The dissertation investigates tests to check if regression models fit well. A new method is explored to assess fit for different types of logistic models. If a model doesn't fit, the test can show which parts are off. The study includes chapters on binary and multinomial logistic models, with tests for each. Simulation studies in R were done to see how well the tests work. A real dataset was used to show how the tests can be applied. Overall, the research provides insights into how to check if logistic regression models are a good fit.