DA0-002 — CompTIA Data+ V2 Cheat Sheet

For best results:

• Use the tables to test “when would I choose this?” rather than memorizing definitions only.
• Practice reading scenario clues: business objective, data source, data type, quality problem, stakeholder, and reporting need.
• Pair this page with timed DA0-002 practice questions to confirm that you can apply concepts under exam conditions.

This page is IT Mastery exam-prep support. It is designed to help you connect key ideas to IT Mastery practice, including original practice questions, targeted topic drills, a full question bank, and answer explanations that clarify why each choice is right or wrong.

1. Scan the high-yield tables first. Mark topics that feel weak.
2. Work topic drills immediately after reviewing a section. Do not wait until you feel “done.”
3. Use mistakes diagnostically. A missed question usually points to a decision rule, vocabulary distinction, or scenario clue.
4. Review explanations, not just answer keys. For CompTIA Data+ V2 (DA0-002), understanding why distractors are wrong is often as important as recognizing the right option.
5. Finish with mixed mock exams. Topic drills build accuracy; mixed exams build exam-day judgment.

Measurement Scales

Exam trap: Do not average nominal labels. Be cautious averaging ordinal ratings; it is common in business reporting, but the scale distance may not be truly equal.

OLTP vs OLAP

Data Modeling Essentials

Common Storage and Processing Concepts

Schema, Grain, and Keys

Decision rule: Before aggregating or joining data, identify the grain. Many analysis errors come from joining tables at different grains and accidentally duplicating facts.

ETL vs ELT

Common Transformation Tasks

Core SQL Patterns

-- Aggregation with group filter
SELECT
    region,
    COUNT(*) AS order_count,
    SUM(order_amount) AS total_sales,
    AVG(order_amount) AS avg_order_value
FROM orders
WHERE order_date >= '2026-01-01'
GROUP BY region
HAVING SUM(order_amount) > 100000
ORDER BY total_sales DESC;

-- Left join to keep all customers, even those without orders
SELECT
    c.customer_id,
    c.customer_name,
    COUNT(o.order_id) AS order_count
FROM customers c
LEFT JOIN orders o
    ON c.customer_id = o.customer_id
GROUP BY c.customer_id, c.customer_name;

-- Window function: rank rows without collapsing detail
SELECT
    customer_id,
    order_id,
    order_amount,
    RANK() OVER (
        PARTITION BY customer_id
        ORDER BY order_amount DESC
    ) AS order_rank
FROM orders;

-- CASE expression for business categories
SELECT
    customer_id,
    total_spend,
    CASE
        WHEN total_spend >= 10000 THEN 'High'
        WHEN total_spend >= 1000 THEN 'Medium'
        ELSE 'Low'
    END AS spend_segment
FROM customer_summary;

Join Types

SQL Exam Traps

Data Profiling Checklist

• Count rows and compare to expected volume.
• Review data types and unexpected type coercion.
• Count NULLs by column and by key business segment.
• Identify duplicate keys or suspicious near-duplicates.
• Check minimum, maximum, mean, median, and outliers for numeric fields.
• Validate categorical values against allowed domains.
• Verify date ranges, future dates, and impossible timestamps.
• Confirm referential integrity across joined tables.
• Compare aggregates to trusted control totals.
• Document assumptions, exclusions, and known limitations.

Cleaning and Remediation Choices

Quality Dimensions

Quality Controls

Common trap: Data quality is not only a technical issue. It depends on business definitions, process design, ownership, and controls.

Measures of Center and Spread

High-Yield Formulas

Use formulas only after confirming the business definition. For example, “growth” may mean year-over-year, month-over-month, compound growth, absolute change, or percentage change.

Distribution Concepts

Common Test Selection

Exam trap: Statistical significance does not prove practical significance, business value, or causation.

Bias and Validity Traps

Method Selection

Model Evaluation Metrics

Visualization Design Principles

Dashboard and Report Types

KPI and Metric Review Checklist

• Is the metric tied to a business objective?
• Is the numerator and denominator clearly defined?
• Is the grain clear?
• Are filters and exclusions documented?
• Is the time period consistent?
• Is there a target, benchmark, or baseline?
• Could the metric be gamed?
• Are leading and lagging indicators balanced?
• Are related metrics needed to avoid misinterpretation?

KPI Quality Checklist

A good KPI is:

• Aligned to a business objective.
• Defined clearly enough that two analysts calculate it the same way.
• Measurable from available or obtainable data.
• Timely for the decision being made.
• Actionable by the audience.
• Contextualized with target, baseline, segment, or trend.

KPI, Metric, Dimension, and Measure

Metric Trap Examples

Chart Selection

Visualization Principles

Dashboard Design

Common trap: A dashboard is not a data dump. It should support decisions, monitoring, and investigation.

Communicating Findings

Exam trap: A technically correct analysis can still fail if it does not answer the stakeholder’s actual question.

Governance Artifacts

Sensitive Data Handling

Access Control Distinctions

Governance Concepts

Protection Techniques

Governance Decision Rules

Validation Before Publishing

• Reconcile totals against trusted source reports.
• Confirm joins do not change expected row counts unexpectedly.
• Test filters, date ranges, and parameter defaults.
• Review outliers and decide whether they are errors or real events.
• Validate metric definitions with stakeholders.
• Check access permissions and sensitive fields.
• Include source, refresh date, owner, and definitions.
• Save query/report version and document assumptions.

Final Cheat Sheet Checklist

Before your next DA0-002 practice session, confirm that you can:

• Explain structured, semi-structured, and unstructured data.
• Distinguish nominal, ordinal, interval, and ratio data.
• Identify grain before joining or aggregating.
• Choose the right join type for a scenario.
• Explain WHERE vs HAVING.
• Handle NULLs, duplicates, missing values, and outliers appropriately.
• Select mean, median, mode, standard deviation, percentile, or IQR based on data shape.
• Recognize correlation without overstating causation.
• Identify bias in sampling and surveys.
• Match charts to analytical questions.
• Design dashboards around audience and decision needs.
• Apply data quality dimensions.
• Use governance concepts such as classification, lineage, stewardship, and least privilege.
• Communicate findings with assumptions, context, and limitations.

Levels of Measurement

Common trap: If a value is numeric-looking but represents a label, it is not automatically quantitative. ZIP codes, account IDs, product codes, and employee numbers are identifiers, not measures.

Workflow Traps

Batch vs Streaming

Common trap: “Real-time” is not automatically better. If the decision is monthly, a well-controlled batch process may be more appropriate than a fragile streaming pipeline.

Cleaning Techniques

Missing Data Decision Table

Aggregation Traps

WHERE vs HAVING

NULL Behavior

Critical rule: Correlation does not prove causation. A relationship may be caused by a third variable, coincidence, reverse causality, or selection effects.

Bias and Sampling

Hypothesis and Inference Basics

Common trap: A statistically significant result may be too small to matter operationally. A non-significant result may still be important if the sample is too small or noisy.

Time-Series Review

Communicating Results

Likely correct choices include dashboard design, KPI alignment, aggregation, and concise visualization.

Scenario 2: Dirty Customer Dataset

• Profile duplicates, nulls, invalid values, and inconsistent formats.
• Define deduplication logic.
• Preserve lineage and document cleaning rules.
• Avoid deleting records without a business rule.
• Validate against authoritative sources where possible.

Likely correct choices involve data quality dimensions, cleansing, validation, and stewardship.

Scenario 3: Unexpected Sales Spike

• Check source system changes, promotions, seasonality, outliers, and data pipeline issues.
• Segment by product, region, channel, and customer type.
• Compare against historical baselines.
• Avoid assuming causation.

Likely correct choices involve diagnostic analysis, segmentation, trend comparison, and validation.

Scenario 4: Sensitive Employee Data

• Classify data.
• Restrict access by role.
• Mask or tokenize when full values are unnecessary.
• Log access.
• Use only the minimum data needed.

Likely correct choices involve governance, security, least privilege, masking, and privacy-aware handling.