Agencies looking to recruit data scientists now have a clearer standard of what a qualified candidate should bring to the job.
The Office of Personnel Management last week released its new occupational series for data scientists in government, a project that’s been in development for more than two years.
OPM’s occupational series for data scientists sets a foundation that will help agencies identify prospective employees with the appropriate skills, and will help build out a federal community of practice around data modernization that gained momentum in recent years with the creation of the Chief Data Officers Council.
CDOs reported significant progress in meeting early goals under the 10-year Federal Data Strategy, but repeatedly sought to increase staffing and raise the overall data literacy of the federal workforce.
More than half of respondents in a recent Data Foundation and Grant Thornton survey of federal CDOs said their office has 10 or fewer full-time staff, and 50% said they hire contractors to help with data product design, data operations and data visualizations.
OPM worked with the Census Bureau, the Chief Information Officers Council, Office of Management and Budget and other agency partners to create the new data scientist series.
The occupational series from OPM builds on guidance it issued to agencies in June 2019, which gave agencies the flexibility to add a parenthetical “data scientist” tag onto positions “that perform data science work as a major portion of the job, and not as a collateral duty.”
Kimberly Holden, OPM’s deputy associate director for talent acquisition, classifications and veterans programs, in a memo to agency human resources directors, said data science work may be found across a range of existing occupational series, including health care, statistics and IT specialist positions.
Holden, in the memo, said agencies can continue to add “data scientist” to job titles across various federal occupations series, in an effort to maximize the flexibility agencies have to manage their workforces.
“Data Scientist work is multifaceted and requires talent from interdisciplinary backgrounds,” Holden wrote.
OMB, in the 2021 action plan for the Federal Data Strategy, said OPM expected to issue the new data scientist job series before the end of the year, which would help agencies more easily recruit in-demand data talent.
The Foundations for Evidence-Based Policymaking Act that Congress passed in January 2019 required OPM to create the data scientist job series.
The 2022 National Defense Authorization Act that President Joe Biden signed Tuesday also directs OPM to “establish or update one or more occupational series” covering federal jobs in data science and data management, as well as software development and software engineering jobs.
Data Foundation Policy Director Corinna Turbes said this provision in the NDAA would help create a “federal workforce that is equipped to manage and leverage data across federal agencies.”
“Improving the recruiting and retention of top talent in the federal government’s workforce is crucial to creating data literate organizations that can fully realize the benefits of leveraging data for evidence-informed decision making,” Turbes said.
Agencies have 12 months to use this new occupational series to classify data science jobs, but must implement the qualifications standards immediately for new hires.
OPM defines the basic qualifications for data science positions as a bachelors-level degree in mathematics, statistics, computer science, data science or a related field.
The agency also considers candidates eligible for data science positions if they have a “combination of education and experience” equivalent to a four-year degree.
OPM’s data scientist job series covers positions that involve pulling and interpreting data sets, as well as visualizing them to highlight their findings.
“The primary purpose of data scientist work is to provide managers or leadership with objectively based information derived from the use of advanced scientific, mathematical, and statistical principles and theories,” OPM states in its position classification. “The results of this data will be used for making data driven decisions regarding the administrative and programmatic aspects of agency operations and management.”
The work of data scientists, OPM notes, may overlap with some of the required skills for related fields, such as computer science, mathematical, statistical, and IT.
The memo also differentiates the roles and responsibilities of data scientists from similar positions, include data analysts, data architects and data engineers.
“Data Scientists primarily apply advanced quantitative, computer science, mathematical, and statistical skills to interpret extremely large and unstructured data sets and asks questions regarding the unknown. This differs from other Data Science work roles such as Data Analysts, Data Architects, and Data Engineers. These other work roles may not require the same skill sets as a Data Scientist and therefore may not be appropriately classified in this series,” the memo states.
The work of federal data scientists includes analyzing performance data to evaluate the effectiveness of agency programs. OPM also expects qualified data scientist candidates to have a baseline proficiency in machine learning and artificial intelligence tools.
“Data Scientist[s] use technology to wrangle enormous data sets and work with complex algorithms. The tools used by data scientist[s] are complex, requiring expertise in coding, prototyping, and integration with complex data systems. The exploration of big data by data scientist[s] requires the use of powerful technology to explore data — such as machine learning and artificial intelligence,” the memo states.
OPM conducted a comprehensive study of data science work, which included focus groups with agency human capital officials, subject matter experts and leadership.
The data scientist occupational series marks the latest in a series of governmentwide efforts this year aimed at bringing data experts into government, and raising the overall data proficiency of the federal workforce.
OPM worked with the Chief Data Officers Council and U.S. Digital Service in January to lead a governmentwide data science hiring initiative.
The initiative gave candidates an opportunity to apply to 50 data science positions through a single application. Ten agencies put out the call to hire program analysts at the GS-13 and 14 levels with at least a year’s worth of prior experience.
Applicants went through a Subject Matter Expert Qualification Assessment (SMEQE) that gave data scientists already working at agencies a seat at the table vetting applications, rather than rely on candidates assessing themselves. OPM received more than 500 applications in less than 48 hours.
OMB also saw its first cohort of federal employees to graduate from its data science upskilling program in Mya.
OMB accepted 61 federal employees from 20 agencies to pilot its Federal Data Science program. Agencies put a call out for nominations and submitted names for the program between July and August last year. Accepted employees then went through training from September through January.