The influential "Common Core State Standards for Mathematics" (CCSSM) expect students to start statistics learning during middle grades. Thus teacher education and professional development programs are advised to help preservice and in-service teachers increase their knowledge and confidence to teach statistics. Although existing self-efficacy instruments used in statistics education focus on students, the Self-Efficacy to Teach Statistics (SETS) instrument measures a teacher's efficacy to teach key CCSSM statistical topics. Using the results from a sample of "n" = 309 participants enrolled in a mathematics education or introductory statistics course, SETS scores were validated for use with middle grades preservice teachers to differentiate levels of self-efficacy to teach statistics. Confirmatory factor analysis using the Multidimensional Random Coefficient Multinomial Logit Model supports the use of two dimensions, which exhibit adequate reliabilities and correspond to the first two levels of the "Guidelines for Assessment and Instruction in Statistics Education" adopted by the American Statistical Association. Item and rating scale analyses indicate that the items and the six-category scale perform as intended. These indicators suggest that the SETS instrument may be appropriate for measuring preservice teacher levels of self-efficacy to teach statistics.