Cost Effectiveness of  B-APT

The capability of a test to save (or make) money for an organisation is known as "test utility". There is a test utility in the use of a valid aptitude test to select applicants for training because it increases the number of people who are above average in training and subsequently in on-the-job programming.

For example, a major insurance company selects 20 applicants each year for programmer training. A training director asks whether testing 80 applicants with an aptitude test would be worth the cost. In the past, 30% of the trainees were unable to complete the course satisfactorily. A conservative estimate of programmer training costs (including salary while learning) is £22,000 per person. At this rate 30% who fail in each training class represent a £132,000 loss for the company. The table below shows the estimated savings using the Berger Aptitude for Programming Test (B-APT).

The values in the table were determined by a formula using test validity, qualifying score, selection ratio, and training success rate. Although a generalised validity study of the B-APT based on 16 companies resulted in a validity coefficient of .74, a more conservative estimate, .65, was used in the formula. The qualifying score was 20 out of 30. Based on this correlation of 0.65 and the qualifying score out of 20, the training success rate was 95%. The table below indicates that there was a net saving of £105,200 in training costs as a result of using the B-APT for one group of 20 trainees.

Costs and Gains When B-APT is Used to Select Trainees