Expectancies of success and values were used to predict success in an online undergraduate-level introductory statistics course. Students who identified as primarily face-to-face learners were compared to students who identified as primarily online learners. Expectancy value theory served as a model. Expectancies of success were operationalized as self-efficacy for learning online and self-efficacy for learning statistics. Values were separated into the worth of learning statistics and the value of grades in the course. The purpose of this study was to determine if there are differences in the variables that may be used to predict final exam scores and successful course completion in typically face-to-face and typically online students, because there are differences in the populations of students who tend to take courses in these two different formats (i.e., traditional and adult learners). In predicting final exam grades there were no interactions with typical mode of instruction, though worth of statistics was a significant covariate and there was a main effect for typical mode of instruction. In predicting successful course completion, there were interactions between typical mode of instruction and one of the online learning self-efficacy subscales as well as the worth of statistics scale. These results are discussed in relation to the application of mainstream motivational models in the populations of traditional and adult learners.