Assessing educational outcomes is a matter of great importance for everyone in a society that relies primarily on public schools to shape the next generation of citizens. Given the vital nature of this task, it is imperative that the researchers who undertake it have an informed understanding of the measures used to evaluate educational outcomes. This dissertation contributes to that task by analyzing grades as a potential complement or alternative to standardized test scores as a measure of educational outcomes.

The research is based on a theoretical framework that emphasizes the importance of both instruction and socialization as goals of the education system. Grades are theorized to assess elements of both of these goals because they account for academic achievement and student behavior, while standardized test scores primarily reflect knowledge and instruction. Therefore, grades are expected to provide a more comprehensive evaluation of student achievement than standardized test scores. In addition, grades function as feedback to students that can affect their future outcomes. Given the comprehensive nature of grades and their role in shaping students' later outcomes, the theory establishes the necessity of developing a thorough conceptualization of grades.

Both quantitative and qualitative data are utilized to test elements of the theory. The analyses examine several aspects of grades and provide information that contributes to a more complete understanding of grades than has been available in the past. The research confirms that grades account for students' behaviors and achievement orientations while standardized test scores do not offer a thorough assessment of these characteristics of students. Observations clarify the importance of behavior for mediating opportunities to learn within a classroom setting and support the inclusion of behavioral assessments in grades. The analyses also highlight the role of teachers in interpreting and evaluating students' behaviors in grades. In addition, teachers are found to mediate the impact of students' behavior on their own academic achievement and on classmates' opportunities to learn.

The models estimated in this study demonstrate that grades can serve as an appropriate supplement to standardized test scores or act on their own as measures of educational outcomes. Their viability as outcomes in analyses is determined based on the variables used to predict achievement and the information in which researchers are interested. Because grades act as a proxy for behavior and achievement orientation and also act as feedback to shape students' academic self-concepts, they are valuable predictors of outcomes that are influenced by those characteristics of students. In many situations, using both grades and standardized test scores to predict outcomes is appropriate because each accounts for different aspects of educational achievement.

Overall, the study establishes the usefulness of grades as measures of educational outcomes. It accomplishes this goal by generating a theoretical basis for understanding the meaning and function of grades. The research also sets a foundation for understanding what grades represent in analyses and how they can be appropriately utilized and interpreted in research.