TY - JOUR

T1 - Statistical power and cost in training evaluation

T2 - Some new considerations

AU - Yang, Hyuckseung

AU - Sackett, Paul R.

AU - Arvey, Richard D.

PY - 1996/1/1

Y1 - 1996/1/1

N2 - Two ways to reduce the costs of training evaluation are examined. First, we examine the potential for reducing the costs of training evaluation by assigning different numbers of subjects into training and control groups. Given a total N of subjects, statistical power to detect the effectiveness of a training program can be maximized by assigning the subjects equally to training and control groups. If we take into account the costs of training evaluation, however, an unequal-group-size design with a larger total N may achieve the same level of statistical power at lower cost. We derive formulas for the optimal ratios of the control group size to the training group size for both ANOVA and ANCOVA designs, incorporating the differential costs of training and control group participation. Second, we examine the possibility that using a less expensive proxy criterion measure in place of the target criterion measure of interest when evaluating the training effectiveness can pay off. We show that using a proxy criterion increases the sample size needed to achieve a given level of statistical power, and then we describe procedures for examining the tradeoff between the costs saved by using the less expensive proxy criterion and the costs incurred by the larger sample size.

AB - Two ways to reduce the costs of training evaluation are examined. First, we examine the potential for reducing the costs of training evaluation by assigning different numbers of subjects into training and control groups. Given a total N of subjects, statistical power to detect the effectiveness of a training program can be maximized by assigning the subjects equally to training and control groups. If we take into account the costs of training evaluation, however, an unequal-group-size design with a larger total N may achieve the same level of statistical power at lower cost. We derive formulas for the optimal ratios of the control group size to the training group size for both ANOVA and ANCOVA designs, incorporating the differential costs of training and control group participation. Second, we examine the possibility that using a less expensive proxy criterion measure in place of the target criterion measure of interest when evaluating the training effectiveness can pay off. We show that using a proxy criterion increases the sample size needed to achieve a given level of statistical power, and then we describe procedures for examining the tradeoff between the costs saved by using the less expensive proxy criterion and the costs incurred by the larger sample size.

UR - http://www.scopus.com/inward/record.url?scp=0030306013&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0030306013&partnerID=8YFLogxK

U2 - 10.1111/j.1744-6570.1996.tb01588.x

DO - 10.1111/j.1744-6570.1996.tb01588.x

M3 - Article

AN - SCOPUS:0030306013

SN - 0031-5826

VL - 49

SP - 651

EP - 668

JO - Personnel Psychology

JF - Personnel Psychology

IS - 3

ER -