Abstract: Over the last decade, the rise of Massive Open Online Courses (MOOCs) has led new players to join this new way of thinking about teaching and learning. The multiplication of courses offered, the broad spectrum of disciplines they now cover, as well as the arrival of new players such as universities raise the question of assessing the quality of content offered in MOOC. This represents a new challenge in terms of quality assessment because MOOCs users appear to be fundamentally different from users of other distant learning courses. This study aims to operationalize a portion of the quality-assessment approach developed by Roy, Poehlluber, Garand and Beauchamp-Goyette (2016) with the introduction of a semi-automated lexicometric content analysis approach of MOOCs users’ appreciation surveys. Specific attention is paid to the reliability and validity of an analytical method that can both reflect the main trends and the specificities of the needs expressed by the students and effectively feed the instructional design process.