This work explores using a low-cost electroencephalography (EEG) headset to quantify the human response to stressed and non-stressed states. We used a Stroop color-word interference test to elicit a mild stress response in 18 test subjects while recording scalp EEG. EEG signals were analyzed using an algorithm that computed the root mean square voltage in the beta, alpha, and theta bands immediately following the presentation of the Stroop stimuli. These features were then used as inputs to logistic regression and k-nearest neighbor classifiers. Results showed that there was a median accuracy of 73.96% for classifying mental state using the O1 sensor on the Emotiv headset.