Here's a scenario. I have a series of three questions regarding a maturity scale, with an array in answer to each

Q1 - select from bad - good - better
Q2 - select from bad - good - better
Q3 - select from bad - good - better

answering 'bad' in Q1 means the answer to Q2 & Q3 should also be bad. Similarly for Q2. And so, if the participant selects bad for Q1, I want to auto fill Q2 & Q3 with bad and move to the next group. And again for Q2 to autofill Q3. If however, the participant selects 'good' in Q1, they should not be allowed to select better in Q2 or Q3. i.e. the next question in a sequence, can not have a better state than the previous.

is this possible?

Many thanks
Philip

Why auto-complete?

Use relevance to hide Q2 and Q3 if Q1 == bad.

Use relevance to hide Q3 if Q2 == bad.

Validation for Q2:
Validation for Q3:

Thank you for the tip, but does that mean Q2 & Q3 would be unanswered in the results?

Also, how does that work if 'good' is selected in answer to Q1, and only 'bad' or 'good' should be in the answer array for Q's 2 & 3?

The questions are meant to test the maturity of a process and if the respondent is achieving bad or good at Q1, they should not be able to record a better result at Q2 & Q3. I hope that makes sense?

Thanks again

Thank you for the tip, but does that mean Q2 & Q3 would be unanswered in the results?

Correct.

Also, how does that work if 'good' is selected in answer to Q1, and only 'bad' or 'good' should be in the answer array for Q's 2 & 3?

The validation equations handle that scenario.

Ok, Tony, thank you. Not having the questions answered isn't necessarily a problem with careful weighting on the Q1 & 2 responses. I think I could make that work, I'll give it a go.

You could probably user an equation question after the 3 questions to check the missing two answers, something like if(Q1==1, Q2=1 AND Q3=1)...

You cannot set questions that are irrelevant. The answer will not be recorded.

tpartner wrote: You cannot set questions that are irrelevant. The answer will not be recorded.

Not sure if Holch's approach is addtional to yours.

But for LS5 the concept of not saving irrrelvant/hidden items should be challenged.

Thank you all for all the advice & tips, but a couple of clarifications:

Q2 and Q3 are relevant, but each question describes a degree of maturity in a process, and if the answer to Q1 is "BAD", then Q2 & Q3 cannot score higher but should be scored. Similarly, if the answer to Q1 is GOOD, Q2 & Q3 should only be scored BAD or GOOD, with BETTER excluded. It follows that if Q1 is scored GOOD, but Q2 is scored BAD, Q3 has to be scored BAD, too.

It seems to me that the only way to achieve this is to hide Q2 & Q3, if the answer to Q1 is BAD, but weight it with a score of ZERO, with that being total score for the section. If GOOD is selected in Q1, weight that score as ONE, display Q2 where BAD is again weighted as ZERO and GOOD as ONE, and if BAD is selected, the total score for the section would be ONE.

The logic being that a group total score of ZERO indicates BAD as the answer to all Q's; GOOD, BAD, BAD would score ONE; GOOD, GOOD, BAD would score TWO; GOOD, GOOD, GOOD would score THREE, BETTER, BAD, BAD would score FOUR and so on.

The tricky part is not hiding Q2 & Q3, but altering the answer array, to hide BETTER as an option in Q2, for example, if GOOD is selected in Q1.

I hope that explains my questions, better?

Thanks again,
Philip

Hi,

Thank you for the tip, but does that mean Q2 & Q3 would be unanswered in the results?

You can set the value afterwards by an equation.

1. Use relevance to hide Q2 and Q3 if Q1 == bad.
2. In an equation "Qset" you set the values of Q2 and Q3, like
{Q2=if(Q1==1,1,Q2)}
{Q3=if(Q1==1,1,Q3)}
with 1= code of "bad".