Using the paired t test because the samples are correlated with each other could be another avenue. This would help us find the p-value and look at the statistical significance in evaluating the quality of care or outcomes in a medical home. This would help us identify how accurate a sample size would be in looking at positive outcomes or negative outcomes?