Halo Effect
Halo Effect
What is the halo effect?
- A cognitive bias in which one trait colors the entire evaluation
- A classic evaluation error in hiring and performance reviews
- Blocked by structure (independent item-by-item evaluation), not by resolutions
What the halo effect really changes is the resolution of an evaluation. Even if a scorecard has five items, once one strong impression takes hold first, the other four boxes turn into places where that impression gets copied. You end up with five scores but only one judgment, and the bigger problem is that the evaluator believes they weighed each item carefully. That is why this bias can’t be fixed by correcting scores after the fact; it shrinks only when you design the order in which scores are formed.
It helps to draw a line between this and similar rating errors. The horn effect is the same mechanism in the opposite direction; the recency effect is a time bias in which the most recent period colors the whole review period; and leniency bias means rating everyone generously, so the ranking among candidates stays intact. The halo effect, by contrast, flips the ranking altogether. The conditions that strengthen it are fairly clear too. It works harder when the criteria are abstract, such as attitude, potential, or personality; when the evaluator has observed the person only briefly; and when one person has to evaluate many people in a day under a tight deadline. Even the same interviewer leans on the halo more in the fifth interview of the afternoon than in the first of the morning.
The most common misconception in practice is that 'adopting a structured interview solves it.' Even with a fixed question list, the impression formed in the first 3 minutes comes back in the follow-up questions. Candidates who make a good impression get questions that help them along, while the others get questions that probe their weak spots. Evaluation form design works the same way: put an overall comments box above the item scores, and people write the summary first and then work backward to fit the item scores to it. When this goes on, rejection reasons converge on a single line: 'not a good fit', and when a candidate or a union later raises a fairness complaint, the company has no records to show.
Ara LimExpert in global leadership, AI, and business communication
John KimInstructor John Kim, author of B2B Marketing Design and CEO of Nanum ConsultingIn 1920, psychologist Edward Thorndike analyzed military commanders’ ratings of their subordinates in his paper "A Constant Error in Psychological Ratings" and discovered, and named, a phenomenon in which the rating of one trait correlates abnormally highly with the ratings of other traits.
Consider a hypothetical case. Company A, an industrial parts manufacturer with 410 employees, hired 62 experienced professionals over the past 2 years, and 19 of them (31%) left within a year. When HR went back through the interview records, the overall comments box on successful candidates’ evaluation forms kept repeating phrases like 'good impression' and 'you can feel the energy.' They redesigned the interview process, starting with the next half-year’s hiring.
- The HR head deleted the overall comments box that had been at the top of the scorecard, leaving only four items: job knowledge, problem solving, collaboration, and tenure stability. For each item, scores could not be entered unless the interviewer wrote one line on 'what this candidate actually did.'
- In an interviewer training session, we handed 14 line managers 30 evaluation forms from last year’s rejected candidates and asked them to rewrite the rejection reasons. Of the 30 forms, 22 said 'not a good fit,' and only 5 cited behavioral evidence.
- They finalized a question list and three follow-up questions for each job in advance, and moved background questions like 'Where did you work before?' to the last 15 minutes of the interview. In the first-round résumé screening, reviewers read only the work-history descriptions, with company and school names masked.
- Scoring was done in the five minutes right after each interview: everyone wrote down their scores on paper first, and only then was the discussion opened. In the first round, when a business unit head opened with 'I thought that candidate was good,' the HR executive cut in: 'Let’s flip the scores over before we start.'
- Six months later, the average gap between interviewers’ scores on a 5-point scale had narrowed from 1.8 to 0.7, and first-year turnover among new hires fell from 31% to 14%. The share of hires from top-ranked universities also shifted, from 54% to 33%.