Background: Observational learning is an effective pedagogical approach that can be used as a method to improve motor skill acquisition and also a useful instrument to promote psychological variables such as self-efficacy by emphasizing the motivational aspect. Although, little research is known about the model type and the observation conditions which will optimize learning. Aim: This study aimed to clarify the effect of using a combination of ideal model-observation and self-observation on self-efficacy and learning the forehand service of table tennis. Materials and Methods: Forty females were assigned to one of four experimental groups. All groups received the same instructions concerning how to perform the task, filled out a self-efficacy questionnaire, and then performed 10 pre-test trials. The acquisition phase included six sessions. The retention test was done 72 hours after the last acquisition session. Mixed ANOVA with repeated measures and Bonferroni post hoc tests were conducted, and one-way ANOVA was used to determine group significant differences. Results: The results indicated that observing the combination of skilled and learning models led to the highest self-efficacy for learning, compare with other experimental models. Also, significant learning of the task was shown in the acquisition phase in which the performance of all three experimental groups was clearly superior to the control group. Although in this phase there was no significant difference between the performances of two groups of combined skilled-learning and skilled. The results of the delayed retention test indicated that the performance scores of the combined skilled-learning group were significantly higher than the other three groups. Also, in this phase, the difference between the combined skilled self-observation and control or skilled groups was not significant. Conclusion: Demonstrating a skilled model is the most popular type of observational learning among the instructors and physical education teachers, it is suggested that adding a learning model to that leads to better performance and enhancing the self-efficacy.
