Written by: Director Robert Groves
At this writing we have completed roughly 62% of the nonresponse interview workload. Our field staff have done a wonderful job especially dealing with balky computer control issues.
I’ve blogged before on problems we’ve had with the software that we use for nonresponse followup.
We have about 500 local census offices that that assign addresses to enumerators for visits to conduct short interviews. These are the addresses that either did not receive a mailed form or did not return one. Close to 600,000 enumerators depend on these offices to provide them with work assignments, receive their daily report of hours, check over their work product, process payroll data, select cases for quality-control reinterviews, and oversee productivity of the work teams.
The “paper-based operation control system” contains a set of software functions that were developed starting in 2008, when the decision was made not to use handheld computers during the nonresponse followup stage.
In the first week of nonresponse followup, the system was giving very slow response times to users throughout the country. This didn’t hurt the assignment of cases to enumerators; they continue to conduct interviews at roughly the right pace. It did affect the checking-in of completed interviews.
Hour-by-hour study of the performance of the system led to some tweaks of the handling of the database, as well as the creation of work-arounds for some functions of the system. Over the past week or so, the performance of the check-in function has improved. We’re nearly through an excessive backlog of completed cases that piled up in offices and are approaching a checked-in total that we wanted at this point. Things are going much better than two weeks ago.
The staff in the local census offices and at headquarters have been stellar in their teamwork to identify problems and forward inventive suggestions to make real-time improvements in the work. I’m proud to have them as colleagues.
We’ve not achieved a trouble-free status and constant vigilance is needed. We are inventing possible work-arounds for the next operations in the census. We need wise management of the use of the software for those operations.
Please submit any questions pertaining to this post to ask.census.gov