Overview
Data science investigates questions using computational and statistical analysis of data. Work often includes defining a problem, obtaining data, cleaning records, exploring patterns, building models, and communicating results. These stages are interconnected: a modeling decision may reveal that the original question was poorly specified or that the available data cannot answer it responsibly.
Core ideas
Data provenance records where information came from and how it changed. Missing values and inconsistent definitions can introduce bias. Training and test sets help estimate predictive performance, but leakage can make results misleadingly strong. Prediction does not establish causation, and a model's usefulness depends on costs, fairness, stability, and the decisions it is intended to support.
Try it
Download a small public dataset, such as monthly weather summaries, and write a notebook that checks column types, missing values, and units. Produce two clear plots and answer one narrow question. Preserve the raw file and document every transformation. Include a short limitations note explaining why your result might not generalize to another location or period.
Study pathways
Study pathways include analytics, machine learning, data engineering, research support, and decision science. Build foundations in probability, statistics, databases, and software practices rather than collecting tools alone. Many roles need substantial domain expertise. Handle personal information only with an appropriate lawful basis, and avoid presenting automated scores as neutral facts when their consequences depend on human choices.