PUBLICATIONS
Council-developed Software

Frequently Asked Questions About the Household Registration System (HRS)
(Answers provided by Bruce MacLeod, Ph.D., University of Southern Maine)

What is the Household Registration System (HRS)?
The Household Registration System is a new software program designed to function as a template for generating Demographic Surveillance Systems (DSS) computer programs. The HRS aims to facilitate the collection, management, and analysis of data generated by DSS.

What does it cost?
It is available free to researchers for noncommercial use. There are two versions: HRS for DOS and HRS2 for use with FoxPro. They can be downloaded from the Web site of the Population Council.   

What kind of surveys can be conducted with HRS2^®?
The HRS2^® is a computer system for the automated generation of database software for longitudinal survival studies. The system maintains a consistent record of significant demographic events that occur to a population in a fixed geographic region. In addition, the HRS2 generates up-to-date registration books that can be used by field workers. HRS sites currently are using the program for research about acute respiratory infection, malaria and other diseases, health system reforms, child survival, and reproductive health.

Why was this program developed?
Critical research needs in the health sciences often require longitudinal population studies, but launching and sustaining such studies often require technical assistance that can be prohibitively expensive in developing countries. With support from the Thrasher Research Fund, the Population Council developed the HRS to reduce the cost and complexity of longitudinal research.    

What are the differences between the DOS-based version of the HRS and the Visual FoxPro version (HRS2)?
I believe that the HRS2 version is substantially better than the DOS version, for these reasons:

• It is quite easy to change ID schemes.
   
• It can accommodate non-resident individuals.
   
• It can keep track of many families in one location.
   
• It can keep track of social groups other than the family.
   
• It can accommodate many different types of relationships in addition to marital relationships.
   
• Data entry is more user-friendly.
   
• Consistency checks are tied to the data—you cannot bypass them by editing the data directly.
   
• Updates to tables are made in a group—if any update fails, none of the updates in a group will go through.

I suspect I have missed a few, but I hope that the list is compelling enough for you to consider using the next version of the HRS (HRS2).

The downside, of course, is that you need to learn a new software system. I believe it is worth it.

How different is the programming language in FoxPro5 from FoxPro 2.6?
It is quite different. There is an object-oriented architecture that allows for the development of sophisticated user interfaces. Database capabilities have also improved considerably with the addition of validation checks attached to the fields and records of the database.

Will FoxPro 5 run my older FoxPro 2.6 applications?
It will—there is an automated conversion program—but I was not pleased with the results. Sometimes it drops the validation checks and it also buries the screens below many layers of "containers."

How does Access compare to FoxPro for the development of longitudinal systems?
Access is a very reasonable database system. Some of the automated builders/wizards are very impressive. FoxPro should have more of these features. Although Access is appealing from a number of perspectives, I have these reservations:

• Access is object-based but has no inheritance—without inheritance you need to cut and paste. Access does not have good buffering capabilities—the key impact on our systems would be that the fields on the screen will probably write directly to the database table (alternatively you can use memory variables the old way). FoxPro buffering capabilities allow for storage in an intermediate buffered area.
   
• In Access, you cannot add properties or methods to screen controls (such as text boxes, command buttons, and the like).
   
• FoxPro has a true object-oriented architecture and Access does not. The object-oriented architecture can be used to embed error handling logic and consistency checking. It is very difficult to do this in an elegant way in Access. Active-X controls or adding new events are ways to get around this problem.
   
• Access's performance under large data loads is reported to be marginal. I have heard this from more than one source, so I am anxious to test the new system on a large number of records.

What are the steps involved in making a transition from my current system to the HRS2?
My biggest concern is that projects have the appropriate resources to make the HRS2 work for them. Here are some of the required resources:

1. A copy of Visual FoxPro, Version 5 or Version 6.
       
2. Books and manuals. The documentation manuals need to be purchased separately from the software. In addition, I recommend the following books:
       
   • The Fundamentals: Building Visual Studio Applications on a Visual FoxPro 6.0 Foundation
     Whil Hentzen. Edited by Doug Hennig.
     Hentzenwerke Press
     ISBN 0-9655093-5-4
       
   • Effective Techniques for Application Development with Visual FoxPro 6.0
     Jim Booth & Steve Sawyer, Ed. Steven P. Dingle
     Hentzenwerke Press
     ISBN 0-9655093-7-0
         
   • Advanced Object Oriented Programming with Visual FoxPro 6.0
     Markus Egger, Ed. Mac Rubel
     Hentzenwerke Press
     ISBN 0-9655093-8-9
     
   • Visual FoxPro 6 Enterprise Development.
     Paddock, Petersen, and Talmage,
     Prima Publishing
     ISBN 0-7615-1381-7   
      
   • any other Visual FoxPro books you find interesting.
      
3. At least one technical person. This person should know some database program and be willing to learning Visual FoxPro.
       
4. A plan to convert your existing data to fit into the HRS2 format. I imagine that this will require some collaboration between your group and me. I have been involved in converting data from three projects and it does require patience and persistence. I have some tools to help us.
       
5. Some adaptation/modification to fit your project. This includes the addition of new fields, modifying screens, and the like.
       
6. Once a version of the HRS2 is adapted to your site, I recommend that you run the system in parallel to the old system to make sure that it is working correctly. You should have a testing phase built into the transition.
            
7. The current documentation and technical manual  for the HRS2 is weak. As time permits, I will be improving this part of the system.

How do you model the population?
A paper that lays out the data model for the HRS2 can be found at:
http://indepth-network.org (click on Publications)

How do you handle the recording of revisits to a household?
I see two solutions to the problem of recording revisitations during a single round:

1) Field-based: Do not put the date of visit until you actually interview the household. I believe that this is the appropriate solution, but it may not be possible due to complications in the field operations.

2) Computer-based:

 

a) Do you have the revisit date? If so, then enter this date into the computer rather than the first visit date.
    
b) You could enter an amendment—add an additional visit and then add the event. This would require removing the rule that you can not have more than one visit per round. I can help you do this if you choose this route.

How do you handle migrations within the study where you do not know the ID of the internal migrant?
If you do not know the ID of the internal migrant, then Q is the appropriate value for the recall ID. The system will then assign an ID to this person. Later, when the ID is known, you edit the recall ID, change from Q to the known value and save. This operation should change all the ID values from the assigned ID to the value put into the recall ID field.

Could you tell me more about INDEPTH?
INDEPTH (International Network of Field Sites with Continuous Demographic Evaluation of Populations and Their Health) is a global network of DSS sites linked electronically. The group's Web site is dedicated to longitudinal and health research.

---