The most commonly used data integrity models today are Bibba, Wilson-Clark and Chinese models. These models are designed for both data integrity protection and confidentiality. Many optimization problems are related to linear programming. In practice, these variables involved are probabilistic. This paper proposes a data integrity model based on data anomalies assuming data are under stochastic linear constraints. An algorithm is constructed using the simplex method to find confidence intervals for the problem solutions. In the end the results from Monte Carlo simulation are compared with those from simplex method.