I have to say I'm really surprised by this. You are totally right--ML resets the skipto value at the start of each questionnaire. I would hesitate to change this though because there may be many experiments out there that count on the fact that this is the way it has always worked. It must simply be the case that in testing, we were never looking at skipto values across different questionnaires.

So long story short--I don't know what to say. I'm wondering if a fix/work around could involve using a responses.xls file. You could add a blank instruction item at the beginning of each questionnaire--each using a automatic duration value of (d1) so you don't even notice it's there. Using the responses.xls file you could assign a skipto value to that first item based on any prior response (or collection of responses) to any prior questionnaire(s). Kind of a pain to bring in the Advanced Features like that if you weren't planning on it. Not too bad though if you have some familiarity with that functionality.

In any case, thank you for bringing this to our attention. And thank you for the great test sample you attached that demonstrated the issue immediately.