CompTIA Data+

SQL is the analyst’s main tool for getting at data. You select, filter, join and summarize tables, and write queries that another person can read and check.

• SELECT, WHERE and ORDER BY
• Joins across tables
• Grouping and aggregate functions
• Subqueries and common table expressions
• Window functions for running totals and ranks
• Writing readable, tested queries

You build A set of queries answering six business questions from a sample sales database.

Real analysis starts by gathering data from several places. You pull from files, APIs, web pages and surveys, combine sources on shared keys, and recognize the risks of each method.

• Exports, files and common formats
• APIs and web data collection
• Surveys and sampling for collection
• Merging and appending datasets
• Extract, transform and load pipelines
• Documenting where data came from

You build A combined dataset built from three sources, with a record of every join.

Before you analyze, look. You profile a dataset to find its shape, its gaps and its oddities, and decide what is a real signal and what is a defect in the data.

• Summarizing columns and distributions
• Finding missing values
• Detecting duplicates
• Spotting outliers and impossible values
• Checking data types and ranges
• Writing up what you found

You build A data profile report for one messy dataset, listing every problem found.

Bad data produces confident, wrong answers. You fix missing values, duplicates and inconsistent formats with stated rules, reshape tables for analysis, and build new fields from existing ones.

• Handling missing data: remove, fill or flag
• Standardizing text, dates and units
• Removing duplicates safely
• Reshaping: pivoting and unpivoting
• Deriving new columns and categories
• Keeping the raw data untouched and the steps repeatable

You build A cleaning script with a log, turning a messy file into an analysis-ready table.

Statistics turn a table into understanding. You summarize centre, spread and shape, compare groups, measure relationships and recognize the common distributions that data follows.

• Mean, median and mode, and when each misleads
• Variance, standard deviation and percentiles
• Common distributions and skew
• Correlation and what it does not show
• Comparing groups and segments
• Sampling and sample size

You build A descriptive summary of a dataset that compares three customer segments.

A difference in a chart may be chance. You use hypothesis tests, confidence intervals and simple regression to judge evidence, and learn how bias and wrong assumptions mislead analysts.

• Hypotheses, p-values and what they mean
• Confidence intervals
• t-tests and chi-square tests
• Simple linear regression
• Correlation and causation
• Bias, confounders and misuse of results

You build A short analysis that tests a business claim and states its limits.

Technique is useless if it answers the wrong question. You translate a request into measures and methods, track trends and key indicators, and find and fix problems in an analysis that does not look right.

• Clarifying a request into a measurable question
• Choosing metrics and indicators
• Trends, seasonality and time series basics
• Segmentation and cohort comparison
• Debugging an analysis that gives odd results
• Stating assumptions and limitations

You build An analysis plan and written findings for one stakeholder question.

A chart is an argument. You choose the right visual for each question, use color, labels and scale honestly, and spot the common ways a chart misleads.

• Matching chart types to questions
• Maps and geographic data
• Color, labels and accessible design
• Axes, scales and misleading charts
• Annotating a chart to carry the point
• Tables and when not to chart

You build A set of five charts that each answer one question, each with a one-line takeaway.

Dashboards turn analysis into a routine tool. You design for a specific audience, add filters and drill-down, connect to live sources and keep a dashboard fast and readable.

• Designing for an audience and a decision
• Layout, hierarchy and restraint
• Filters, parameters and drill-down
• Connecting to data and refreshing it
• Performance and maintenance
• Sharing and access

You build An interactive dashboard that tracks five measures for a team.

A report is only useful if it can be trusted. You write analysis for different readers, reconcile numbers against sources, test your reports for errors and version them as they change.

• Static, recurring and ad hoc reports
• Writing a data story for a reader
• Reconciling totals against source systems
• Testing calculations and filters
• Version control and distribution
• Reviewing a report with its audience

You build A validated monthly report with a reconciliation sheet and change log.

Analysts handle data other people own. You document lineage and definitions, apply retention, privacy and access rules to personal data, and measure data quality so users know what to trust.

• Metadata, catalogs and data dictionaries
• Data lineage and ownership
• Retention, privacy and personal data
• Compliance requirements for analysts
• Access control and sharing rules
• Quality dimensions and quality checks
• Master and reference data

You build A data dictionary, lineage note and quality checklist for one dataset.

One project covers the full cycle. You take a business question, source and prepare data, analyze it, present it visually and document how it was governed, then defend it to a skeptical audience.

• Framing the question and the success measure
• Sourcing, cleaning and exploring the data
• Analyzing and testing the findings
• Visualizing and reporting
• Documenting governance and limits
• Presenting and answering challenges

You build A complete analysis with queries, cleaning log, dashboard and written recommendation.