How to allocate scores to candidates in an interview

By MUNGAI KIHANYA

The Sunday Nation

Nairobi,

30 November 2008

Human resource managers face a difficult problem when selecting the best candidate after a series of interviews. This arises because it is not possible to interview more than one person at the same time, thus comparison is not easy.

To illustrate the problem, suppose a vacancy attracts 100 interested applicants. The first step is short-listing those to be interviewed. This is easy to do because the application documents are all available at the same time and they can be compared, cross-checked and re-checked to ensure that only the best candidates are called for interviews.

The interview stage is the tricky part. It is common to rate candidates on a scale of, say, one to ten and then to compare the scores after all interviewees have been seen. But since there is no basis for comparison at the beginning, how many points should be allocated to the first interviewee?

Suppose he appears to be very good and we give eight points. But then, when the next candidate comes in, we find that she is at least twice as good as the first one! What score shall we give her? We can’t allocate 16 since our scale only goes to ten!

So we cancel the score of the first candidate and change it from eight to, say, four points and then award the second interviewee a score of eight. But still, the third applicant may turn out to be twice as good as the second one (whom we’ve already given eight points) and we have to go back and make cancellations and change the scores again!

Now that doesn’t look very professional, does it? Anyone looking at the score sheet might think that we are cooking the results to favour certain candidates for some unstated reasons. So here is a better way of allocating the points.

The first step is to change from a limited scale – the one to ten points – to an open ended range. That way, if a candidate who is much better than those already seen shows up, we will have the flexibility to allocate as many points as deemed necessary. In the above example, if the scale was open-ended, the second interviewee could have been allocated a score of 16 and the third one 32 points.

But what if a very poor candidate appears after we have already assigned a very low rating of, say 2 points to a previous one? Should we then give a negative score? No; that would be unkind!

To avoid that trap, the scoring should start at a fairly high value, say 100. Note that the 100 is NOT the highest NOR the lowest score. It is just the starting value. In other words, it is the number of points to be allocated to the first candidate, regardless of the performance in the interview.

Once the 100 points are awarded to the first interviewee, the second one is assessed in comparison to this. If she is twice as good, she score 200; if half as good she gets 50.

Now since we may not be able to memorise the performance of more than about three previous candidates, every assessment should then be done in comparison with the previous two interviewees.

Thus the third candidate is compared to the first and second; the fourth to the second and third…the tenth is compared to the eight and ninth…and so on. This way, we shall have a distribution of scores at the end of the interviews and these can be sorted to give the order of performance – from best to worst.