Data Analytics Fundamentals · Course lab · about 180 minutes · 7 tasks · marked out of 100, pass at 60

One question, answered honestly, on one page

The situation

An academy owner says: "Check how our evening batch is doing." That is not a question yet. You will turn it into one, build a register that could answer it, clean it, analyse it, chart it honestly, and deliver a one-page finding with its limits stated — the complete cycle a working analyst runs every week.

What you'll be able to show

  • Turn a vague request into a question with an explicit comparison
  • Clean a register and write the cleaning note that makes the numbers trustworthy
  • Answer with pivots and formulas, and cross-check one number by hand
  • Chart with a finding as the title, and show how the same data could be made to lie

What you need

  • Excel or Google Sheets
  • A 60-row register you invent to be plausible (students, city, batch, month, fee due, fee paid, attendance)
  • One page for the finding

Tasks

  1. 1The question
    Write the decision behind "check the evening batch", the specific question, the comparison you will make (evening vs morning? this term vs last?), and one way the obvious comparison could mislead.
    A correct result: Four sentences: decision, question, comparison, trap.
  2. 2The register, dirty
    Build 60 rows with the columns above. Then dirty it as real data is: three spellings of one city, two duplicate rows, five fees written as text with Rs, four blank fees, and one attendance of 140%.
    A correct result: 60 rows with the fifteen defects present and listed.
  3. 3Clean it, and say so
    On a copy, fix each class of defect. Write a cleaning note: what was wrong, how many rows, what you did (including what you did with the blank fees and the 140%). Verify SUM(fee paid) against a hand-checked figure.
    A correct result: A note with counts per defect class, and a SUM that matches your hand check.
  4. 4Answer it
    With pivots and formulas: enrolments and fee collected per batch; attendance average per batch; fee due vs paid per batch. Cross-check one pivot number by filtering the rows by hand.
    A correct result: Three answers, one of them hand-verified with the filter shown.
  5. 5Three honest charts
    A sorted bar of fee collected per batch with a zero axis; a line of admissions per month; a scatter of fee paid vs attendance. Each gets a title that states the finding, not the topic.
    A correct result: Three charts whose titles are sentences.
  6. 6Make it lie, then label it
    Copy the bar chart and truncate its axis so a small difference looks huge. Put the two side by side and write one line on how the same data now lies.
    A correct result: Two charts of the same data and one sentence.
  7. 7The one-pager
    One page: the finding as a sentence with a number, two charts with finding-titles, a recommendation labelled as a recommendation, three limits of the analysis, and the cleaning note attached.
    A correct result: One page a manager reads in three minutes and can act on — or push back on, because the limits are there.

What to hand in

The workbook (original and clean sheets, pivots, charts), the cleaning note, and the one-page finding.

How it is marked

CriterionPoints
Question has decision, comparison and a named trap 15
Cleaning is complete and documented with counts 20
Answers are correct and one is hand-verified 20
Charts are honest with finding-titles 15
The lying chart is shown and explained 10
One-pager has finding, recommendation and three limits 20
Total · pass at 60 100

Hand in your lab

Create a free BvLogic ID to hand in your lab, get it marked, and have it on your certificate.

Create your BvLogic ID

Already have one? Sign in and this course will be added to it.