The attribution of learning success or failure is crucial for students' learning and motivation. Effective attribution of their learning success or failure in the context of a small private online course (SPOC) could generate students' motivation toward learning success while an incorrect attribution would lead to a sense of helplessness. Based on the attribution theories developed by Wiener and Kelly with learning analytics technology, this article developed a systematic framework for learning analytics in SPOC in an undergraduate course. A total of 197 undergraduates participated in the experiment and were randomly assigned to a treatment group (98 students) and a control group (99 students). The treatment group applied the learning analytics system framework in a SPOC, and the control group applied the traditional SPOC teaching method, and other conditions are consistent. Structural equation modeling is adopted to explore the correlations among attribution variables (i.e., ability, effort, difficulty of the task, and luck), motivational variables (i.e., endogenous motivation, and extrinsic motivation), and learning achievement. The results indicated that all four attribution variables have a significant correlation with both endogenous and extrinsic motivation and have a significant effect on learning achievement. Students in the experimental class exhibited better outcomes than students in the control class with respect to learning motivation and learning achievement. The findings validated the effectiveness of using the SPOC analytics system to facilitate SPOC learners making effective attributions, leading to the enhancement of learning motivation and achievement.