Predictive Validity of an Assessment Center Program in Big
Communities
The second city of
Sweden
,
Goteborg
, uses the
Assessment
Center
method for selection and development of operative and strategic
managers. The used set-ups of the centers are the traditional ones with
exercises, problem solving, meetings and interviews. The specific
exercises are for both centers group discussion, presentation, external
meeting, and for the operative center planning and internal meeting, and
for the strategic center in-basket and fact finding. The contents of the
centers and their impact on the development and selection of managers
will be presented and discussed. The assessors are trained in real
settings. The reliability of the ratings is estimated in specific
studies. For the follow-up study two questionnaires were constructed,
one for the nearest chief’s and one for the employees’ judgment.
There are two sets of results in the study. In the first place the
predictability of success in a center from back-ground and ability data
are presented. One result is the heavy effect of general ability and
educational level on the results in the centers. In the second place
predictive validity coefficients are presented for the regression of
criterion factors on abilities, exercises, and behavior factors. There
are validity coefficients at the level of 0,50. The follow-up study
revealed several interesting methodological problems that could be
discussed. Other aspects to be mentioned in this respect are information
and involvement issues, the presentation of a scientific based study to
laymen, and economic pros and cons for the use of
Assessment
Center.
Speaker
Bertil Mardberg
City of Goteborg, Sweden
