In a data-reading interview exercise, explain what the numbers support and what they leave unknown. Check the units, denominator and comparison before making a recommendation. A confident statement is not useful if it answers a different question from the one the data can address.

The interview questions and answers below use an original fictional help-desk fixture. These are practice numbers, not research findings or a real organization's results.

Read the fixture

Week | Requests received | Requests resolved
A    | 80                | 72
B    | 100               | 85

Not provided: starting backlog, case complexity,
staffing, reopened cases or time spent per request.

Read each column literally. The fixture does not say that the resolved requests were all received in the same week. That missing relationship matters when interpreting a percentage.

Question: did resolved volume increase?

Sample answer: “Yes. Resolved volume increased from 72 to 85, which is 13 additional resolutions. That is about an 18.1% increase relative to 72. I would not yet conclude that individual productivity improved, because staffing and case complexity are not provided.”

The calculation is 13 divided by 72. State the base when using a percentage. Thirteen more cases is an absolute change, while 18.1% is a relative change.

Question: did the resolution rate improve?

Sample answer: “Resolved divided by received is 90% for A and 85% for B. That ratio fell by five percentage points. However, it is not necessarily the completion rate of each week's incoming cases because we do not know which cases were resolved.”

This answer gives the arithmetic and protects its meaning. A ratio can be calculated correctly while being given an unsupported label.

Question: can we calculate the backlog?

Sample answer: “Not its total size without the starting backlog and movement rules. If we assume no reopening, deletion or other adjustments, net unresolved volume increased by eight in A and fifteen in B. Those are changes under assumptions, not complete backlog totals.”

Keep the assumption visible. If cases can reopen or move between queues, the simple subtraction may not describe the whole process.

Question: what would you ask next?

  • How are received and resolved defined?
  • Are the periods comparable?
  • What was the starting backlog?
  • Did staffing or case mix change?
  • What decision is this analysis meant to support?

Choose questions that affect the decision. Do not request every possible field without explaining how it would help.

How would you recommend an action?

State a provisional interpretation and the next check. For example, rising incoming work and a larger unresolved difference justify investigating capacity and case mix. They do not prove that staff are performing poorly.

When practicing, ask someone to change one number or add a missing condition. Recalculate and revise the conclusion. That tests your reasoning more effectively than memorizing the sample answers.

Before finishing, separate observations, calculations, assumptions and recommendations. A good data-reading answer lets the listener inspect each step instead of accepting an unsupported story about the numbers.

References and next practice

careerservices.fas.harvard.edu — career

Browse interview topics · Practise a question in Interview Challenge · How to Prepare for an Interview