In the XLS form we have 3 different sections (begin/end repeat):
listing of the current household members
members who have left the household during the recall period
members who died during the recall period.
The instruction to start one of these sections is:
“Press “Add Group” - to add another household member until all members are listed. When listing is complete, press “Do Not Add” to continue with the questionnaire.”
Automatically, a group is added for the listing of the current household members which is fine as we will always have at least one person.
But I would like to avoid that the two sections about members who have left / who died be set as compulsory questions/group.
Currently, a group is automatically presented. Even if no data are entered into the questionnaire/section, we have an incorrect total of members who left/died (1 instead of 0) because of the presence of the group.
An alternative is to add two new questions:
“What is the total number of household members that LEFT this household during the recall period?”
and “What is the total number of household members that DIED during the recall period?
If the answer is “0”, then the questions related to the persons that left and died will not appear.
If the answer is “2” the group of questions will appear twice.
But I would like to avoid adding new questions.
Any guidance on potential ways to avoid adding new questions ?
Welcome to the community, @fannette! Looking at your scenario I would advise you to use the repeat count header to repeat a group of questions till you wish. Maybe you could have three questions for this at the beginning:
listing of the household members currently available at the household
members who have left the household during the survey period
members who died during the survey period.
Once you have this number you could create 3 groups for each question asked above and then control the iteration of the group as needed.
Maybe our community have better suggestions for this.
Thank you for your answer it is highly appreciated
Yes, my plan B is to use “repeat count” → ${total_depart} and “repeat count” → ${total_died}. It is working well.
But we would like to avoid to introduce 2 additional questions in our survey (What is the total number of members who left/died?).
The problem is that we have none required variable before to start the group - just a note saying “MEMBERS WHO LEFT HOUSEHOLD: Please complete the following information for all household members who moved away from the household during the recall period”. With the instruction (hint): " Press “Add Group” - to add another household member until all members are listed. When listing is complete, press “Do Not Add” to continue with the questionnaire."
We would like that the enumerators feel free to skip a question/section if needed, and go back later, so we don’t want to add a required variable.
@fannette, the form design would also differ or vary based on how convenient the data manager would be managing the captured data. Some would like to keep all the variables within a roster while other would like to separate them as needed. So ultimately if you are the data manager it would depend on how convenient you structure your variable for the form.
Hello,
We suggest to frame with a group. Otherwise you may get ODK/Collect problems with 0 repeats or going back and changing the repeat control variable.
Hi @fannette
I know you are familiar with the skip logic criteria, so apologies if this link belabors the point. To propose a solution, you need to think about the design, as hinted by @Kal_Lam. I would generally structure my design questions as follows:
Do you want the “details of the member who left” to be in the same datasheet as the “details of the household members”? If so, then keep the details of the members who left in the same group as household members by simply adding a question (did this member go?) in the repeat group.
Do you want the datasheets to be separate? If so, then keep your design as it is. However, you need to add a skip logic on the group for “members who left”. It is this skip logic that will make the entire group appear or does not appear.
In case what we have provided does not seem helpful, you could post your XLS form and screenshots of what you see so that we can support you further.
Thank you very much for your answer. It is highly appreciated
Yes, we would like to have 3 different tabs in the output: one for the current household members, one for all household members who left and one for all household members who died. (Option 2).
Each section is independant so there is no question where to add a skip logic…
We only have a note before the repeat group…
See the XLS file in attachment. SMART_Survey_MORTA_EN_FR_03.xlsx (22.4 KB)
Any suggestion on how to add a skip logic in this context?
No unfortunately I can not add a skip condition here (except ${consent}=1).
The listing of the current household members (hh roster - previous section) is not related to the listing of the household members who left…
Hello @fannette ,
You can just to the same as for HH members. Ask for the number left and use it as repeat_count and for the relevants (see above). You can treat 0…n who left. (similar to 1… n HH members).
