Contrast bias: Definition, examples, and how to prevent It in hiring

You interview two candidates back to back. The first is weak. The second gives a decent answer, and suddenly they seem exceptional. Then a great candidate walks in, and that same decent performance feels average. Nothing about the candidates changed. your reference point did.

That is contrast bias. It happens when candidates are judged against the people interviewed before them instead of against a consistent standard for the role. Research on real interview sequences shows that previous candidates can influence how evaluators assess the next person, making a structured scoring process important for keeping hiring decisions focused on job-related evidence.

TL;DR

• Contrast bias happens when an interviewer judges a candidate against another candidate rather than against a consistent job-related standard.
• It can make the same candidate appear stronger or weaker depending on who was interviewed immediately before them.
• Research on real interview sequences has found meaningful effects from the previous candidate on subsequent evaluator decisions.
• In performance reviews, a similar problem is often called contrast error.
• The practical response is to use predefined scoring criteria, evaluate candidates independently, and reduce unnecessary sequence effects.

What is contrast bias?

Contrast bias, also called the contrast effect, is a cognitive bias where an evaluator judges a person against someone they assessed recently instead of against a consistent standard. In hiring, this can make a candidate seem stronger after a weaker candidate and less impressive after a stronger one, even when the candidate's own qualifications have not changed.

The effect can be difficult to notice because the comparison feels like normal calibration. After several weak interviews, an average answer may sound unusually strong. The problem is that the reference point has shifted from the requirements of the role to the candidates seen immediately before.

Why does contrast bias matter in hiring?

Contrast bias matters because the order in which candidates are evaluated can influence how their performance is perceived. Research on real interview sequences has found that the quality of a previous candidate can affect the assessment of the next candidate.

There can also be a fairness dimension. Research on sequential spillover effects has found that candidate characteristics can interact with sequence effects, meaning that who is evaluated immediately before a candidate can influence how that candidate is rated. This makes interview order and evaluation structure relevant not only to consistency, but also to fair hiring processes.

The practical takeaway is simple: when candidates are evaluated against a predefined standard rather than against one another, there is less room for the previous candidate to become the benchmark.

What causes contrast bias in interviews?

Contrast bias is caused by an evaluation process that never fixes its reference point, so the most recent candidate becomes the reference point by default. Three conditions make it much worse, and all three are scheduling decisions rather than personality traits.

Back-to-back scheduling is the first. Sequence effects need a fresh memory to work with, and a day of interviews stacked without breaks keeps the previous candidate loaded and available.

The second is scoring after the fact. An interviewer who writes all five scores at the end of the day is not scoring five candidates. They are ranking them, which is a different task with a different answer. Ranking is comparative by construction, so contrast bias is not a risk there, it is the method.

The third is a vague rubric. "Strong communicator" has no fixed meaning, so it borrows one from context, and context is whoever spoke last. A rubric that says what a 3 sounds like, with an example answer, gives the score somewhere else to anchor. That is also the difference between a structured and an unstructured interview, and it is the difference that matters most here.

Affinity bias definition and how it differs

An affinity bias definition is straightforward: affinity bias is preferring a candidate because they remind you of yourself through a shared school, hometown, accent, hobby, or career path. It is worth separating from contrast bias, because the two get treated as interchangeable in training decks, and they need opposite fixes.

Contrast bias is about the candidate's neighbor. Affinity bias is about the interviewer. Blind screening helps a lot with the second and does almost nothing for the first, which is why teams that anonymize resumes are often surprised to find their interview scores still drift by slot.

How do you stop contrast bias in hiring?

You stop contrast bias by removing the interviewer's ability to compare candidates at the moment of scoring. Not by asking them to try harder. Five changes, in the order they pay off.

1. Write the rubric before the first interview. Define what a 1, 3 and 5 look like for each competency, with an example answer at each level. A score with a written anchor cannot borrow one from the previous candidate.
2. Score immediately, then close the document. The score goes in before the next candidate joins. No end-of-day batch scoring, no adjusting yesterday's numbers once today's candidates are in.
3. Ask every candidate the same questions in the same order. The US Office of Personnel Management describes a structured interview as one where all candidates are asked the same predetermined questions in the same order and all responses are evaluated using the same rating scale and standards for acceptable answers. That second half is the part teams skip.
4. Randomize the running order and cap the day. Four interviews with real gaps beats seven back to back. If sequence moves scores, then at minimum stop letting application date decide the sequence.
5. Put an objective measurement in front of the interview. A scored work sample or skills test taken before anyone meets anyone gives every later judgment a fixed anchor that no interview order can move.

Skills-first screening is doing the heavy lifting in that list, and the evidence behind it is not a vendor claim. The large 2022 reanalysis of selection method validity by Paul Sackett and colleagues, which revised many long-quoted estimates downward, still places structured interviews among the strongest predictors of job performance while years of education and general years of experience sit among the weaker ones. Structure is the thing that survives every reanalysis, even when the exact coefficients are argued over.

Pro Tip: Audit for this before you redesign anything. Pull the last 90 days of interview scores, tag each one with its slot position in the day, and compare the average score in slots 1 and 2 against slots 4 and 5. If later slots run consistently lower, the running order is doing some of your hiring. It is a 20 minute query and it turns an argument about whether bias exists into a number the hiring manager can see.

One caveat worth naming. Randomizing order and fixing the rubric will not help if the questions themselves are not tied to the job. A perfectly consistent process for measuring the wrong thing just gets you reliably wrong answers, which is why building the assessment around the actual competencies comes before any of the scheduling work.

How does a scoring standard replace comparison?

Testlify turns candidate evidence into a structured hiring decision by combining assessment results, AI insights, reviewer feedback, references and interview data, while final judgment stays with the hiring team. Applied to contrast bias, its job is narrow and specific: give every evaluator a number that already exists before the interview starts.

A few mechanics matter more than the rest here. Test weighting is fixed in advance, from x0 to x5 per test, so the relative importance of each competency is a decision the team makes once rather than a feeling that shifts across an interview day. Benchmarking reports a candidate's percentile and rank against a defined benchmark rather than against whoever sat in the room before them. Score thresholds gate stage changes, so a candidate advances by clearing an absolute bar, not by outshining their neighbor.

On the AI side the product carries its own disclaimer, and it is the right one: AI scores and insights are for guidance only, use human judgment for final decisions. Including the AI score in the final average is a toggle a customer can leave off. That matters for a bias conversation, because swapping a human comparison habit for an unexplainable machine score would not be progress.

Most teams reading this already run an applicant tracking system they trust, and there is no reason to change it. Testlify integrates as the screening and interviewing layer through 100+ ATS integrations and leaves the ATS as the system of record, which is how assessment results end up on the candidate record without anyone rekeying a score. Teams with no applicant tracking system at all can use the simple built-in pipeline instead, though it is deliberately basic and is not meant to stand in for a dedicated system of record.

Hire on evidence, not on interview order

If your interview scores still depend on who went first, a scored assessment before the first conversation is the shortest path out. Book a demo to see how a fixed scoring standard and role-relevant pre-employment tests give every reviewer the same anchor.

Frequently asked questions (FAQs)
What is an example of contrast bias in an interview?