Background: While the effectiveness of tutoring-style videos has been reported in previous studies conducted in laboratories, how these types of videos facilitate students' learning experiences and achievement has not been much explored in prior studies, which prevents discussion on how to design such tutoring-style videos. Objectives: In this study, we addressed the limitations reported in studies conducted in authentic classes and attempted to explore students' learning patterns and experiences with tutoring-style videos involving interactions between the instructor and agents. In addition, we attempted to reveal what instructional strategies can be used in tutoring-style videos to enhance students' engagement and achievement. Methods: The study took place in an undergraduate introductory biology class offered at a university located in the Midwest of the United States. The same instructor taught two sections of this course which lasted for 15 weeks during a Spring semester. The two sections were randomly assigned to an experimental group in which participants watched tutoring-style videos and a control group with participants watching traditional videos. The experiment was administered in a module lasting two and a half weeks focused on microbiology. Surveys and tests were conducted to determine whether the tutoring-style videos had a positive impact on student engagement and achievement. In addition, a video content analysis was carried out to elicit insights on how to design effective tutoring-style videos. Results: For students' emotional engagement and perceived usefulness, no significant difference was found between the two groups, even though the descriptive statistics indicated slightly higher scores from the experimental group than the control group on both subscales. In terms of achievement, the participants reported significantly more learned concepts from watching rich-interactions in the tutoring-style videos than the participants in the control group; and they also reported a significantly smaller number of learned concepts from watching one-interactions in the videos than the control group participants. We also found that instruction in the tutoring-style videos used effective strategies for facilitating students' participation and cognitive processing. Conclusions: Our findings indicate that tutoring-style videos can be effective learning materials by allowing students to engage in interactions between instructors and agents in the videos. The more instructional strategies used to facilitate the interactions between the instructor and the agents, the higher engagement can be expected from viewers. Our study would inform the design and development of effective tutoring-style videos.