The interviewer glances at her sheet and asks about a customer who was upset with you. You explain that you always stay calm, listen first and look for a fix. It sounds reasonable, yet it names no customer, no day and nothing you actually did. In 62 videotaped interviews for a real research assistant job in Switzerland, applicants answered past-behavior questions with a story, one specific episode, less than one time in four.¹ Far more often, they summarized what they usually do, and opinions and self-descriptions also outnumbered stories.

The STAR method is built to prevent that. It gives a behavioral answer four parts: the Situation, your Task, the Action you took and the Result. The US Office of Personnel Management (OPM) recommends it to federal hiring teams, with situation or task as one part.² To build a STAR answer:

1. Pick one real episode, your best one
2. Set the scene in two sentences
3. Spend most of the answer on what you did
4. Close with the result and what it shows
5. Practice out loud, then let the follow-up questions come

Behavioral questions come late in a longer process; the targets, weekly plans and interview practice that lead up to them are laid out in our piece on starting a job search from scratch. This page stays with one thing: the shape of a single behavioral answer.

What the STAR method is, and who created it

STAR is a four-part outline for answering behavioral interview questions: describe the situation or task, the action you took, and the result. Development Dimensions International (DDI), a firm that sells interview training, says it introduced STAR in 1974 as part of Targeted Selection, its interviewing program. DDI writes the first two letters as one step, “ST”, because the situation and the task do the same job: they give context. It presents STAR as a way for interviewers to collect objective data.³

That origin claim comes from DDI’s own website. We found no independent record from the 1970s that confirms the date, so treat it as the company’s account.

Myth

The STAR method is a script for telling interviewers what they want to hear.

Fact

By DDI's account, it was built for interviewers, to collect objective evidence. What it asks of you is a true episode with a clear action and result.

STAR began on the other side of the table: selection researchers still describe it as a guide to interviewer prompting.¹ The idea underneath is older than any acronym. OPM calls it behavioral consistency: past behavior is the best predictor of future behavior.⁴ In academic research, Gary Latham and colleagues described the situational interview, which asks what you would do, in 1980.⁵ Tom Janz compared the patterned behavior description interview, which asks what you did, with unstructured interviews in 1982.⁶

1. 1Situationwhere and when, in a sentence or two
2. 2Taskwhat was yours to do, and what was at stake
3. 3Actionwhat you did, step by step: the longest part
4. 4Resultwhat happened, and what you took from it

Knowing where STAR began tells you how to use it. The letters are the interviewer’s checklist: if a part is missing, the interviewer has to ask for it, and not every interviewer will. Give all four parts without being prompted, and the evidence is on the table whether or not the follow-up question comes.

Why STAR fits how structured interviews are scored

A structured interview uses fixed rules for asking questions, observing answers and rating them, so every candidate is judged the same way. OPM notes that interviews with more structure show higher validity, meaning they predict job performance better, and that a panel usually scores each answer separately before discussing differences.⁴

OPM’s 2008 practical guide shows what that looks like from the inside. Each question targets at least one named competency, and interviewers keep planned follow-up questions, called probes, in three groups: the situation, the action and the outcome.⁷ Those are STAR’s parts under different names. A candidate who covers all three has answered the probes before they are asked.

The rating scale is the part most candidates never see. In OPM’s example for interpersonal skills, the levels rise with the difficulty of the situation you handled: from the simplest situations at level 1 to exceptionally difficult ones at level 5. Staying courteous with an employee frustrated by a payroll problem sits in the middle. Defusing an emotionally charged meeting with outside stakeholders sits at the top.⁷

1. Situation and task: set how hard the problem was; in OPM’s example rating scale, the levels rise from the simplest situations (level 1) to exceptionally difficult ones (level 5)
2. Action: answers the action probes, such as how you responded and what you did first
3. Result: answers the outcome probes, such as what happened and whether you would do anything differently

That changes how you choose a story. The situation sets the ceiling for your score, and the action shows whether you reached it. A smooth story about an easy problem can only score so high.

So before an interview, read the job ad for the competencies it names and hold each story you plan to tell up against them: is its situation as hard as the problems this job will bring?

Why most answers to “tell me about a time” questions miss

Most answers miss because candidates describe a pattern instead of an episode. In a 2014 study of real interviews by Adrian Bangerter and colleagues, the most common reply was a “pseudo-story”: a summary of situations that recur, with no single time or place.¹

The applicants were mostly students applying for a research assistant job, and the interviewer deliberately asked no follow-up questions. Stories and pseudo-stories both went with higher hiring recommendations from professional recruiters, while self-descriptions such as “I’m a good communicator” went with lower ones. The study is small and correlational, so it shows what recruiters rewarded, not that storytelling caused the difference. Even the real stories contained more statements about the situation than about the applicant’s actions or results.¹

A randomized experiment published in 2026 points to a practical fix. Koralie Orji and Bangerter asked 274 US adults with customer-facing work experience, recruited online, to answer three past-behavior questions in a recorded mock interview. Wording that asked people to recall a time they performed at their best produced modestly more stories and higher interview ratings. Unlimited preparation time did neither, and the authors suggest preparation may need structure such as STAR to help.⁸

A smooth story about an easy problem can only score so high.

One habit helps you catch pseudo-stories in rehearsal: when you hear yourself say “usually”, “always” or “I tend to”, stop and start again with “one time”.

Building a STAR answer in five steps

Pick one true episode, set the scene briefly, spend most of your time on your own actions, and finish with the result. The steps follow OPM’s probes and the storytelling research above; we found no study that tested STAR coaching on its own in real hiring.

1. Pick one real episode, your best one

OPM advises hiring teams to ask for the most, last, worst, least or best example of a competency, so expect superlatives.² Choose the episode where you did the job best, not a typical week. In the 2026 experiment, simply asking people for their best example raised their ratings a little.⁸ You can ask yourself that question before the interviewer does.

2. Set the scene in two sentences

Say where you were, what went wrong or what was needed, and what was yours to handle. Stop there. The situation matters because it sets the level on the rating scale, but it is also where answers sprawl, and every sentence spent on context is one not spent on what you did.

3. Spend most of the answer on what you did

Describe your own steps in order: what you noticed, what you decided and what you did first. OPM’s action probes ask exactly this, including what the first thing you did was.⁷ The panel is rating you, not your team, so say “I” for your part and credit colleagues in a phrase.

4. Close with the result and what it shows

Give the outcome, with a number if you have an honest one, and one line on what you took from it. OPM’s outcome probes include whether you would do anything differently, so a candid line on what you would change belongs in the answer.⁷

5. Practice out loud, then let the follow-up questions come

Tell each story aloud a few times before the interview, then stop at the result and let the interviewer probe. Keep it true: in an experiment within a 2007 series of studies by Julia Levashina and Michael Campion, past-behavior questions were more resistant to faking than situational ones.⁹

Notice what the example leaves out: no opinion about customer service and no claim to be calm under pressure. The situation takes two sentences, the action takes the most room, and the result ends with a lesson the interviewer can probe.

WiserHours

Test each story against this list

• It is one episode with a time and place, not a pattern
• Its situation is as hard as the problems this job will bring
• The situation and task take two sentences at most
• Most of it describes what I did, in order
• It ends with a result and what I learned
• Every detail is true and I can answer follow-up questions on it

Find this checklist online:
https://wiserhours.com/job-search/star-method/#checklist-star-method-story-check

Past behavior or hypotheticals: what a 2002 meta-analysis found

Questions about what you did and questions about what you would do both predicted job performance well in a 2002 meta-analysismeta-analysis: A study that combines the results of earlier studies on the same question into one overall estimate. Pooling makes the estimate more precise, but it cannot repair the studies it pools: a meta-analysis of surveys is still survey evidence.Full entry in the glossary by Paul Taylor and Bruce Small. Past-behavior questions came out ahead, but mainly when answers were scored against rating scales that describe what each level looks like.¹⁰

The study

Moderate evidence

Taylor and Small (2002): questions about what you did versus what you would do

The authors pooled studies that checked interview scores against everyday job performance as rated by supervisors or peers. Past-behavior questions scored against descriptively anchored rating scales reached a corrected validity of .63, against .47 for situational questions scored the same way. Past-behavior questions without such scales fell back to .47.¹⁰

Those figures are correlations between interview scores and performance ratings, corrected upward for measurement problems, so trust the ranking more than the size. A 2022 re-analysis by Paul Sackett and colleagues concluded that range-restriction corrections, one of the two adjustments in the 2002 study, had substantially overstated how well many selection methods predict performance, although structured interviews still ranked first; it did not re-test the past-behavior versus situational comparison.¹¹

The comparison is also contested. Earlier meta-analyses had favored situational questions, and Taylor and Small argue that the main reason is that early past-behavior studies rarely used rating scales.¹⁰ OPM’s guidance simply calls both formats effective.⁴

The practical point survives the argument. In OPM’s guide, each level of a rating scale comes with examples of behavior, and interviewers use them as a guide when they rate an answer.⁷ Picture two candidates asked about a disagreement at work. One says she is good at bringing people together; the other describes the budget meeting where two managers kept talking past each other, and the one-page compromise she drafted that both signed. Only the second answer contains anything a rater can match to an example.

If you get a “what would you do” question instead, the same parts still help. OPM’s guide gives interviewers situational probes in the same three groups: why the situation arose and what matters most in it, what you would do first, and how your action would be received.⁷ Answer with a plan in that order, and if you have faced something similar, say so briefly and explain what you did then.

How much evidence sits behind each tip

The case for STAR is indirect: it comes from the way structured interviews are designed and rated, plus a few studies of how candidates answer.

Read the rows as a guide to how far to lean on each tip. Answering with one specific, true episode is supported by how recruiters rated answers in real interviews, though that study was correlational. Choosing a hard situation follows from how one government rating scale is written, so treat it as a sensible bet rather than a tested finding.

The bottom line

A behavioral question asks for evidence, and the STAR method is simply the shape that evidence takes on an interviewer’s rating sheet. Pick one true episode whose difficulty matches the job, keep the situation short, spend your words on what you did, and end with the result. In the one real-interview study we found, recruiters gave higher recommendations to answers built on events and lower ones to answers that described the candidate.