Example · Not a real student
Data analysis report
Fictional student · Example case: Check a library’s borrowing claim
Before you read your report
This report covers only this 60-minute work sample. The level describes the work you submitted, not you as a person. A person reviews every attempt and makes any selection decision.
Skill level on this work sample
Developing
You can answer some questions with data. Build accuracy on harder queries and data checks.
Overall score
56/100
You use the data carefully and say what is still uncertain. Your SQL needs more accuracy on date ranges, and your message should lead with the key number.
Skills shown in this work
Open a skill to see what it means, what you did and what to practise next.
SQL50/100Can you write queries that return exactly what was asked?
Based on your checked answers and written work. How sure we are: high (your answers give clear support for this feedback)
Your queries return the right columns, and you count each record once. One query used the wrong date range, so its total was off.
“SELECT COUNT(DISTINCT loan_id) AS loans FROM loans WHERE due_date < '2026-08-01';”
Try this: Write the date range in words before writing the filter, then check the first and last rows of your result.
Data quality63/100Do you notice rows that should not be counted, and fix the numbers?
Checked answers: 50/100. Written work: 75/100. How sure we are: medium (based on only a few answers)
You notice the repeated record IDs and deal with them before counting. You do not mention other rows that might not belong in the total.
“Some record IDs appear twice, so I counted each ID once before comparing the two periods.”
Try this: Before any total, list the kinds of rows that should not count and how you found them.
Analysis50/100Do your conclusions follow from correct numbers, with their limits stated?
Based on your checked answers and written work. How sure we are: medium (based on only a few answers)
You compare rates instead of raw totals, which is fair. You do not say which records you included in each period.
“On-time returns went up after the email, but the second period is shorter, so part of the change may be chance.”
Try this: State which records you included in each period and why.
Business judgement75/100Would your conclusion and next step help the manager decide?
Based on your written answers. How sure we are: medium (based on only a few answers)
Your next step is cautious and practical: keep the email and collect a fairer comparison first.
“Keep the email for one more month and compare the same length of time before deciding.”
Try this: Say what result next month would make you stop the email.
Communication50/100Can a busy, non-technical manager understand your message and act on it?
Based on your written answers. How sure we are: low (we need more detail from your answers)
Your message is short and honest about what is still uncertain. A busy reader would still need the key number in the first line.
“The email may be helping. We need one more month of data to be sure.”
Try this: Put the main number first, then one sentence on what it means and one on what to do.
Checked answers
Your SQL was run again on your dataset, and your numbers and choices were compared with the data.
| Question | Result | What it means |
|---|---|---|
| Understand the tables · wording hidden | Correct | Your choice matched the data. |
| Write SQL · wording hidden | Not correct | Your query ran, but its date range left out some rows. |
| Check data quality · wording hidden | Correct | Your total matched the data after removing the problem rows. |
| Test a claim · wording hidden | Did not run | Your query did not run. Check the column names against the tables. |
| Recommend a next step · wording hidden | Not answered | No answer was saved before the stage locked. |
Strengths to keep
- You count each record once before comparing periods.
- You compare rates, not raw totals.
- You say clearly what is still uncertain.
Your next steps
- Practise date filters on a small public dataset and check the first and last rows each time.
- Before any total, write down which rows should not count.
- Rewrite your message with the key number first.