This paper proposes a new mathematical model for the scheduling of employees. The problem consistsin finding and optimal scheduling method that will balance the workload and avoid dissatisfaction and low performance problems, due to work overload or sudden changes in the workload. We will employ hypothesis testing based on the normalized Bernoulli trial, in order to do a statistical comparison between the tasks assigned by the proposed model with those assigned by the Zhirong Liang, Songs han Guo, Yanzhi Li and Andrew Lim model, which minimizes the difference between minimum and maximum workloads assigned. As a result we observed that the assignations generated by the models do not match in more than 90% of the cases. Further more,we found that the values of the object functions, which were gathered from solutions provided by the current model, were never better than those obtained by this paper proposes.