I am building a form where I have alternate between male and female respondents. E.g.,
I go to household 1, the respondent is a male.
In the sampled household 2, the kobo collect needs to prompt enumerators that your last survey was with a male and this survey should be with a female. However, consider a case where no one is at home and data on gender is not collected in this household.
When respondent goes to household 3, ODK collect needs to remember responses from household 1 and prompt enumerators to survey a female.
Challenge arises when gender is not recorded in a form if the household is empty.
@Ashwini, in case 2, if there is no one in the household maybe the enumerator will not be filling the records from the household. Hence in this case, the third household becomes the second household. So, here you should be able to know that male has already been recoded previously and now the female should be recorded.
We will record data for second household which will indicate that there is no one in the household. However, gender information will not be collected. So this second household will still be saved without gender. Using the last-saved function, the kobo is pulling missing data for gender in third form as there is no gender recorded in second saved form.
@Ashwini, what would it look like to have the screening question at the very beginning (even before the informed consent)? Doing so, the survey could stop at that very first screening question and you would not even reach the respondent's gender question. In this way, maybe you should be able to achieve what you have planned.
But our community should still be able to provide you a different approach which should be much easier.
Maybe you could add a variable in each case, which will store the “last” available gender information before this case. So, you could take this for going on if the new case doesn’t get gender info. (You might also need to set a “default” if the very first case has empty gender information).
