Aligning scores in an evaluation meeting does not necessarily lead to fairness. What should be aligned first are the facts observed.
In an evaluation meeting, a manager who gives a high score alone is asked for an explanation.
“Considering the balance with other departments, isn't that a bit too high?”
“Employees of the same grade are generally around this level.”
“Let's lower it by one level this time and look at it again next time.”
It is difficult for a manager to argue back when the issue of overall fairness is raised.
Even if they explain the individual's achievements, the response is, “There are others who are working hard as well.”
Conversely, there are managers who give low scores.
When asked for the reason, they reply, “I'm still not confident in entrusting them with it,” or “I feel they lack perspective.”
The discussion continues in the meeting, and in the end, the scores fall within a certain range.
HR feels that they have successfully adjusted the discrepancies between evaluators.
However, is it really the evaluation criteria that have been aligned?
Hasn't the score just shifted toward those with the loudest voices?
Have they not aligned only the conclusions while leaving the differences in the facts they observed?
Fairness in evaluation is not about giving the same score.
What should be questioned first is not the difference in scores, but the difference in observations that led to those judgments.
Evaluators do not necessarily look at the same employee.
Even if they use the same grade, the same evaluation items, and the same five-point scale, the work that the evaluators are looking at is different.
One manager looks at the fact that the person won a large project.
Another manager looks at the fact that the person increased support from those around them during that project.
One manager looks at the fact that the person spoke actively in meetings.
Another manager looks at the fact that the person gathered concerns from stakeholders and built consensus before speaking.
One manager looks at the results at the end of the term.
Another manager looks at the failures and corrections that occurred during the term.
When scores differ, we suspect a discrepancy in evaluation criteria.
However, before that, the scope of observation might be different.
If the facts being seen are different, it is natural for the judgments to be different.
If you adjust that only by scores, even important information will be erased.
Score alignment makes it easy to create the appearance of fairness.
Evaluation meetings have the role of aligning standards across the company.
If the leniency or strictness of each department is left unchecked, the same work might receive different treatment. There is also a risk that evaluations will be decided solely by a manager's assumptions or personal closeness.
Therefore, calibration is necessary.
However, adjusting the evaluation distribution is not the same as adjusting the quality of judgment.
Lowering scores that are far from the average.
Aligning the number of high-rated employees by department.
Considering conclusions that did not face strong objections as fair.
These are convenient for shaping the results.
However, the lack of observation, bias in evidence, and differences in role expectations remain unchanged.
Rather, managers learn.
Giving a high score makes explaining it troublesome.
Giving a low score leads HR to ask for justification.
If you choose a safe score from the start, the meeting ends quickly.
In this way, evaluations become aligned.
At the same time, the motivation to observe work closely weakens.
Addressing 'observation gaps' changes the questions asked in meetings.
Here, I would like to introduce the perspective of observation gaps.
Observation gaps are differences in judgment criteria that arise because each evaluator sees different situations, time periods, results, impacts on others, and performance conditions.
Observation gaps are not just errors to be eliminated.
They are also an entry point for bringing together work that was not visible to a single supervisor.
For example, when evaluating an employee with high sales results, do not start the discussion from the score.
With which client and what kind of results were achieved?
What results did the individual create directly?
To what extent did support from others and existing assets influence the outcome?
In that process, what was left behind in terms of team results and future reproducibility?
After laying out the facts, make a judgment based on the expectations for the grade level.
Then, the dispute over 'a 5 or a 4' changes into a discussion about 'how well they met the role expectation of leaving behind a reproducible sales process, despite high individual results.'
Before aligning scores, it becomes clear what is being evaluated.
In evaluation meetings, confirm four types of evidence.
To make them usable in tomorrow's evaluation meeting, divide the judgment materials into four categories.
Evidence of the situation.
When, in what work, and what happened?
Instead of saying 'they were always proactive,' state, 'when customer requirements changed, they organized the scope of impact and presented two response plans to the relevant departments.'
Evidence of the results.
What changed as a result of that action?
It is not just about sales. Look at the impact on work, such as reduced rework, faster decision-making, and juniors being able to use the same procedures.
Evidence of conditions.
Under what support, authority, relationships, and difficulty levels were the results achieved?
Do not attribute results solely to the individual's ability; view them alongside the conditions under which they were demonstrated. Looking at conditions is not to discount the results, but to judge reproducibility.
Evidence of change.
What have they become capable of doing compared to the beginning of the period?
Did they repeat the same mistakes? Did they correct them after being pointed out? Did they move from a state of needing support to being able to do it alone?
When these four are brought to the table, the evaluation meeting ceases to be a competition of impressions.
Even if the scores differ, you can discuss which evidence was weighted more heavily.
Increasing the amount of evidence does not necessarily lead to fairness.
Caution is needed here.
Increasing records does not automatically make evaluations correct.
Number of remarks, number of projects, meeting memos, chats, comments from others.
If only information that is easy to collect increases, easily measurable actions will be overvalued.
High-profile work is easy to record.
Work that prevents problems before they happen is hard to record.
People who speak in meetings are visible.
People who did the groundwork to make it possible to speak are hard to see.
Therefore, it is necessary to ask not only about the quantity of evidence but also about what is missing from the records.
The individual's self-description, the supervisor's observations, the voices of those involved, and the numbers each only see a part of the picture.
Bringing multiple pieces of information together is not to create a single truth.
It is to make more explainable judgments while being aware of the parts that are not visible.
Fairness is not a conclusion without hesitation.
It is the ability to explain why that conclusion was reached from limited evidence and to update it if necessary.
HR should change the language of the meeting before changing the distribution.
If scores are split in an evaluation meeting, before asking "Which is appropriate?", HR should ask in the following order:
"What specific situations are you each looking at?"
"If you separate actions from results, what was happening?"
"What conditions made those results possible?"
"Are there any facts that would lead to an opposing judgment?"
The last question is important.
Ask those who gave a high evaluation for facts that would lead to a low judgment.
Ask those who gave a low evaluation for facts that would lead to a high judgment.
Change the meeting from one that defends your own conclusion to one that also searches for evidence that might shake that conclusion.
Even so, judgments may still differ.
In such cases, there is no need to force everyone to adopt the same perspective.
Along with the final decision, document what remains unconfirmed and what should be observed in the next period.
While finalizing the current evaluation, use it as material to improve the quality of the next one.
Try hiding the scores for a moment at the next evaluation meeting.
At the next evaluation meeting, do not look at the scores for the first few minutes.
First, confirm the expected role.
Next, bring together evidence of the situation, results, conditions, and changes.
After that, look at each evaluator's scores and reasons.
This alone will change the order of the meeting.
When people look for facts after seeing scores, they tend to gather material that supports their own conclusions.
When people think about scores after looking at facts, room is created to reconstruct their judgments.
Do not fear variations in evaluations.
What you should fear is pretending to have seen the same thing and aligning only the conclusions when you have actually seen different things.
Fair evaluation is not born from meetings where scores line up neatly.
It begins with meetings where participants bring what they have seen, acknowledge what they have not seen, and document the reasons for their judgments.
Related Articles
#AI
#HR
#HR
#HRWork
#FutureOfHR
#PersonnelSystem
#EvaluationSystem
#EvaluationMeeting
#PersonnelEvaluation
#SystemDesign
#Fairness
#HumanResourceManagement
#TalentManagement
#StrategicHR
#HRBP
#Managers
#Management
#Feedback
#OrganizationDevelopment
#HumanResourceDevelopment