Knowledge that Transforms

Using Williamson's (1971) control loss model, the relative merits of dual hierarchies (P. B. Evans [Evans, P. B. 1975. Multiple hierarchies and organizational control. Admin. Sci. Quart. 20 250–259.), strong vertical dyadic linkages, and mentoring for reducing overall control loss are explored. The potential impact for each organizational structure or process is assessed. Simulations are performed to examine the impact of each of these organizational arrangements for organizations with hierarchical levels ranging from 3 to 11 levels. In addition, the effects of deteriorating control at each level are investigated. In the discussion of appropriate control strategies, a number of other organizational characteristics are discussed. I conclude that: Under conditions of sequential or reciprocal interdependence, building uniformly high control efficiency throughout the organization is essential. In small organizations (and those with high turnover) dual hierarchies and cadre relations are appropriate mechanisms for reducing control loss. In large organizations (and those with low turnover) mentoring and cadres are appropriate.

This paper focuses upon the maximization of labor productivity through selection of the optimal number, duration, and placement of rest breaks. It is demonstrated that resolution of the limited validity associated with previous models leads to a mixed-integer quadratic programming formulation of this problem. Efficient analytic procedures are developed which allow solutions to be obtained in minutes using a hand-held calculator. Then, the characteristics, properties, and potential effectiveness of optimal solutions are examined through an illustration from an existing work environment. Issues relevant to the implementation of the models are presented in a summary of an initial experiment in which one of the models in this paper was used to select rest break policy. The results of this experiment indicated that the output achieved using the model's policy was 13% higher than the output achieved when the subjects used their own rest break policies. Several future research topics are also identified.

The classical confidence interval estimator commonly used in simulation is compared with four new estimators based on standardization of a time series presented in a previous paper. These new interval estimators are shown to have asymptotic properties that strictly dominate the classical estimator when used with data from independent simulation replications or means of batched observations from a single simulation run. Two of the new estimators also can be used with a single unbatched replication of a simulation program, a situation where the classical estimator is not defined.