STAR interview answers for internships: two complete examples

What each part needs to do

Situation: where the problem happened. Task: what you were responsible for. Action: what you did and why. Result: what changed. You do not need to announce those four labels when speaking.

Berkeley recommends practicing relevant experience before interviews. These two answers are our own fictional practice examples. Use their structure, not their invented history.

Sources[1]

Example 1: catching an error

Question: 'Tell me about a time you caught a mistake before a deadline.'

'In a statistics class, our team was summarizing a survey, and I was responsible for checking the spreadsheet before we wrote the findings. I noticed repeated rows. Rather than deleting every repeated respondent ID, I checked the raw files and import history. One source file had been imported twice. I kept the original, removed the extra imported copy from our working data and asked a teammate to check the record count. We corrected the summary before submitting it. Afterward, I added an import log so we could catch the same problem earlier.'

Why that answer is more than 'I am detail-oriented'

The answer identifies one decision: check the source before deleting records. Repeated IDs might represent legitimate repeat responses; a duplicated import is a different problem. The student explains how they distinguished the two and checked the correction.

It also keeps the result modest. The team corrected its class submission. The answer does not need invented savings, an exceptional grade or a claim that the student rescued the project.

Be ready for the follow-up: 'How did you know which file had been imported twice?' In your own story, you should be able to explain the check you actually performed.

Example 2: a mistake you actually own

Question: 'Tell me about a time you missed a deadline.'

'For a club workshop, I was responsible for getting the room request submitted. I assumed another committee member had sent the form and did not confirm it. When I checked the day before the event, our preferred room was unavailable. I told the organizer that I had missed the check, found two available rooms and contacted the registered attendees after the organizer chose one. We held the workshop, but the move created extra work and some confusion. For later events, I put a named owner and a confirmation link beside each booking task.'

The correction does not erase the mistake. That is why this answer fits the question better than a story about working too hard or caring too much.

Cut the parts that do not change the answer

Remove a long history of the club, unnecessary software names and adjectives such as 'exceptional'. Keep the constraint, the choice and the result. Practice from prompts rather than a memorized paragraph so you can answer an interruption naturally.

When the question changes, select the relevant part of a true story. Do not convert a teammate's action into your own or change a successful outcome into a failure to make a prepared answer fit.

Replace ‘50% more accurate’ with what actually happened

Fictional source notes: a student found four repeated order IDs in a 40-row import, kept the original file, checked the import history with a teammate and submitted a corrected class report. No accuracy study was performed.

Repair the unsupported answer
Weak claim	Use instead
The spreadsheet was messy	Four order IDs appeared more than once in a 40-row import
I used advanced analytics	I compared the IDs and checked the import history
I improved accuracy by 50%	We corrected the confirmed import problem before submitting the report
Everyone loved the result	The actual feedback, if any; otherwise leave it out

Your turn

Tell the story in a short answer. Then explain why four repeated IDs do not necessarily mean four rows should be deleted.

See a worked answer

For a class report, I was checking a 40-row order import when I found four repeated order IDs. I kept the original file and checked the import history with a teammate before changing the data. We corrected the import problem and submitted the revised report. The useful lesson was to investigate repeated IDs before relying on the summary.

Four repeated IDs could involve more than four extra rows, or legitimate line items. You need the row pattern and the table's intended unit of observation before deciding what to remove.

For your own application

For each result in your own answer, ask what was measured and what you personally changed. A corrected report is a result; it does not need an invented improvement percentage.

01What each part needs to do

02Example 1: catching an error

03Why that answer is more than 'I am detail-oriented'

04Example 2: a mistake you actually own

05Cut the parts that do not change the answer