In the past decade a great amount of experimental and observational research has been devoted to discovering if there is a connection between electromagnetic fields (EMF) and leukemia. Meta-analysis was used by Kheifets et. al. (1997) [KAB97] to estimate the association between occupational jobs exposed to Electromagnetic Fields (EMF) and leukemia. In this paper, I discuss new approaches for evaluating publication bias and heterogeneity in research studies. I then look at how fixed, random, and mixed effects models can be applied to combining the estimates from observational epidemiology studies. These methods were then applied to re-evaluating Kheifets et. al. work on estimating the effects of occupational EMF exposure on causing leukemia. My analysis confirmed Kheifets et. al. results. However, the new statistical and graphical methods provided a clearer picture of the data and stronger evidence for the conclusion.