Office of Personnel Management

http://www.opm.gov/

Milestone 8 - August 31st 2015

OMB Review Complete: OMB has completed the agency review for this milestone. Agencies should contact their OMB desk officer if anything looks incorrect.

Leading Indicators

These indicators are reviewed by the Office of Management and Budget

Review Status complete
Reviewer Rebecca Williams
Last Updated October 6, 2015, 3:53 pm EDT by Rebecca Williams

Assessment Summary

OPM did not make progress this quarter on any of the IDC requirements.

Action Items: 1. This quarter OPM has not seen any growth in EDI/PDL datasets. While reductions in dataset count may be due to consolidating records, OPM must prioritize expanding their EDI/PDL every quarter until it is complete. 2. OPM only includes Public Domain/licensing information for 0.8% of their datasets, which is the same percentage as last quarter. To ensure the legal reuse of federal open data is conspicuous OPM must prioritize including an explanation for all datasets that do not include a Public Domain URL. 3. OPM must prioritize updating their engagement strategy to include a transparent two-way feedback mechanism and to inform data improvements to include in IDC Use & Impact reporting.

Inventory Composition

Public Dataset Status

Dataset Link Quality

Status Indicator Automated Metrics
Overall Progress this Milestone
Inventory Updated this Quarter
702 Number of Datasets
6 Number of APIs
Schedule Delivered Crawl details
1 Bureaus represented
8 Programs represented
603 Number of public datasets
56 Number of restricted public datasets
43 Number of non-public datasets
Inventory > Public listing
12.5% Percentage growth in records since last quarter
1180 Spot Check - datasets listed by search engine
Agency provides a public Enterprise Data Inventory on Data.gov
0.8% License specified Crawl details
Status Indicator Automated Metrics
Overall Progress this Milestone
624 Number of Datasets Crawl details
Number of Collections Crawl details
466 Number of Public Datasets with File Downloads Crawl details
6 Number of APIs Crawl details
756 Total number of access and download links Crawl details
Quality Check: Links are sufficiently working Crawl details
545 Quality Check: Accessible links Crawl details
199 Quality Check: Redirected links Crawl details
10 Quality Check: Error links Crawl details
2 Quality Check: Broken links Crawl details
0% Percentage growth in records since last quarter
624 Valid Metadata Crawl details
/data exists Crawl details
/data.json Crawl details
Harvested by data.gov
504 Views on data.gov for the quarter
Status Indicator Automated Metrics
Overall Progress this Milestone
Description of feedback mechanism delivered Crawl details
Data release is prioritized through public engagement
Feedback loop is closed, 2 way communication
See below Link to or description of Feedback Mechanism
https://www.opm.gov/about-us/open-government/contact-us/
Status Indicator Automated Metrics
Overall Progress this Milestone
Data Publication Process Delivered Crawl details
Information that should not to be made public is documented with agency's OGC
Status Indicator Automated Metrics
Overall Progress this Milestone
See below Open Data Primary Point of Contact
donna.seymour@opm.gov
POCs identified for required responsibilities
Status Indicator Automated Metrics
Overall Progress this Milestone
Identified 5 data improvements for this quarter
See below Primary Uses
(1) Research; research and insights for recommendations on improving government. (2) USAspending: inform data quality analysis as well as advocacy. (3) IGs: a group of volunteers, including Sunlight staff, are scraping the public IG reports from around the federal government. While most IGs make their reports publicly available, few do so in particularly useful ways. This project will allow sites like Sunlight's Scout.sunlightfoundation.com tracking tool to help users surface and understand IG reports. (4) regulations.gov: powers docketwrench.sunlightfoundation.com, allowing users to track rulemaking dockets and analyze large amounts of comments. (5) Program/performance audit (6) Improved economy and efficiency of federal programs. (7) Research on improving federal efforts to recruit, develop and retain expert employees.
Value or impact of data
See below Primary data discovery channels
Colleagues, email correspondence with agency officials, agency website
See below User suggestions on improving data usability
(1) A larger portion of the data on FedScope (http://www.fedscope.opm.gov/). (2) Easier time completing memoranda of understanding (MOUs) for sensitive data. (3) Data currently only available for download in PDF formats should be offered in other, more machine-readable formats. (4) Easier access. (5) More staff to handle requests. (6) Allow CPDF/EHRI to be connected to FEVS data. (6) EHRI payroll data need internal controls applied and a thorough scrub for validity and reliability. Reliability of certain fields in the personnel data are unknown too. Additional assessments of fields not covered in the 2013 report would be useful.
See below User suggestions on additional data releases
(1) Federal Employee Viewpoint Survey (EVS) data connected to anonymized Central Personnel Data File (CPDF -- replaced by Enterprise Human Resources Integration, or EHRI) data. (2) OPM should release the data that powers the Plum Book, which the GPO publishes. (3) See also the answer to the question about usability.
Digital Analytics Program on /data

Automated Metrics

These metrics are generated by an automated analysis that runs every 24 hours until the end of the quarter at which point they become a historical snapshot

data.json
Expected Data.json URL http://www.opm.gov/data.json (From USA.gov Directory)
Resolved Data.json URL http://www.opm.gov/data.json
Number of Redirects
HTTP Status 200
Content Type application/json
Valid JSON Valid
Detected Data.json Schema federal-v1.1
Datasets with Valid Metadata 100%(624 of 624)
Valid Schema Valid
Datasets 624
Datasets with Distribution URLs 75.3% (470 of 624)
Datasets with Download URLs 74.7% (466 of 624)
Total Distribution URLs 756
Total Download URLs 535
Total APIs 6
Public Datasets 541
Restricted Public Datasets 53
Non-public Datasets 30
Bureaus Represented 1
Programs Represented 8
License Specified 0.8% (5 of 624)
Datasets with Redactions 0.0% (0 of 624)
Redactions without explanation (rights field) 0.0% (0 of 624)
File Size 1.22MB
Last modified Monday, 02-Mar-2015 16:04:35 EST
Last crawl Monday, 31-Aug-2015 00:00:22 EDT
Analyze archive copies Analyze archive from 2015-08-31
Nearby Daily Crawls
/data page
Expected /data URL http://www.opm.gov/data (From USA.gov Directory)
Resolved /data URL http://www.opm.gov/data/
Redirects 1 redirects
HTTP Status 200
Content Type text/html; charset=utf-8
Last crawl Monday, 31-Aug-2015 00:00:15 EDT
/digitalstrategy.json
Expected /digitalstrategy.json URL http://www.opm.gov/digitalstrategy.json (From USA.gov Directory)
Resolved /digitalstrategy.json URL http://www.opm.gov/digitalstrategy.json
Redirects
HTTP Status 200
Content Type application/json
Valid JSON Valid
Last modified Friday, 03-Apr-2015 09:22:08 EDT
Last crawl Monday, 31-Aug-2015 00:00:15 EDT
Digital Strategy

Date specified: Monday, 16-Mar-2015 00:00:00 EDT

Date of digitalstrategy.json file: Friday, 03-Apr-2015 09:22:08 EDT

1.2.4 Develop Data Inventory Schedule - Summary

Summarize the Inventory Schedule


Our open data leads continue to work with program offices to document data definitions and standards and publish releasable metadata. We have worked one-on-one with these offices to obtain the metadata and spread awareness about the importance of providing open data in machine-readable formats. Now that our inventory has begun to reach a good level of maturity, we are enriching and opening our inventory by providing more detailed metadata, better organizing our inventory, and providing more data to the public in machine-readable formats. One example of better organization was made possible with the late 2014 release of version 1.1 of the metadata schema: parent-child relationships. We are taking full advantage of this new feature. Since completing the initial inventory in November 2014 and moving to version 1.1 of the schema in February 2015, we have begun to turn our attention to providing machine-readable versions of data that we currently make available only in formats such as PDF. However, we continue to identify data assets that have fallen through the cracks and add them to the inventory. Besides maintaining the inventory, we will conduct qualitative analysis to develop categories among the data assets and modify our Metadata Repository (MDR) to incorporate these categories and accommodate the data assets.

1.2.5 Develop Data Inventory Schedule - Milestones

TitleData asset identification
DescriptionIdentification of data assets from throughout OPM
Milestone Date2014-02-28
Description of how this milestone expands the InventoryIncreased the number of data assets in the inventory
Description of how this milestone enriches the InventoryProvided a broader picture of OPM's data
Description of how this milestone opens the InventoryCreated a foundation for public release of data
TitleWider data asset identification
DescriptionIdentification of data assets at a more granular level and from a wider range of programs
Milestone Date2014-05-31
Description of how this milestone expands the InventoryIncreased the number of data assets in the inventory
Description of how this milestone enriches the InventoryProvided an even broader and deeper picture of OPM's data
Description of how this milestone opens the InventoryBuilt a stronger foundation for public release of data
TitleTargeted asset identification
DescriptionIdentification of missing data assets and targeting them for inclusion
Milestone Date2013-08-30
Description of how this milestone expands the InventoryIncreased the number of data assets in the inventory
Description of how this milestone enriches the InventoryContinued to expand our understanding of OPM's data
Description of how this milestone opens the InventoryHelped identify data for immediate or future release
TitleComplete inventory
DescriptionRelease of the complete inventory
Milestone Date2014-11-30
Description of how this milestone expands the InventoryBrought stragglers into the fold
Description of how this milestone enriches the InventoryProvided a nearly full understanding of OPM's data
Description of how this milestone opens the InventoryMade as much of the full inventory available to the public as legally appropriate and practicable
TitleMetadata Schema v. 1.1
DescriptionMoved the inventory files to v. 1.1 of the metadata schema. In the process, designated some entries as parents and created hundreds of children, along with parent-child relationships
Milestone Date2015-02-28
Description of how this milestone expands the InventoryProvided individual entries for child datasets
Description of how this milestone enriches the InventoryParent-child relationships show how one dataset connects to another. This change helps the public better understand OPM's data
Description of how this milestone opens the InventoryProvided more links to OPM's publicly available data, organized in a useful way
TitleDeeper Look into Employee Services
DescriptionFocused on Employee Services data to fill in holes about areas such as employee relations, labor relations, veterans employment, and official time for unions representing federal employees
Milestone Date2015-05-31
Description of how this milestone expands the InventoryAdds entries regarding Employee Services
Description of how this milestone enriches the InventoryProvides a fuller understanding of OPM's government-wide policies
Description of how this milestone opens the InventoryFederal job series data are now available in a machine-readable format (CSV), making the data easier to work with

1.2.6 Develop Customer Feedback Process

Describe the agency's process to engage with customers


On both http://www.opm.gov/digitalstrategy and http://www.opm.gov/data, we have placed links to the open government topic area in our frequently asked questions system. Users can submit questions or requests for  data using that system. OPM's open government staff redirects these requests to the proper program offices. The program office staff are responsible for responding to these requests. OPM's Data Governance Board considers remaining requests monthly and works with program offices to release that data as practicable.

1.2.7 Develop Data Publication Process

Describe the agency's data publication process


The Data Governance Board (DGB) oversees the processes of inventorying and releasing data. The members of the DGB are technically competent and represent OPM's major data owners. We are currently focusing on inventorying our data and are temporarily accepting program offices' assessments of the public access level of their data assets. However, as we further open our data, when a program office deems that the data's access level should be public, the DGB will offer a determination as to whether the data could be harmful because of the mosaic effect (i.e., two or more independently harmless data sets could be compared or unified to inadvertently identify an individual or otherwise cause harm) or other potential negative consequence that was unforeseen by the program office. If the DGB does not find any such potential harms, it will recommend to OPM's Investment Review Board (IRB) that the data be released. The IRB, in turn, will make a recommendation to OPM's Office of the General Counsel, which will make the final determination. Any program office that labels its data restricted public or non-public is required to provide justification. Reasons for restricting some of the data but not all of it (restricted public) fall primarily into the category of personally identifiable information, or PII. Social security numbers are the most obvious example, but information about law enforcement or homeland security personnel that could make them easy to target would also be off-limits. In these cases, we will release non-attributable or less granular data at the appropriate time. Reasons for restricting all of the data (non-public) will fall primarily into the category of security. For example, certain data about Continuity of Operations (COOP) are not be releasable because they could compromise the agency's ability to operate securely in a national emergency.