In this study, I delved into a dataset provided by the Centers for Disease Control and Prevention (CDC), with the objective of exploring the relationship between diabetes and two predictor variables, namely inactivity and obesity. The aim was to employ statistical methods to investigate whether these variables are significant predictors of diabetes. Firstly, I conducted a t-test to compare the means of two groups, namely those with diabetes and those without. It’s worth noting that t-tests are known to have certain built-in assumptions, as elucidated in the Wikipedia article on the subject. However, in many practical applications, t-tests exhibit robustness even when these assumptions are not fully met. Nonetheless, for datasets that deviate significantly from normality, the assumptions underlying t-tests can render the estimation of p-values unreliable. In order to address this issue and obtain a more reliable assessment of the significance of the observed difference in means, I employed a Monte Carlo permutation test. This computational procedure allowed me to estimate a p-value under the assumption of a null hypothesis, which posits that there is no genuine difference between the two groups. It’s crucial to emphasize that simply applying a t-test using appropriate software may not provide an intuitive understanding of how the p-value was calculated. Therefore, the adoption of a Monte Carlo procedure in this study was not only useful but also informative. It offered a more robust approach to estimating the p-value, considering the non-normal nature of the data. This analytical framework facilitated a deeper and more nuanced exploration of the relationships between inactivity, obesity, and diabetes, shedding light on potential predictors and their impact on the dataset.