Interview Scorecard

What an interview scorecard is

An interview scorecard (بطاقة تقييم المقابلة) is a written form on which the interviewer records an assessment of each candidate against criteria agreed for a specific job before any interview begins. For each criterion it carries a weight, a scale with clearly described scores, and a box for the evidence on which the judgement rests. It is a tool for recording, not for asking: the questions are put in the interview, and the scorecard keeps what was heard in a form that allows comparison.

It is the written side of the structured interview, an interview in which every candidate is asked the same questions in the same order and the answers are assessed on a scale fixed in advance. Without the scorecard, the interview stays structured in its questions alone, and its scores become impressions written down after the session ends.

What an interview scorecard contains

• Criteria. These are the qualities and abilities the work requires, derived from job analysis rather than from the interviewer’s personal estimate.
• Weight. This is the share of each criterion in the final score, and the weights total 100.
• Scale. It describes what each score means in observable terms, so that a 4 from one interviewer is not a 3 from another.
• Evidence. This is what the candidate said or did, written in the candidate’s own words rather than as a general description. The difference between “excellent communicator” and “explained how he rewrote a monthly report for an angry customer” is the difference between an impression and evidence.
• Recommendation. This is the interviewer’s decision after the scores, with a minimum that may not be crossed where one is set.

The scorecard has room for different kinds of questions. It records answers to behavioral interview questions, which ask the candidate to describe a situation that really happened, and answers to the situational interview, which presents a situation that has not yet occurred and asks what would be done in it. Each type has its own scale, since the answer in the first is an event and in the second an intention.

How weights change the result between candidates

Suppose the scorecard for a senior accountant carries these criteria and weights: technical knowledge 50, problem solving 20, communication 20 and teamwork 10. The scale runs from 1 to 5. These figures are hypothetical.

• The first candidate. The scores are 5, 3, 2 and 2, so the simple average is (5 + 3 + 2 + 2) ÷ 4 = 3.0.
• The second candidate. The scores are 3, 4, 4 and 3, so the simple average is (3 + 4 + 4 + 3) ÷ 4 = 3.5.

By simple average the second candidate leads. Once the weights are applied, the first leads: 5 × 0.5 + 3 × 0.2 + 2 × 0.2 + 2 × 0.1 = 3.7, which is 74%, against 3 × 0.5 + 4 × 0.2 + 4 × 0.2 + 3 × 0.1 = 3.4, which is 68%. The order reversed because technical knowledge weighs more than the other criteria, and the calculation simply applied what the organisation had decided.

In a scorecard, the weights therefore carry the decision. They are written before any candidate is seen. If they are changed after the scores are known, the change becomes a way of justifying a choice the interviewer already leaned toward. A minimum is also added for a decisive criterion, for example that technical knowledge must not fall below 3, so that strength in one criterion does not cover weakness in the decisive one.

How a scorecard is built for a new job

The scorecard starts from the job, not from a stock template. Job analysis shows what separates those who succeed in the job from those who struggle, and this is condensed into a number of criteria that the session has room to test. The weight of each criterion is then set by what follows from failing in it: a criterion that halts the whole work when it is missing takes a larger weight than one that training can make up.

Next, each criterion receives a description in observable points, and we suggest that it covers at least the low, middle and high scores. For problem solving, for example, score 1 means describing the problem without analysing its causes or putting forward options. Score 3 means identifying the causes, proposing one solution and trying it. Score 5 means comparing options, choosing by a stated standard and then explaining how the result was checked. Scores 2 and 4 are left for what falls between the descriptions. The scorecard is tried in a trial interview before it is adopted, which shows whether two interviewers understand the same description in the same way.

What is written in the evidence box

The box holds what the candidate said or did that bears on the criterion, in the candidate’s own words or close to them. When the candidate is asked about a disagreement with a colleague under teamwork, a good piece of evidence is a line such as “said he asked his colleague for a meeting and proposed a shared schedule of tasks”, and a poor one is “seems cooperative”. The first can be reviewed and discussed. The second is an opinion that cannot be tested.

Anything unrelated to the criteria of the job stays out of the box: age, family status, or what is said about the candidate from outside the interview. If any of it comes up in the session it is not recorded, and if it is recorded, the assessment file then holds something that the person reading it should not build a decision on.

How the scores of more than one interviewer are recorded

In a panel interview, a session in which one candidate sits before several interviewers, each member completes a scorecard alone before anyone speaks. If they discuss first, the opinion of the first speaker fixes the opinion of the rest. Suppose three interviewers gave the candidate 2, 4 and 5 for communication. The average is 3.67, and the gap between the lowest and the highest is 3 points.

In this example the average hides the dispersion. A large gap signals that one interviewer saw something the others did not, or that the scale is understood in different ways. We suggest a rule written in advance: any gap of two points or more on one criterion requires each interviewer to present the evidence behind the score, after which each may adjust the score, with the reason for the adjustment written down.

How the interview result is combined with other selection tools

The scorecard summarises the interview, and the interview is one stage of selection. If the process includes a work sample test, a short task taken from the job itself and assessed against a written standard, the two scores are combined with weights the organisation decides in advance.

Suppose the interview score is 74% and the test score is 80%, with weights of 60 for the interview and 40 for the test. The combined score is 74 × 0.6 + 80 × 0.4 = 76.4%. With weights of 50 and 50 it would be 77.0%. The difference is small in this example, but it can reverse the order between two candidates who are close, so the weights are written before assessment begins, just as the weights of the criteria are.

Where a scorecard does not help

A scorecard is of little help when the candidates are few. If 3 of the 4 candidates who entered the selection stage are hired, the ratio is 0.75 (the number hired ÷ the number who entered the stage, which is the selection ratio). Hiring then becomes acceptance more than selection, and the scorecard turns into documentation of a decision already taken. By contrast, if one of ten who entered is hired, the ratio is 0.1, and the scorecard has wide room to sort.

It also does not help when interviewers have not been trained to use it, when it is filled in after the decision to justify it, or when it is asked to rule out every error of judgement. It organises judgement and does not replace it.

What to tell a candidate who was not accepted

The candidate needs to know that the decision was built on the criteria of the job. The reply names the criteria used for the assessment, without the detailed scores or what others said. It is worded to describe the organisation’s decision, without comment on the candidate personally. The organisation keeps the line of correspondence open for anyone who wants to understand where the gap lay, and whoever asks is answered with what relates to the criteria alone.

What spoils a scorecard

• Writing days after the session. Selective memory takes the place of observation.
• A score with no evidence. A score of 5 with no line to justify it can be neither reviewed nor defended.
• Impression carried between criteria. If the candidate impresses the interviewer on one criterion, the interviewer raises the scores on the others, which is called the halo and horn effect. The remedy is to record each criterion separately with its own evidence.
• Early judgement. An opinion forms in the first minutes and the remaining answers are then interpreted through it, which is the first impression error.
• Too many criteria. A scorecard with twelve criteria may not fit what the session can test, and some may be filled by guesswork.
• Criteria that do not belong to the job. Appearance or style in social conversation are examples, and they bring into the score traits the work does not need.

How a scorecard is reviewed after hiring

Its value shows after some months. The organisation compares the scores of those hired with what their performance shows after a set period, which is one aspect of quality of hire. If a criterion carried a large weight and did not relate to performance at all, the weight is reconsidered, or the scale that measures it.

The review has limits that should be stated. The sample may be small for a single job, and the performance of those rejected is unknown, so it is not known whether they would have failed. For that reason the weights are not adjusted abruptly after one or two hires, and the results of several batches are gathered before anything is changed. The scorecard is also reviewed when the work itself changes: a job that has had new duties added may need a criterion that was not on the scorecard.

Scorecards are kept according to the organisation’s policy on recruitment records, since they show on which criteria the decision was built. We suggest that what is written in them about a candidate who was not hired not be used for any purpose other than the one for which it was collected.

This is an explanation of the concept, not legal advice.