Foundations: Data, Data, Everywhere
Core concepts
Data is a collection of facts; information is data that has been processed, organized, and given context so it can drive a decision. Candidates must know the six-phase data analysis lifecycle - Ask, Prepare, Process, Analyze, Share, Act - and what happens in each. Ask defines the business problem, Prepare builds a trustworthy dataset, Process cleans it, Analyze finds patterns, Share communicates results, and Act applies them.
Know the data-type taxonomy. Quantitative (numeric) data splits into discrete (countable whole units) and continuous (measurable, can be fractional); qualitative data is categorical. The four measurement scales are nominal (labels, no order), ordinal (ordered but with uneven gaps, e.g., satisfaction ratings), interval (equal gaps, no true zero, e.g., temperature in Celsius), and ratio (equal gaps plus a meaningful zero, e.g., revenue or height). Also distinguish structured data (rows and columns) from unstructured data (text, images, audio) and semi-structured formats such as JSON and XML.
Other terms to recognize: big data (volume, velocity, variety), root cause analysis, gap analysis, data-driven versus data-inspired decisions, and the division of labor among data analyst, data scientist, and data engineer.
Traps
Questions love interval versus ratio: a zero in Celsius does not mean no temperature, so ratios are meaningless there. Another common trap asks which lifecycle phase an activity belongs to - writing a measurement plan is Ask, joining tables is Prepare/Process, building a dashboard is Share.
Rule of thumb
Ask whether zero can mean none. If yes, the scale is ratio; if no, it is interval. A 20-degree day is not twice as warm as a 10-degree day, but 20 dollars is twice 10 dollars.