Hello everyone. I created a roster to capture the different household members with their first names and only the initials of their last names in Question 2. E.g. John K, David C etc.
• In Question 2a, it is required for only the first name of each member to be captured and it should appear serially when listing the household members in Question 2.
• Question 5 will only appear if there is a household member between the ages of 5-17 years.
• In Question 5a, the study wants to know the first name of each other members of the household and Question 5b wants to know the relationship of any household members aged 5-17 years to each of the other members in the household.
Do you have a question related to this work? Is there a specific problem that you're running into? If so, how can we reproduce it? What have you done to try to address the need you have?
Yes, indeed. The work has been included along with the question, and I'm attaching it here for your reference as well.
Currently, the main challenge lies in not fully understanding how to approach the work.
Up to this point, I've attempted various solutions, but none has proven effective.
You need to make use of indexed-repeat. This example is very close to what you are trying to do:
The attached form implements a version of this (see the calculation column). I made two other optional changes to make it easier to script. In my version:
Note cells B19 and B28. Enumerators enter the name once in SecB1_Q2 but they do not enter the name again in SecB1_Q2a and SecB1_Q5a. This is my personal preference, but it will save time and result in cleaner data.
Note cells K25 and J26. In the SecB1_Q5 repeat group, they will be asked Person 1's relationship to Person 1. For instance, if the household members are Matt, Alex, and Peter, the script will go through every HH member. For Matt, it will ask Matt's relationship to Matt, to Alex, and to Peter. I'm sure there is a way to avoid this -- if that's important, maybe someone else can advise on the best way.
Hint: Do systematic pretests, incl. moving back and adding, removing and changing info from roster members. Also, check calculations concerning those repeat updates. You might need to pack the repeat content in an additional inner group (with a relevant) to avoid update inconsistencies. Furthermore, be aware that Collect and Enketo may act differently on those updates.
Thank you for your assistance, it has been valuable. However, I've encountered an issue where the names referenced in SecB1_Q2b and SecB1_Q5a are mentioned in notes and they do not appear on the datasets.
Can you see the variables (columns in your data) indexed_name1 and indexed_name2 in your dataset? Or whatever you named the equivalent calculate fields?
If not:
Can you see the other variables (e.g., SecB1_Q3 and SecB1_Q5b) that are inside the repeat groups? Or are they also missing?
Yes, I can see the variables within both indexed_name1 and indexed_name2, and I'm able to transfer the names into the 2a. and 5a. field via copy and paste.
Is there a way to enforce a restriction to allow only one response per relationship?
For instance, if a name is chosen as a biological father, another name should not be selected as another biological parent for the same child.
If it were me, I would have a calculate that sums the number of times any HH member has been selected as the biological parent, and have that trigger a note at the end of the repeat group that says "More than 1 person is listed as the biological father for _____. Please go back and correct this." I would make the note required so the enumerator cannot pass it without resolving the issue.
But I think there will be more direct solutions in the thread wroos suggested.
Here is a very quick version to illustrate the logic. sum_repeat.xlsx (10.4 KB)
Hint: You might also need more checks, e.g.
- corresponding sex of parent
- maximum limit for age-difference between biological mother (parent) and child
- maximum age for mother on birth of child.
Maybe also review your analysis requirements, please. International MICS (Unicef) and DHS surveys target "vulnerabilty" (incl. orphanhood), i.e. biol. parent(s) alive? and living in the household?
I attempted to complete the form after inputting additional information, but I encountered repeated rejection of question 10. Despite my efforts to troubleshoot the issue, I was unable to resolve it, suggesting a possible oversight on my part. Moreover, questions 11 and 14, which are contingent upon the identification of the biological mother and father in the roster of question 10, respectively, also present challenges due to the skip logic.
For further clarification and to enhance my understanding, I have attached the file. Your assistance in resolving these issues is greatly appreciated.