Tour · Data Analytics & Visualization
For the K–12 Data Analytics & Visualization Specialist role
Five stops, a few minutes each. Each one opens a tool in its own tab, says what to try and what to notice, and quotes the line of the job description it answers.
Everything here runs on a made-up school – the tools' real code, nothing sent anywhere. For what the live tools have done with real classes, see the AI Lessons and moving off spreadsheets case studies.
-
A briefing a leader can read in a minute
Lion Tracks (in development) · Briefing, viewed as Leadership · whole school
Try
- Read the headline and the four figures under it.
- Set Division to Middle and Subject to Math; then change "Viewing as" to a teacher.
- Turn on Show the evidence and pick a sentence; open How ratings are worked out.
Notice
- The headline says what changed, why, and where to act first – in plain words.
- Every figure has its comparison and its number of students; retest-flagged tests are left out, and it says so.
- A teacher sees only their class: the same page, the right access.
Answers:
Lion Tracksexplain patterns, recommendations, assumptions, and limitations to nontechnical audiences
·disaggregate data appropriately
·define measures, filters, access
-
Three messy exports, one clean picture
Lion Tracks · Bring in the files, then Data & method
Try
- Open Bring in the files: the MAP Growth, CAT4 and DIBELS exports as they arrive.
- Press Reconcile, read what was matched, cleaned, merged or left for a person; look at the same join in SQL.
- Open Data & method and its list of measures.
Notice
- Leading zeros, "Last, First", duplicates and retests are cleaned by rules you can read – and anything uncertain is left for a person, not guessed.
- Every measure has a definition, a calculation, a source, a refresh schedule and who sees it.
Answers:
Lion Tracksextract, transform, reconcile, and combine data from master data sources using spreadsheets, SQL
·define measures, filters, access, calculations, data sources, and refresh schedules
·reduce duplicated and manual reporting
-
One school, four dashboards
AI Lessons · the principal's dashboard (then switch "View as" in the sample bar)
Try
- Read the four tiles, then Show them under "Need a nudge".
- Switch Lesson overview and Attainment by strand between Chart and Table.
- Filter by a class or a teacher; then view as the teacher and as the AI lead.
Notice
- A number always leads to a list of people to act on.
- Every chart has a table; color is never the only signal.
- Each role sees only what it should – the rules are enforced by the server, not the page.
Answers:
AI Lessonsdesign, develop, and maintain dashboards and reports
·appropriate access, confidentiality
·accessible visual hierarchy
-
Evaluating an AI pilot, with the evidence
AI Pilot Evaluation (in development) · a made-up MagicSchool pilot
Try
- Read the Report and its recommendation.
- Open Survey: the questions themselves, then before and after.
- Open Comments and change a theme – watch the agreement figure move.
Notice
- Groups smaller than five are suppressed, and it says why.
- AI suggested the comment themes; a person checked each one, and the agreement (with Cohen's kappa) is shown, not assumed.
- Limitations come before conclusions.
Answers:
AI Pilot Evaluationprogram evaluation, surveys, and strategic initiatives through appropriate measures, quantitative and qualitative analysis
·Validate AI-assisted work against source data
·check for bias
-
Data a teacher can act on
MAP Dashboard · Spoken Briefing and the questions along the top
Try
- Step through the Spoken Briefing (or press Play).
- Ask Who needs support? and How do I group them?
- Open the seating planner under More analysis.
Notice
- The questions are teachers' questions, not the export's column names.
- It ends in actions – groups, seats, goals – not just charts.
Answers:
MAP Dashboardexplain findings and limitations to nontechnical audiences
·Build staff data literacy
·turns data into clear, useful, and actionable information
If you have more time
The same role, from other angles.
- Data AnonymizerNames swapped for codes in the browser before a spreadsheet goes to an AI tool –
protect sensitive information
. - Five traps when reading assessment dataA one-page staff guide –
common analytical pitfalls
. - Moving three tools off spreadsheetsA case study:
data migration, testing, and onboarding
.
Opens in its own tab