usaspending-forum
[Top] [All Lists]

[usaspending-forum] Data Quality Problems with Agency Codes, including e

To: usaspending-forum@xxxxxxxxxxxxxx
From: susan.turnbull@xxxxxxx
Date: Mon, 31 Dec 2007 09:42:10 -0500
Message-id: <OF0376ED95.6A4833BE-ON852573C2.005013B3-852573C2.0050C3BC@xxxxxxx>
(Message received by STurnbull from Cedric Stroud, for posting to 
USASpending forum, including Data Concerns Table missing from the first 
posting)    (01)

I think modeling the USASpending.org site after the OMBWatch site, was a 
good decision. Searching and sorting on the Federal funding data, and 
quickly being able to retrieve data with options from summary records down 
to full, detailed records will make this site useful for a wide audience 
of researchers.     (02)

One big concern I have is with a problem that should be fixable without 
too much effort. There needs to be better validation of the Agency codes. 
I have been working with the Federal funding data since the late 1980's, 
so I am aware of the various publications used for verifying that this 
data meets certain Federal standards.    (03)

FIPS 95 and FIPS 55 have been important tools for reporting data 
accurately.  Even if a document is outdated, for instance FIPS 95-1,it 
still should be referenced when an unfamiliar Agency code appears in the 
Contracts or Assistance data.     (04)

My concern for the immediate future is who will be making sure "Unknown" 
doesn't appear in data that obviously should save that information?    (05)

If the Federal Government awards money to a recipient, I don't think a 
Congressman nor a citizen wants to see a record that indicates that the 
Agency that gave away millions or even billions of dollars can't be 
identified.  I know that is not the case, based on my research on your Web 
site, but I recommend cleaning it up early before someone thinks there is 
a greater problem.     (06)

My concern for the long-term future is as to what mechanisms Federal 
agencies have implemented to insure the quality of the data that is being 
reported to the USASpending site? If an agency doesn't appear to know its 
own agency codes and/or if GSA isn't validating the agency codes of the 
data that is submitted, it gives me less confidence about the rest of the 
data.  The FIPS publications that I mentioned are going and agencies are 
supposed to make sure their data is current and accurate, so it is 
important that each one understands how important this is and enacts the 
appropriate measures.     (07)

I have included in the table below some of the types of problems I 
noticed.  I have also included references to relevant documentation to 
help in resolving them.     (08)

I have not written this email to cause anyone harm or embarrassment, but 
the nature of this data requires that all efforts are taken to insure its 
accuracy.  With that in mind, you don't have to publish this email to your 
public forum, so long as the issues are addressed. You are free to decide 
that one way or another.     (09)

You can also contact me if you have any questions.     (010)

Thanks.     (011)

See USASpending Web Site - Data Concerns Table 
http://colab.cim3.net/file/work/USAspending/comments/USASpendingWebSite_DataConcerns.xls    (012)




Cedric Stroud 
Sr. Software Engineer 
U.S. House of Representatives 
House Information Resources 
CAO Advanced Business Solutions 
Web Solutions Branch 
2nd and D Sts., S.W., Room 640 FHOB 
Washington, D.C. 20515 
202-226-6438     (013)

How am I doing? Please take a few minutes to complete this survey.  Thank 
you. 
http://housenet.house.gov/keywords/survey/web     (014)

CAO Web Assistance Hotline: 202-226-2140 
CAO Web Assistance: webassistance@xxxxxxxxxxxxxx     (015)



_________________________________________________________________
Message Archives: http://colab.cim3.net/forum/usaspending-forum/
Subscribe/Unsubscribe/Config: 
http://colab.cim3.net/mailman/listinfo/usaspending-forum/
Shared Files: http://colab.cim3.net/file/work/USAspending/  
Community Wiki: http://colab.cim3.net/cgi-bin/wiki.pl?USAspendingGov
Community Portal: http://colab.cim3.net/
To Post: mailto:usaspending-forum@xxxxxxxxxxxxxx    (016)
<Prev in Thread] Current Thread [Next in Thread>
  • [usaspending-forum] Data Quality Problems with Agency Codes, including examples, susan . turnbull <=