Rosters ending, or not showing all occurrences using CSEntry

[jbthompson21]

I am Having a strange issue come up during a training we are conducting prior to the start of field work that is concerning. Randomly sections that are rosters whether they be entering plots until they are completed or multiply occurring rosters where questions are repeating but changing what they're asking about based on %getocclabel%, these sections are just ending and moving on to the next form prematurely.

I can't figure it out because it is happening randomly and sometime isn't a problem but then starts in the middle of a survey. The only way I can get them to show up is to enter the case tree, select the last occurrence, hold it down and then select add occurrence and it adds what should have come up automatically. This is a time consuming process especially with a long survey and something that would be difficult to train enumerators how to do. It is also strange because using the same tablet, running CSEntry, if I open a new case most of the time I won't experience the same issue.

Has anyone else run into this problem? I can't figure it out as it doesn't seem to be a coding issue as most of the time on most tablets it works just like it should with no issues. But running a large survey with a big team i'm worried it will be a big problem. Any help or suggestions would be extremely helpful! Thank you.

[josh]

Is this only hapenning in modify mode (reopening a previously completed questionniare) or does it also happen in add mode (adding a new questionnaire)? In add you can continue to add rows to the roster until you get to the maximum occurrence. In modify mode, the number of occurrences is set based on the number of occurrences set when the case was first entered. To add additional occurrences beyond that in modify mode you have to either use the case tree as you are doing or use the insert function from logic.

Are you using operator controlled mode or system controlled mode? Rosters behave a bit differently in the two modes. In system control the interviewer can't manually exit a roster before the end, it must come from program logic.

Do you have logic to end the rosters (using endgroup for example or skipping to somewhere outside the roster)? Could be a bug in your logic that causes the roster to end.

[jbthompson21]

Thanks Josh!

It is in add mode. I'll test out a few of your other suggestions as far as operator and system controlled and see what I can come up with through trial and error. I do have codes to end rosters, or they just end when they come to the end as when i'm pulling dynamically using %getocclabel%. But it seems fairly random as to when it happens.

I'll test some things and report back if I can get it figured out.

Thanks!

[jbthompson21]

Have been banging my head against the wall over this issue of multiply occurring rosters randomly only allowing the first occurrence then skipping the rest and moving on. Then from that point on doing the same for all following rosters. The only way I can seem to fix this is go into the case tree and manual add occurrence which isn't feasible given the length of the survey.

Just wondering if anyone else has experienced this at all. Is there a new version of CSEntry that could be the reason for this? I'm stumped because of the seeming randomness of it happening and at this point it's potentially putting the start date of our field work in jeopardy. Just wondering if anyone has any idea about this issue or has seen it or heard about it before?

Thank.

[Maria]

One idea would be to group the issues in different forms by sections or pages of your survey that facilitate movement within the tree.
good luck