How to Prepare for CDMP® Data Modelling Specialist Exam

Key Takeaways

Exam Overview & Strategy: Prepare for the CDMP exam by tagging DMBOK for easy concept location during the test.
Practical Training Methodology: The preparation uses Steve Hoberman’s question-based method for building relational and dimensional models.
Core Exam Categories: The exam tests DMBOK fundamentals, focusing on introduction, activities, tools, and best practices.
Physical Implementation Challenges: Exam challenges include managing supertypes, subtypes, and the implications of denormalisation on performance.
Lineage and Business Communication: Data models aid business communication by clarifying horizontal and vertical lineage and validating software requirements.
Schemas and Notations: The exam tests modelling notations and the translation of logical models for various database types.
Visualisation and Readability: Normalisation is essential, but clear visual design ensures models are easily understood by stakeholders.
Data Modelling Governance: Assess modelling practices, integrate into SDLC, maintain scorecards, and capture metadata within models.
Exam Pitfalls and Practice: The exam includes tricky formats; practice quizzes help identify knowledge gaps before testing.

Webinar Details

Title: How to Prepare for CDMP® Data Modelling Specialist Exam
Date: 2023-08-10
Presenter: Howard Diesel
Meetup Group: African Data Management Community Forum
Write-up Author: Howard Diesel

How can Attendees Benefit from the Session?

This webinar details a session focused on the Data Modelling and Design specialist exam, following a prior session dedicated to data governance. Howard instructs attendees to enrol in the Canvas platform to access relevant sample questions designed specifically for data modelling preparation. The session outlines the upcoming schedule, noting that the subsequent week will cover data quality, followed by comprehensive overviews of other specialist courses.

Figure 1 CDMP Specialist Exams

Why is it Important to select the Appropriate Specialist Exam Based on one’s Professional Role?

Howard emphasises the importance of selecting the appropriate specialist exam based on an individual’s professional role and departmental requirements. He explains the exam’s structure, categorising questions into practitioner levels (derived from the DMBOK), associate levels (linked to other DMBOK areas), and master levels (requiring industry experience). A highly recommended preparation strategy is to tag the DMBOK reference material to efficiently locate essential concepts during the open-book examination.

Furthermore, attendees are informed that data modelling principles overlap significantly with data governance and master data management, accounting for 70% to 80% of the cross-knowledge-area questions. Attendees aiming for a master-level certification are also advised to be prepared to retake the exam, as the rewrite fee ($211) is significantly lower than the initial attempt ($311).

Figure 2 CDMP Specialist Exam Q&A

Figure 3 Specialist Exam Sample Quiz Using Instructure

Figure 4 Exam Critical Success Factors

Figure 5 Modelware Systems Training Courses

Figure 6 Data Modelling & Design

What Does the “How to Model Data” Course Curriculum Include?

The session introduces the “How to Model Data” course, modelled on Steve Hoberman’s methodology. This instructional program is designed to be accessible to novices while remaining highly beneficial for experienced professionals due to its systematic, repeatable, and question-based approach. The curriculum is strictly practical, focusing on the construction of relational and dimensional models from inception.

Over three days of training, attendees complete exercises covering every aspect of data modelling, culminating in a fourth-day use-case study in which they must produce conceptual and logical models, as well as an analysis for dimensional modelling. The course incorporates dedicated exam preparation and assigns practical homework exercises to rigorously reinforce the techniques.

Figure 7 How-to Model Data Master Class

What are the Core Categories of the Data Modelling and Design Exam?

The Data Modelling and Design exam is structured into several core categories: introduction to data modelling, activities, tools, best practices, and governance. Approximately 83% of the exam content is concentrated strictly within the introduction, activities, and best practices sections. Attendees are expected to understand the foundational relationship between data models and broader enterprise architecture frameworks, such as the Zachman Framework.

Furthermore, the syllabus requires a deep understanding of essential data modelling components outlined in the DMBOK, including entities, attributes, relationships, domains, and the precise standardisation of supertypes and subtypes. The examination strictly assesses the ability to identify cardinality, optionality, and various relationship structures.

Figure 8 Data Modelling & Design

Figure 9 What you need to study

Figure 10 Introduction, Uses & Methods

What Complex Questions about Denormalisation Methodologies are included in the Exam?

A critical area of examination focuses on the physical implementation of data models, particularly the handling of supertypes and subtypes. Attendees must understand the implications of implementing subtypes as an identity, rolling them down into partitions, or absorbing them into a supertype. Howard emphasises that decisions made by database administrators regarding subtype absorption can introduce potential data quality issues, necessitating strategies to resolve and adequately prevent such errors. Additionally, the exam features complex questions on denormalisation methodologies, requiring attendees to understand different approaches to denormalisation and evaluate their direct impact on physical database implementation.

What is the Difference Between Horizontal and Vertical Data Lineage?

Effective data modelling requires establishing clear parameters, which the methodology refers to as setting “camera settings” to formally determine if a model should be concrete or abstract. Understanding data lineage is also critical for business communication. Howard differentiates between horizontal lineage, which maps the flow of data, and vertical lineage, which systematically connects a database instance, through physical, logical, and conceptual models, directly to the business glossary. Visualising the subject area model is highlighted as an effective technique for logically communicating a data strategy to stakeholders. Models are also utilised to accurately evaluate software package requirements by confirming whether proposed systems can adequately enforce established business and relationship rules.

Figure 11 Activities

What Criteria are Essential for Selecting the Appropriate Schema in Data Modelling?

A foundational step in data modelling is selecting the appropriate schema and defining whether the model objectively represents an “as-is” or “to-be” state. Attendees are systematically tested on various data modelling notations, including Data Vault, anchor modelling, and object-oriented frameworks. Understanding how business requirements translate across different database environments is essential; for instance, a relational model may be used for conceptual and logical designs, while a NoSQL environment (such as MongoDB) requires defining specific JSON document structures for the physical model. The formal curriculum also covers fully communication-oriented modelling (like CaseTalk) and Object-Role Modelling (ORM).

Figure 12 Tools: Notations and Tooling

Why is Diagram Visualisation Crucial in Data Modelling?

The best practices category extensively covers normalisation (ensuring data aligns with “the key, the whole key, and nothing but the key”) and the strategic use of denormalisation for system performance. Beyond structural integrity, diagram visualisation and layout are paramount; if a model is excessively difficult to decode, its communicative value is completely lost. Data architects are formally responsible for establishing diagramming principles that leverage layout, formatting, colour, contrast, flow, and balance to ensure clarity.

For optimal readability, crossing lines should be avoided, and relationships within the model should ideally follow a logical, clockwise reading pattern. Howard also highlights the highly effective technique of abstracting complex physical models to simplify presentations for non-technical audiences.

Figure 13 Best Practices (Improving Your Data Models)

Figure 14 Visualisation – the Four Pillars of Data Visualisation

Figure 15 Purpose – Why this Visualisation

Figure 16 Content – What to Visualise

Figure 17 Structure – How to Visualise

Figure 18 Design Principles

Figure 19 Techniques: Diagramming Clarity

Figure 20 Layout Dos & Don’ts

Figure 21 Best Practices (Improving your Data Models)

What does Data Modelling Governance Encompass?

The final operational component is Data Modelling Governance, which encompasses assessing practice maturity, integrating modelling securely into the Software Development Life Cycle (SDLC), and actively managing data model scorecards. Governance strictly requires capturing critical metadata—such as Personally Identifiable Information (PII) classifications and designated data stewards—directly within the modelling tools.

Regarding the exam itself, attendees are warned about occasionally frustrating question formats, particularly “true or false” questions embedded within multiple-choice distractors, and complex inquiries regarding physical implementation and normalisation trade-offs. Following the formal presentation, attendees are directed to complete a timed 30-question Canvas quiz to assess baseline readiness and identify specific areas requiring further review.