Based on the cognitive-affective theory, the present study designed a science inquiry learning model, "predict-observe-explain" (POE), and implemented it in an app called "WhyWhy" to examine the effectiveness of students' science inquiry learning practice. To understand how POE can affect the cognitive-affective learning process, as well as the learning progress, a pretest and a posttest were given to 152 grade 5 elementary school students. The students practiced WhyWhy during six sessions over 6 weeks, and data related to interest in learning science (ILS), cognitive anxiety (CA), and extraneous cognitive load (ECL) were collected and analyzed through confirmatory factor analysis with structure equation modeling. The results showed that students with high ILS have low CA and ECL. In addition, the results also indicated that students with a high level of self-confidence enhancement showed significant improvement in the posttest. The implications of this study suggest that by using technology-enhanced science learning, the POE model is a practical approach to motivate students to learn.