this is most likely because the revisiting on the questions is set to automatically decide, rather than resume or reset, it can behave different on each computer. In the testing it may have appeared to reset whilst on your computer it may decide to resume and keep the last decision. It can REALLY throw you off with testing:) The same for the gameboard slide, I havn't really looked into the mechanics, but I suspect it is because you partially regulate the states via variables and partially via clicks, you won't get this inconsistency if the variable is always the decider, using state combined with state via variable and letting revisiting decide on things is likely to throw you off.
I'm going to see if I can figure out where it goes wrong,(I highly recommend doing this as this is where you learn most) the way I do that is
1. Safe the file as a copy test file
2. delete everything that is not necessary,
3. create a textfield with a reference to the variable that causes the issue and that way check if the variable is doing what it should be doing.
4. describe the logic of the failure to myself. e.g. state changes/has changed or is reverted when I get back. the value should be 'x' and so on.
5.It becomes a simple excersice in of deduction till you discover the flaw.
If you describe it logically there can only be 1-3 causes, which you test one by one:)