CDS logo
Focused certification exam prep
Start practice

What Is A CDS?

TL;DR
  • CDS here means Certified Data Scientist - Associate/Practitioner, issued by the Institute for the Certification of Computing Professionals (ICCP).
  • The accessible Data Science Exam outline lists 12 official topics, from business questions through reproducibility, without published percentage weights.
  • The ICCP store lists $350 per professional exam and $249 per foundational exam; confirm which applies to your tier.
  • Active certification runs on a 3-year professional-development cycle; general ICCP tiers cite 60 hours (Associate) or 120 hours (Practitioner).

What "CDS" Means on This Page

"CDS" is an overloaded acronym. In finance, medicine, government, and software it stands for entirely different things, and a quick web search for the letters will mix them together. On this site, CDS means Certified Data Scientist - Associate/Practitioner, a professional credential for people who work with data to answer business and research questions. If you arrived here wondering what a CDS is in the data-science sense, you are in the right place. If you want the short-form definitions, our pages on what CDS stands for and CDS meaning cover the vocabulary question directly.

The credential is built around a simple premise: data science is not just modeling. A certified data scientist is expected to start with the business problem, understand where data lives, apply sound statistics, write code, clean messy inputs, explore, model, infer, predict, reason about causality, communicate findings in writing, and make the work reproducible. That arc is exactly what the 12 official exam topics trace.

Who Issues the Credential

The certifying body is the Institute for the Certification of Computing Professionals (ICCP). ICCP is a long-standing certification organization for computing and information-technology professionals, and it runs its credentials through a tiered structure. The Associate and Practitioner labels in the CDS name reflect that tiering: they signal the level at which a candidate is certified, not two separate subjects.

Because the credential comes from a professional certification institute rather than a vendor or a university, the emphasis is on demonstrated competence plus professional conduct. Candidates agree to an ethics statement as part of the general ICCP application, and certification is maintained through ongoing professional development rather than a one-time exam. For a fuller picture of how the program is framed, see our overview of what CDS certification is.

Why the tier matters: General ICCP rules attach different maintenance requirements to the Associate and Practitioner tiers. Which exams and how many a given CDS tier requires has not been verified here, so treat any claim about a "single exam" or "two-exam" path with caution until you confirm it with ICCP directly.

The 12 Official Topics, Grouped by How You Will Use Them

The accessible Data Science Exam outline lists 12 official topics. It does not publish percentage weights or a revision year, so no one can honestly tell you that a given topic is "20% of the exam." What you can do is understand each topic and study in proportion to your own weak spots. For a topic-by-topic walkthrough, see the complete guide to all 12 CDS content areas. Here is a practical grouping.

Framing the work

Domain 1: Business & Technology Issues: Starting with the Question First - What Problem are you Trying to Solve?

The credential opens with the question, not the algorithm. Candidates should be able to translate a vague business need into a defined analytic problem.

  • Distinguishing a real business problem from a request for a specific technique
  • Understanding the technology environment the analysis must live inside
  • Recognizing when data science is, and is not, the right tool

Domain 5: The Data Analytic Question and Types of Data and Reporting

Different questions call for different analyses, and different data types constrain what you can do with them.

  • Matching the type of question (descriptive, exploratory, inferential, predictive, causal) to the approach
  • Recognizing data types and what each permits
  • Understanding how results will be reported to the audience that asked

Data foundations

Domain 2: Data Storage, Big Data and Sources

Where data comes from and how it is stored shapes everything downstream.

  • Common storage approaches and the trade-offs between them
  • What "big data" changes about how you work
  • Evaluating the reliability and origin of data sources

Domain 4: Programming Skills

Data science is executed in code. This topic checks that you can reason about programming tasks used in analysis.

  • Reading and reasoning about analytic code
  • Manipulating and transforming datasets programmatically
  • Understanding how code structure supports repeatable work

Domain 6: Tidying the Data - Data Cleaning and Quality

Real data is rarely analysis-ready. This is where much practitioner time goes, and the exam reflects that.

  • Identifying missing, inconsistent, and malformed values
  • Assessing data quality before trusting any result
  • Structuring data so it can be analyzed cleanly

Quantitative core

Domain 3: Mathematics and Statistical Data Science

The mathematical and statistical vocabulary that everything else relies on.

  • Core probability and statistical concepts
  • The mathematics that underpins common methods
  • Reading statistical output with understanding rather than by rote

Domain 7: Exploratory Analysis

Looking at data before committing to a model.

  • Summarizing and visualizing distributions and relationships
  • Spotting outliers, patterns, and anomalies
  • Using exploration to refine the original question

Domain 8: Statistical Modeling and Inference

Drawing conclusions about a population from a sample, with honest uncertainty.

  • Fitting and interpreting statistical models
  • Quantifying uncertainty and understanding what inference does and does not justify
  • Knowing the assumptions behind the methods you use

Domain 9: Prediction and Machine Learning

Building models whose goal is accurate prediction rather than explanation.

  • The distinction between predictive and inferential goals
  • Evaluating predictive performance without fooling yourself
  • Recognizing overfitting and how it distorts results

Domain 10: Causality

The hardest conceptual shift: when does an association justify a claim about cause?

  • Why correlation alone does not establish causation
  • Study design and the reasoning behind causal claims
  • Confounding and how it undermines naive conclusions

Communication and trust

Domain 11: Written Analysis

An analysis nobody can understand does not change a decision. This topic treats clear writing as a core data-science skill.

  • Structuring a written analysis for a stated audience
  • Reporting findings, limitations, and caveats honestly

Domain 12: Reproducibility

Whether someone else, or you in six months, can rerun the work and get the same answer.

  • Documenting data, code, and decisions so the analysis can be repeated
  • Why reproducibility is a quality and trust issue, not a nicety

How Testing and Registration Work

General ICCP examinations use approved proctoring, and the options include arranged remote or in-person sessions. The specific testing vendor and workflow for the CDS have not been verified here, so confirm scheduling mechanics with ICCP rather than assuming they match another certification you have taken. Two items that candidates often ask about, whether the exam is open-book and whether it is adaptive, are likewise unverified. Plan on a conventional proctored session until the issuer tells you otherwise.

The general ICCP application asks for a resume and an agreement to the ICCP ethics statement, along with a candidate application fee. The exact CDS education, experience, training-hour, and reference prerequisites need to be confirmed with the issuer, so do not treat any blog post, including this one, as a substitute for ICCP's own eligibility language. Our CDS requirements guide collects the eligibility questions you should be asking, and the CDS exam dates page covers scheduling considerations.

Verify before you pay: The calculator policy and delivery format are not confirmed. If you plan to use a calculator or expect a particular question navigation style, ask ICCP in writing before test day so you are not surprised.

Fees, Tiers, and Maintenance Hours

The published numbers that can be stated with confidence are limited, and it is better to say so than to guess. Here is what the issuer's materials support:

ItemWhat is known
Professional exam fee (ICCP store)$350 per exam, including proctoring
Foundational exam fee (ICCP store)$249 per exam
Older FAQ figure$299 (superseded by current store listing)
Candidate application fee (general)$45
Associate maintenance (general ICCP tier rule)60 hours per cycle, $35 annual fee
Practitioner maintenance (general ICCP tier rule)120 hours per cycle, $75 annual fee
Active certification cycle3-year professional-development cycle

Notice what is missing: a total certification-exam fee. Which exam SKUs the CDS tier requires, and how many, is not verified, so adding up a single total would mean inventing a number. Likewise, the Associate and Practitioner maintenance figures are general ICCP tier rules, and their applicability to a specific candidate should be confirmed. For a fuller walkthrough of how these pieces combine, see the CDS certification cost breakdown.

Key Takeaway

Budget in layers: application fee, then exam fee or fees (current store pricing differs between professional and foundational exams), then ongoing maintenance hours and annual fees. Ask ICCP which exam SKUs your tier requires before you commit to a number.

Who Hires Certified Data Scientists

The CDS is aimed at people who do or want to do analytic work across industries, and the 12-topic structure shows the kinds of roles it maps to. Organizations that need people who can move from a business question through cleaning, modeling, and written communication tend to value this breadth. Typical employers include companies with analytics or business-intelligence functions, consultancies that deliver analysis to clients, and public-sector or research organizations that need defensible, reproducible results.

Certification does not replace a portfolio or experience, but it can signal that you understand the full workflow, including the less glamorous steps like data quality and reproducibility that hiring managers often find lacking. Job titles vary widely, from analyst and data scientist to analytics consultant, so read postings for the underlying duties rather than the label. Our pages on CDS jobs and the CDS salary guide go deeper, and the ROI analysis helps you weigh the credential against the cost. No specific salary figures are asserted here because none are verified for this credential.

Sequencing the 12 Topics in Your Study Plan

Generic study advice is everywhere; what is useful is deciding which CDS topics to tackle in which order and why. A sensible arc follows the way the topics build on each other. Treat this as a template and stretch or compress it to fit your own background.

Weeks 1-2

Foundations first

  • Domain 1 (problem framing) and Domain 5 (question and data types) set the vocabulary used everywhere else
  • Domain 3 (math and statistics) is the heaviest prerequisite, so start it early rather than last
Weeks 3-4

Data handling

  • Domain 2 (storage and sources), Domain 4 (programming), and Domain 6 (cleaning and quality)
  • These are practical; work examples on real messy data rather than only reading
Weeks 5-7

Analysis and modeling

  • Domain 7 (exploration), Domain 8 (modeling and inference), Domain 9 (prediction)
  • Keep inference and prediction clearly separated in your notes
Week 8

Reasoning and communication

  • Domain 10 (causality), Domain 11 (written analysis), Domain 12 (reproducibility)
  • Finish with a timed practice pass across all 12 topics

Place causality after modeling and inference on purpose: you cannot reason well about cause until you are comfortable with what a model and a confidence statement actually claim. For a deeper plan, use the CDS study guide, and keep the one-page cheat sheet nearby for final review. To test yourself against exam-style questions, try the CDS practice tests on the main site.

Common Mix-Ups to Avoid

Confusing this CDS with others

Because the acronym is shared, searches for "CDS pass rate" or "CDS passing score" can surface information about unrelated credentials. Treat any number you find with suspicion unless it is tied specifically to the ICCP Certified Data Scientist. No pass rate or passing score is asserted on this page because none is verified for this credential; if you want to see how we handle the question, read about the CDS pass rate and the CDS passing score.

Assuming it is only about machine learning

Only one of the 12 topics is explicitly about prediction and machine learning. The rest cover framing, data, statistics, communication, and reproducibility. Candidates who study only modeling tend to be underprepared for the cleaning, causality, and written-analysis topics.

Treating the outline as a full blueprint

The accessible outline lists the 12 topics but is not a verified complete multi-exam pathway blueprint. Use it as your content map, but check with ICCP about how the full certification path is assembled. If you are still weighing difficulty, the difficulty guide discusses what makes the content challenging without claiming unpublished statistics.

Frequently Asked Questions

What does CDS stand for on this site?

It stands for Certified Data Scientist - Associate/Practitioner, a credential issued by the Institute for the Certification of Computing Professionals (ICCP). It does not refer to any other credential that happens to share the acronym.

How many topics does the CDS exam outline cover?

The accessible Data Science Exam outline lists 12 official topics, running from business and technology issues through reproducibility. The outline does not publish percentage weights or a revision year.

How much does the exam cost?

The ICCP store lists $350 per professional exam, including proctoring, and $249 per foundational exam. An older FAQ cited $299. Which exams your tier requires is unverified, so no total is stated; confirm with ICCP.

How do I keep the certification active?

Active certification runs on a 3-year professional-development cycle. General ICCP tier rules cite 60 hours and a $35 annual fee for Associate, and 120 hours and a $75 annual fee for Practitioner. Confirm applicability to your situation with ICCP.

Is the CDS exam open-book or adaptive?

Neither open-book status, calculator rules, nor adaptive delivery has been verified for this exam. Ask ICCP before test day rather than assuming the format matches another certification you have taken.

Ready to pass your CDS exam?

Put this into practice with free CDS questions across every exam domain.