Aldo Bazán-Ramírez ^1* Walter Capa ² Roberto Chávez-Nava ² Mónica C Dávila- Navarro ³ HOMERO ANGO-AGUILAR ⁴ Edmundo Hervias-Guerra ² Catalina Bello-Vidal ²

• ¹ Universidad Nacional José María Arguedas, Andahuaylas, Peru
• ² Universidad Nacional Federico Villarreal, Lima, Peru
• ³ Instituto Tecnológico de Sonora (ITSON), Ciudad Obregón, Sonora2, Mexico
• ⁴ Universidad Nacional de San Cristobal de Huamanga, Ayacucho, Peru

The evaluation of teaching performance in didactic interactions has generally been approached from the student's perspective, with little literature on evaluation based on the teacher's own perspective. Having an instrument that allows self-evaluation of teaching performance during didactic interactions will contribute to improving the quality of teaching and learning at the university level. Objectives: the validity and reliability evidence for the scores of the self-evaluation questionnaire of the teacher's didactic performance are examined. The self-assessment of didactic performance is analyzed and the general didactic performance of the teacher and of his interactive episodes is contrasted according to some socio-academic characteristics. Method: Methodological, instrumental, descriptive, comparative and predictive cross-sectional study. A total of 203 teachers participated, of whom 64 were professors from a public university in northwestern Mexico, and 139 professors from two public universities in Peru. Results: Satisfactory evidence of content-based validity was found, Aiken's V estimates were higher than 0.70 for all items in the criteria: clarity, pertinence and relevance. As for the evidence of validity based on the internal structure of the construct, satisfactory results were found with the confirmatory factor analysis strategy for a multidimensional model of seven oblique factors (CFI = 0.99, TLI = 0.98, RMSEA = 0.05, SRMR = 0.06) and a second-order factor model made up of two general factors and seven specific first-order factors (CFI = 0.98, TLI = 0.98, RMSEA = 0.06, SRMR = 0.07). Evidence of convergent and discriminant validity was also acceptable. The reliability for the overall score of the questionnaire as well as for the interactive episodes evidenced McDonald's ordinal alpha and omega coefficients ≥ 0.94 and H coefficient ≥ 0.95. With respect to the comparative analyses according to the professional training discipline and the sex factor, differences were found for a small effect size (d > 0.20) in favor of education teachers and women both in general didactic performance and in the teaching and formative evaluation factors. Likewise, being an education teacher and a woman is more than twice as likely to have optimal teaching performance. As for the nationality of the teachers, no significant differences were found.