Literature Review on Performance Appraisal: Themes, Methods and Findings

Performance appraisal research spans five recurring strands: the method families, fairness and justice perceptions, rater bias, how employees react to feedback, and the awkward finding that many appraisal systems fail to improve performance at all.

The hard part of this literature is not the volume of findings but their disagreement with each other, since the field is unusually critical of itself. That disagreement is not noise to be tidied away. It is the most reportable thing about the literature, and once you can see where it comes from, the topic starts to organize itself.

What a literature review on performance appraisal has to cover

Almost every credible review divides the ground the same way, and it is worth adopting that division rather than inventing one. There is a methods section, describing the families of appraisal instruments and what each is for. There is a fairness and justice section, covering how employees perceive the process. There is a bias section, covering the systematic errors raters make. There is a reactions section, covering feedback acceptance. And there is an effectiveness section, asking whether any of it works.

The methods form the natural spine, because nearly every other debate attaches to one of them. Five families cover the ground.

Graphic rating scales

The oldest and simplest family. The rater scores the employee on a set of dimensions or traits using a numeric or descriptive scale, and the result is a number that can be compared across people and years.

What distinguishes it is also its weakness: it rates abstract attributes rather than observed behavior or delivered results. That leaves it more exposed than any other family to the judgment errors described further down — halo effects, leniency, and the tendency to score the impression already formed.

Behaviorally anchored rating scales (BARS)

BARS keeps the scale but replaces the adjectives. Each scale point is tied to specific narrative examples of what good and poor performance actually look like, developed from critical incidents collected in the role itself.

The distinguishing feature is that the anchors are observable behaviors rather than descriptions of a person. Two raters looking at the same behavior have less room to interpret it differently, which is the whole design rationale. The cost is development effort: anchors have to be built role by role, which is why BARS is more common in the literature than in practice.

Management by objectives and results-based appraisal

Here performance is evaluated against objectives agreed in advance. The classic MBO cycle runs through four steps — setting goals, planning, self-control, then evaluating performance — with objectives written to be specific, measurable, attainable, results-oriented and time-limited. OKR-based reviews are the contemporary variant of the same logic.

The distinguishing feature is that it judges outcomes against a pre-agreed standard rather than judging the person. That relocates the decisive moment: in a results-based system, what happens at the goal-setting conversation matters more than what happens at the review.

360-degree and multi-source feedback

Ratings are collected from several organizational levels at once — peers, subordinates, superiors, sometimes customers — and combined into a single picture of how someone works.

The distinguishing feature is the removal of the single supervisor as sole rater. That trades one person’s blind spots for a broader but noisier composite, which is why the literature treats 360-degree feedback as strong for development and considerably more contested for pay and promotion decisions.

Continuous performance management and self-appraisal

The newest family covers frequent, lightweight check-ins and structured employee self-assessment, replacing or supplementing the annual event. This is the family behind the widely reported corporate move away from once-a-year ratings.

It is defined by frequency and developmental purpose rather than by any instrument, and it is the only family that structurally requires the employee to contribute something. That makes it the point where the methods literature and the participation literature meet.

Reviews normally close the methods section with the same argument: there is no single best method. As appraisal-methods overviews such as the University of Wisconsin-Green Bay’s Introduction to Human Resource Management set out, the right instrument depends on the decision it has to support, because the evidence needed to justify a promotion is not the evidence a manager needs in order to coach someone.

What the performance appraisal literature actually finds

This is where a review earns its marks, and where the contradictions live.

Start with effectiveness, because it sets the tone for everything else. The CIPD’s rapid evidence assessment of the research literature on the effect of performance appraisal on workplace performance is the standard reference point, and its conclusion is deflationary: the evidence that appraisal reliably improves workplace performance is weaker and far more conditional than practitioner assumption suggests. The mechanism is worth stating explicitly in your review. Appraisal is not one practice but a bundle — rating, feedback, goal setting, and often pay linkage — whose components pull in different directions, so the aggregate effect washes out.

The critical strand follows from that. A 2025 review titled The Good, the Bad and the Ineffective: An Appraisal of Performance Appraisals catalogs what undermines these systems: biased and unfair processes, a lack of transparency about procedural clarity, and insufficient attention to the psychological and political influences acting on a rating. Note what is absent from that list. These are process failures, not instrument failures. The form is rarely the problem.

Fairness has since become a construct in its own right, with a dedicated systematic review building a conceptual framework for enhancing fairness in appraisals, and a 2025 study in Cogent Business & Management asking whether public-sector appraisals are fair and which practices reduce bias. The structural point behind that volume of work is that perceived fairness, rather than measured accuracy, is what predicts employee attitudes and behavior afterward.

Bias supplies the explanation. There is strong and consistent evidence that when one person evaluates another’s performance, systematic errors of judgment occur, and confirmation bias is repeatedly singled out: evaluators attend to information confirming what they already believe about the employee rather than assessing actual performance. For anyone who will be appraised, that finding has a blunt implication. A rating partly measures an impression that was formed over months, which means the work of shaping it happens long before the meeting.

Against that fairly bleak picture sits the one consistently positive finding, and it is the one you should give the most space. Employee involvement in the process improves goal setting, evaluation and feedback acceptance. It is also the only major lever the appraised person controls. Participation converts an appraisal from a verdict delivered into a conversation contributed to, and accepting the feedback is the precondition for any behavior change that follows. Most students write that sentence as a research finding and then walk into their own first review having never acted on it, so it is worth seeing what you already bring before you are sitting on the receiving end.

Two further sources are worth having for structure. Brown, O’Kane, Mazumdar and McCracken’s scoping review in Human Resource Development Review (2019) is the most citable single mapping of the field, and a legitimate methodological model if your own review charts breadth and gaps rather than pooling effects. And Keegan and colleagues, in the Human Resource Management Journal (2019), found employees in project-based organizations absorbing appraisal work that theory assigns to line managers — a direct challenge to the supervisor-subordinate model the rest of the literature assumes, and good material for a gaps section. The newest strand, the use of machine learning to counter attribution errors and rater bias, is still largely conceptual rather than evaluative, which makes it another honest gap to name.

The skills behind a review that goes well

Read enough of this literature and a pattern surfaces that the studies themselves rarely state directly. The variables that keep predicting whether an appraisal works — participation, feedback acceptance, the quality of the supervisor relationship, whether results were visible in the first place — are not properties of the form. They are things people do. And things people do can be learned.

Working with Your Manager is the closest match to what the research keeps measuring. A review is not something done to you; it is something you prepare for, contribute to, and steer toward the future. In practice that means arriving with your own record of what you delivered, settling goals and decision-making authority in advance rather than in the meeting, and treating the conversation as one point in an ongoing partnership rather than a verdict handed down. The participation finding above is, behaviorally, a description of this skill.

Building Self-Awareness covers the receiving end. Most of what the literature files under feedback reactions is really the difficulty of hearing an evaluation without immediately defending against it — understanding it first, adding your own view second, reflecting afterward. It also accounts for the gap those studies keep measuring between how people rate themselves and how their supervisors rate them. Blind spots are ordinary, and the way to shrink them is to ask for specific, future-focused advice rather than wait for the annual form to tell you.

Influence deals with everything that happens before the meeting. If ratings partly reflect what a manager could observe and recall, then a reputation built on consistently delivered results and straight talk is worth more than any case you argue in the room. This is not impression management. It is the ordinary discipline of making real work visible while it is still happening.

The free Job Skills Test measures all three, alongside nine other work skills the framework treats as buildable rather than fixed, so a few minutes with it will tell you which to build first — and anything it flags is a starting point, not a verdict.

What this means once you are the one being appraised

You may already recognize some of this from how you handle feedback or prepare for a conversation that matters. Most people arrive at these skills the slow way, through a few uncomfortable reviews, which is exactly why the research finds so many systems underperforming — the behaviors that make appraisal work were never taught to either side of the table.

They can be, though, and none of it asks you to become a different person at work. The skills that matter most for you right now are probably a short list, not a long one. What does change is the weight they carry: the further into a career you go, the more of your progression runs through conversations of this kind. Having read this far, you have already done the part most people skip, which is thinking about the appraisal before being handed one.

See where your own skills stand

The literature is a map of what happens in the room. The remaining question is what you would bring into it.

The free Job Skills Test is a short self-assessment of your work skills. It shows you where you stand across all twelve, and which ones would make the biggest difference to you now — a concrete starting point rather than a general recommendation. If a performance review is somewhere in your near future, it is a sensible thing to do before it arrives rather than after.

Get my skills profile

Free, takes 7 minutes, and your results appear the moment you finish.