Public Health Data Communication

Proof required

Submit your data cleaning log and dataset summary. The log must document each decision made (what was found, what was done, why). Submit the first 20 rows of the cleaned dataset as a table.

What gets checked

• Data source is a real published public health dataset — not a simulated or fabricated dataset.
• Cleaning log documents the decision for each issue — 'duplicates removed' is not a decision; 'found 3 duplicate patient IDs where data for the same patient appeared twice with different dates; kept the most recent record' is.
• Missing data handling is documented — different strategies for missing data (removal, imputation, flagging) must be justified based on the variable's role.

Common mistakes

• Dataset that is already clean — the milestone requires encountering real data quality issues.
• Cleaning log that lists what was done without explaining why each decision was made.
• Missing data treated as zero rather than as missing — a common error with significant analytical consequences.

Resources

What a verifier looks for

• Ask the student to describe the most significant data quality issue they found and how they handled it.
• Check that the cleaning log documents decisions, not just actions.
• Ask how they handled missing data for the most important variable in their dataset.

You'll sign in first, then come straight back here.

Proof required

Submit all 5 visualisations as a PDF or image file with titles, axis labels, data source, and captions. Each must use one of the five required chart types for its stated purpose.

What gets checked

• Title states the insight — 'Life Expectancy by Local Authority' is a label; 'Life expectancy is lowest in the most deprived decile across all English regions' is an insight title.
• Chart types match the data and purpose — using a pie chart for time series data is a mismatched type.
• All five required types (time trend, geographic, demographic, correlation, distribution) are represented.

Common mistakes

• Visualisations that describe the data without communicating an insight — a chart with a variable name as its title is not a public health communication.
• Chart type mismatch — choosing a chart type because it looks good rather than because it fits the data.
• Missing data source citation on one or more visualisations.

Resources

What a verifier looks for

• Ask the student to explain the insight communicated by their most complex visualisation.
• Check that the chart types are appropriate for each data type — ask why they chose each chart type.
• Verify that all five required chart categories are present.

You'll sign in first, then come straight back here.

Proof required

Submit your 3-page report (with all 5 visualisations embedded), written feedback from a public health professional or academic, and a one-paragraph revision note on the most important change you made based on the feedback.

What gets checked

• Executive summary states the public health message and both recommendations in plain English — not a description of the analysis.
• Language is accessible — if a non-expert could not understand the main message after reading the executive summary, the report has failed its audience.
• Feedback is from a public health professional or academic — not a peer student.

Common mistakes

• Report written in academic register rather than for a lay commissioning audience.
• Recommendations that are too general — 'address health inequalities' is not a recommendation a local authority can act on.
• Feedback from a peer student without public health expertise.

Resources

What a verifier looks for

• Ask a non-expert (not the reviewer) to read the executive summary and state the main message — tests accessibility.
• Check that the recommendations are specific enough to be acted upon by a named decision-maker.
• Verify that the reviewer has public health professional or academic credentials.

You'll sign in first, then come straight back here.