Peer instruction (PI), which can be used in various disciplines, is a student-centered teaching method that engages students during class through structured, frequent questioning and is often facilitated by classroom response systems. To help students resolve misconceptions about a subject matter, they are provided with conceptual questions, which is the fundamental characteristic of any PI class. While there are different forms of the PI method, it is usually organized as follows: First, students individually give answers on the provided multiple-choice questions. Next, they discuss their answers with classmates and then individually give answers again on the same questions. In the last step, the instructors provide explanations on the solutions of the questions. When the literature on PI is examined, it appears that researchers follow different strategies to create effective learning environments, which allow student collaboration. In this study, the purpose was to investigate the effect of PI on preservice mathematics teachers’ (PMTs) achievements on the statistics and probability topics. In Spring 2016 semester, 46 third-year middle grades PMTs, who have been attending to a statistics and probability course, participated in the study. At the beginning of the semester, the PMTs were given an achievement test on the statistics and probability topics. Next, they were randomly divided into two equally achieving groups (22 and 24 students) based on the test results. Next, in each group, two PMTs formed a subgroup. To increase PMTs’ engagements in whole-class discussions and to escalate their motivations towards to the classroom topics, they were allowed to choose their discussion partners. Each week, two groups attended together to a two-hour class, which was taught by one of the authors. Next, both groups attended a two-hour PI session in which the PMTs worked on computers, and the same questions used in both PI sessions. Two different scoring mechanisms used in calculating PMTs’ scores. In one group, PMTs received 40% of the total score if they individually calculated the correct answer; after discussing with the discussion partner, the PMTs received 30% of the total score if they still choose the same correct response; and they received 30% of the total score if discussion partner also choose the correct response in their second response. The scoring for second group was 50%-40%-10%. A total of 10 weeks PI classes are planned for both groups. The data collection process is still ongoing. During the last week, each group will be given the same achievement test. The final test results of the two groups will be compared with pretest scores, and statistical analysis will be performed in order to understand the efficiency of the two different accountability scoring mechanisms on PI.