As mentioned, the twelfth grades students of MA ATQIA divided into two groups, which are the control group (CG) and the experimental group (EG). The aim of conducting the pre-test of Quasi-experiment is to evaluate the level of existing knowledge that respondents gained before getting any treatment as the preparation for post-test. During pre-test evaluation, a TOEFL preparation test focused only in Structure and written expression in section 2 about 20 questions in 20 minutes divided into 10 questions of structure and 10 questions of written expression delivered to all groups of students. A teacher assigned to give a brief introduction about what should students do and face from the test. Then, all groups of students asked to answer questions by themselves with limited of time. The answers and score from the test then recorded.

After conducting the pre-test, the preparation for post-test made. The aim of conducting the post-test Quasi-experiment is to identify a significant difference between two different participants after undergoing a particular treatment. Again, students in (CG) went through the same activity as in the pre-test. Students in (EG) given a mobile application called Duolingo. Duolingo application used to leverage both four major skills in languages learning such as reading, listening, writing, speaking, it also teaches the students about translating and understanding skills. Duolingo includes several subjects, such as modal verbs, plurals, food, animals, which further separated into several lessons that adjusted to the learner’s level. Interactive social compete provided in this application the students in the form of rewarded points. They can collect points according to their success, so they can compete against their friends to see who has the highest amount of points. The students in (EG) were given a limited time to use the Duolingo application for 2 weeks. In this group have experience using smartphone and tablets. As first motivation to the group (EG) to use Duolingo in their daily life of learning languages, the teacher gave a reward for a student with the highest score in Duolingo application. After two weeks passed, post-tests were given to all groups with the same format with pre-test but the different question in the same difficulty. The result of the effectiveness testing will also discuss in the next section.

In order know the motivation response through the application. After post-test conducted. All participants (EG) asked to complete an exit survey with close-ended questions design related to their experience with Duolingo and their motivation through the application. Open-ended questions allow the respondent or participant to give a greater variety of responses, but it makes the statistic analyze difficult because the collected data must be coded or reduced in some manner.  Closed-ended questions which this paper uses are easy to analyze statistically. The researcher limits the responses that participants can give (Jackson, 2009). 

Hypotheses evaluation applied to summarize the relationship between the samples that have used in the testing. For testing the hypotheses, one-way analysis of variance (ANOVA) had applied (Coakes & Ong, 2011). The one-way ANOVA calculation produced ? value. If the calculated ? is higher than 0.05, the null hypothesis accepted. Meanwhile, if the calculated ? value is lesser than 0.05, the null hypothesis rejected (Sekaran & Bougie).