Technology acceptance models presuppose that technology users have clearly defined attitudes toward technology, which is not necessarily true. Complementary, social-psychological research proposes attitude strength (AS), a construct that has been so far insufficiently examined in the context of technology acceptance. Attitudes toward technology might become weaker after frequent changes in the used technology. This study examines the relationships between AS and educational technology acceptance predictors. In the case of N = 225 German undergraduate students of Educational Sciences, "millennials" using the learning management system Moodle, and based on structural equations modeling (SEM) and fuzzy set qualitative comparative analysis (fsQCA), we found significant relationships between AS and acceptance predictors. Further results suggest two situations leading to technology acceptance, one in which students are performance-oriented and comply with faculty recommendations; the other in which students are technically experienced and will accept any technology, but avoid technical problems and effort. While the latter situation is only vaguely suggested by SEM, it is much clearly indicated by fsQCA. For acceptance research, we conclude that current acceptance models should be extended by AS, and employ fsQCA. For educational practice, we recommend using fsQCA to assess acceptance predictors when educational technology is implemented in higher education.