Allow search of assignments/interviews by enumeration area

assignments
interviews
headquarters
supervisor
(Arthur Shaw) #1

Use case: Suppose that data collection is done by mobile teams that progress enumeration area (EA) by EA. Many headquarters and supervisor actions would want to address EAs. Headquarters would want to review all interviews in just completed EAs. Supervisors would want to confirm that all interviews have been completed and reviewed for the EA that the team is planning to leave soon. Supervisors would want to make all assignments for the next EA. But there is no easy way to query assignments or interviews by EA.

Desired changes: While Survey Solutions has no way of guessing how assignments/interviews map to EAs, this mapping could be provided by a simple set of user inputs. When questionnaires are imported, users could indicate which combination of variables uniquely identify an EA. Alternatively, users could add an optional _ea or _cluster variable to the tab-delimited file that defines assignments. In the former case, SuSo could construct an EA identifier. In the latter case, the end user would provide the EA identifier. (In many cases, such an identifier is already among the user-provided set of identifying variables being pre-populated in the assignments file.) In both cases, SuSo would have what it needs to provide an EA filter for assignments/interviews. (Obviously, if the user provides no inputs to identify EAs, SuSo cannot enable the EA filter.)

Current work-around: In the absence of an EA identifier, users can use the text field to query assignments/interviews, and return the assignments/interviews that match the search terms (i.e., EA identifiers).

Further notes: This new functionality–particularly if implemented via a user-provided EA identifier–could open the door for other beneficial features. When there are multiple HQ users, they are often assigned (outside of SuSo) to monitor work for certain EAs. If an EA variable existed, SuSo would have the necessary inputs for providing a filtered view to that HQ user of his/her area of responsibility. When supervisors/interviewers are assigned assignments, these are often done by EA. For example, supervisor A gets EAs 1-10, while supervisor B gets EAs 11-20. If the EA variable existed, this could offer HQ another way of making assignments that is closer to they way they currently work. In effect, this EA to supervisor mapping could replace the optional _responsible column in the tab-delimited assignment file.

(Efoster1) #2

Also, the way our survey is set up is that a number of EAs are assigned to each supervisor, who then assigns each EA to one interviewer on the team. There is no easy way to select all households in one EA and assign them all to one interviewer.
The “Identifying Questions” column shows the variable names, but you can’t include them in the search. Even though “cluster code: 2” appears, you can’t search for “cluster code: 2” and you enter just 2, you also get every cluster and phone number and everything with a number 2 in it.