100-Q1_1,Q1_1)}
It's possible, but I do not recommend at all.
Much better you randomize, but for all questions in the same direction.
One respondent always sees "different - identique" the other "identique - different"
JOffm
Volunteers are not paid.
Not because they are worthless, but because they are priceless