KAI: To self-score or to not self-score

– Deborah B. Pettry

The KAI is simple to administer and score. Indeed, it seems so simple that we practitioners – always pressed by clients for time and cost savings – are (wrongly) tempted to use it as a self-scored instrument.

During a recent KAI Advanced Workshop, it was emphasised, once again, that the KAI must not be used in a self-scored format. Rather, it should be collected upon administration and scored by the certified administrator.

Here is my confession: When I first started, I occasionally used the KAI as a self-scored instrument. This was my rationale: How could I score KAIs for 70 people in a two-hour session? Were the dangers listed by Kirton really likely to occur? Certainly I agreed that we owe our clients accurate feedback – but how real is the chance of mis-scoring? I believed that there was little danger.

As it happens, I had administered the KAI to a group of non-profit organisation CEOs the week before attending the Advanced Workshop. They had self-scored in the session, then transferred their own scores to the Feedback Summary and given me their inventories. Before leaving for another workshop, I had asked my assistant to check their scoring for me.

Here is what she had found:

  • Of the 13 inventories, 7 had to be corrected
  • Three persons erred in entering the scores in the correct blanks
  • Two persons made addition errors
  • One person totally missed an item
  • One person greatly under-used the “Hard/Very Hard” response
  • Three persons made more than one kind or error

Okay, the errors were not major, but they were frequent. And they were made by a small group of intelligent, responsible people in a low pressure situation where I could check over their shoulders to be sure they “had the general idea” in scoring. What may be the error rate in a group of 70 people with an instructor rushed for time and no possibility of glancing over shoulders?

How can I (and others like me) meet client time constraints without violating the measure? Here are some ideas: (1) When possible, ask people to complete the instrument in advance, providing it to you for scoring before you meet with them; (2) If you have less than two hours and no chance of advance scoring, do not use the instrument itself. Explain the concepts and allow people to estimate their relative positions in problem-solving preference without actually administering the inventory. (3) If you have up to three hours, administer the inventory at the start and ask an assistant to score while you explain the concepts and provide opportunities to work with them. Feedback the scores during the final hour. Do not use factor scores, as there is not enough time to explain them thoroughly; (4) If you plan to use factor scores, schedule a minimum of four hours for the total session; (5) Estimate that it will take 10 minutes to score, proof and record each respondent’s scores.

Schedule your staffing and agenda accordingly, and (6) Bring calculators!

So okay, Dr. Kirton, you’re right. From now on the KAI is not a self-scored instrument in my practice.

Originally published in KAI News 1993

EDITORS NOTE: We now have a computer program, KAI Scoring Assistant, which can substantially speed up the scoring process for consultants.