DataScientist

This working day is an example; employers organize the work differently.

// Selected official Tasks — DCWF 5.3

• 5120 · Core Task: Conduct hypothesis testing using statistical processes.
• 5924 · Core Task: Train and evaluate machine learning models.
• 5927 · Core Task: Write and document reproducible code.

// Selected official Knowledge — DCWF 5.3

• 21A · Core Knowledge: Knowledge of statistical/machine learning algorithms.
• 75A · Core Knowledge: Knowledge of mathematics, including logarithms, trigonometry, linear algebra, calculus, statistics, and operational analysis.
• 102 · Core Knowledge: Knowledge of programming language structures and logic.

// Selected official Skills — DCWF 5.3

• 172 · Core Skill: Skill in creating and utilizing mathematical or statistical models.
• 6490 · Core Skill: Skill in assessing the predictive power and subsequent generalizability of a model.
• 6651 · Core Skill: Skill in Regression Analysis (e.g., Hierarchical Stepwise, Generalized Linear Model, Ordinary Least Squares, Tree-Based Methods, Logistic).

DCWF describes work rather than setting pay. Compare jobs by duties, experience, location and employer.

// US pay and outlook context

BLS reports a US median annual wage of $120,230 in May 2025 and projected employment growth of 35% from 2025 to 2035 for data scientists. The median covers workers at different experience levels; it is not starting pay.

Read BLS: Data Scientists

// Check actual opportunities

• Compare the advertised salary range, experience requirements, working location and benefits.
• Separate internships and junior roles from senior positions. A national median is not a guaranteed salary.
• For federal jobs, read the grade, eligibility and education or experience requirements in the announcement.

These are suggested ways to build relevant skills. They are not an official DoD qualification pathway.

// Grades 9–12: start at school

Build algebra and statistics skills, then study probability, calculus and linear algebra as available. Learn Python and practice comparing two explanations rather than assuming the first pattern is the answer.

// Considering or studying a degree

Consider statistics, mathematics, computer science, data science or related engineering. Look for statistical reasoning, algorithms, experimental design and machine learning. Some research-focused openings may require graduate study; check the actual job posting.

// Early career and skills-first route

Search for data science internships and junior positions alongside analyst roles that include coding and statistics. Show that you can test a model fairly, write code others can run and explain errors.

// A project for your portfolio

Predict a value using a small public dataset. Keep some examples separate for testing, compare a simple method with your model and record the errors. Explain why success on this dataset may not transfer to another setting.

// Show what you can do

• Explain the question or goal, the source of your data and your own contribution.
• Include the steps you followed, checks you ran, results and limitations.
• Use public data with suitable permission or fictional examples. Do not use private student or employer information.

Credentials can help you structure learning or show knowledge. Choose them based on the work in your target job postings, alongside practical experience.

Official DoD 8140 guidance

Employers may use different titles for similar work. These links open US job searches; they are not endorsed vacancies.

Search federal opportunities on USAJOBS

// Match the work, not just the title

• Compare the duties with the role description. Similar titles can cover different work.
• Check the required experience and whether the opening is junior, specialist or leadership level.
• Federal and defense jobs may specify citizenship, clearance or other eligibility requirements; read the actual announcement.

// You may enjoy this if you…

• Enjoy math, coding and experiments.
• Like investigating results that challenge your first idea.
• Want to explain both the strengths and limits of a prediction.

// Think carefully if you…

• Only want to use AI tools without learning how to test them.
• Dislike statistics, debugging or experiments that do not work.