Two studies were performed to assess the effectiveness of a newly developed automated facial expression detection tool in assessing the quality of trainee engagement with training material relative to traditional Kirkpatrick level-1 style assessment survey techniques. The first study (N=17) established broad support for the superiority of automated facial expression recognition tools compared to traditional techniques. A follow up study (N = 37) was performed to refine and extend findings from the first study. Taken together, the two studies showed the superiority of automated facial expression detection in volume of data collected, identification of key performance indicators, and suitability for distributed online and other computer based learning environments. Both experiments were facilitated on consumer spec PC laptops using standard built in web camera hardware to collect participant facial expression feedback in reaction to viewing video based training content. Results show promise in future implementation regarding improvements to quantity, quality and method for data collection and analysis of user engagement with digital training.