Working with data is a position of trust. Employers increasingly screen for data ethics and governance — knowing how to use data responsibly and lawfully is now a core professional skill, not an afterthought.
Data literacy
Data literacy is the ability to read, understand, question and communicate with data. It means knowing what a metric really measures, spotting when a chart misleads, and asking "where did this come from and can we trust it?" It's the skill that turns numbers into good decisions.
Personal data and GDPR
Much business data is about people. Under GDPR (and similar laws), personal data (PII) — anything that can identify someone — must be handled lawfully:
- Lawful basis — you need a valid reason to process it (consent, contract, legitimate interest…).
- Purpose limitation & minimisation — collect only what you need, for a stated purpose.
- Storage limitation — don't keep it longer than necessary.
- Security — protect it (access controls, encryption).
- Individual rights — people can access or request deletion of their data. Analysts should anonymise or pseudonymise personal data wherever possible.
Data governance
Governance is the framework of policies, ownership and standards that keeps data accurate, secure and used properly — data owners, access controls, definitions ("what counts as an active customer?") and audit trails. Consistent definitions prevent the classic "your number doesn't match mine" argument.
Avoiding bias and misleading analysis
Ethical analysis means not fooling yourself or others:
- Sampling bias — is your data representative, or skewed?
- Confirmation bias — don't cherry-pick data that fits the answer you wanted.
- Misleading visuals — truncated axes, cherry-picked timeframes and dodgy scales distort the truth.
- Correlation ≠ causation — two things moving together doesn't mean one causes the other. Present data honestly, including uncertainty and caveats.
Put it to work
For a dataset you use, identify any personal data and how it should be protected, and check one chart you've seen recently for a misleading axis or timeframe.