Measuring Preservice Teachers’ Beliefs About Grading-as-Feedback: Calibrating Novices in a World of Mixed Messages

• In Event: Saturday Roundtable Session 9:45 am
  In Roundtable Session: The Teacher Side of Assessment: Exploring Biases, Beliefs, and Attitudes (Table 11)

Abstract

Objectives and Theoretical Framework
The purpose of this study is to measure preservice teachers’ beliefs about the construct “grading as feedback” (GAF). Teachers’ beliefs about classroom assessment likely differ by experience levels, exposure to curricula, and prior knowledge (Author, 2017). Scholars have demonstrated beliefs influence actions teachers take in the classroom (Rubie-Davies, Flint, & McDonald, 2011; Woolfolk Hoy, Davis & Pape, 2006). We know teachers’ beliefs guide their decision-making processes and practices regarding grading (Brookhart, 1994) and feedback (Brown, Harris, & Harnett, 2012; Hoy & Davis, 2006). The aim is to develop a valid and reliable measure of preservice teachers’ beliefs regarding the GAF construct.
The theoretical frameworks guiding the study emphasize expected variation among novice teachers. These frameworks emphasize:
1. the knowledge base on beginning teachers and their schema, prior knowledge, p-prims (See e.g., diSessa, 1983), and misconceptions (See e.g., Levin, 2014; Simmons et al., 1999);
2. the modeling of Teacher Learning Progressions (TLPs) in formative assessment practices (Author & Author, 2019, 2020/21); and
3. the principles of construct modeling (Wilson, 2005), which can be used to reliably and validly differentiate novices to better support development and growth.
Our constructivist perspective on teaching teachers is that novices differ from experts in myriad, significant ways (Borko & Putnam, 1996; Livingston & Borko, 1989).
Methodology and Data Source
We administered the survey in the context of a required post-baccalaureate course within a three-semester-long teacher preparation Master's program. All respondents took the survey on the first day of class in the first semester. IRT analysis of the survey employed a partial credit model (Wright & Masters, 1982). The qualitative data was analyzed using deductive coding processes (Miles, Huberman, & Saldana, 2014). Initial coding employed Wiggins’s (2012) criteria.
The sample size (n=42) was sufficient for a pilot calibration study. The items design used a mix of item formats (see Figure 1). The outcome space design was Likert-style with three fixed-choice options: true, false, and not sure. Respondents were prompted, “Explain your answer,” which yielded constructed responses for all items.
Results
Descriptive statistics from the GAF survey indicate three roughly even groupings of pre-service teacher beliefs: those who believe grades are feedback, those who do not believe grades are feedback, and those who are unsure if grades are feedback. Figure 2 presents the results for all items. Thematic analyses of open-ended prompts show that true/false/not sure explanations by preservice teachers were not well grounded in research and relied on respondents’ experiences rather than on declarative or procedural knowledge. Table 1 shows coded misconceptions.
IRT analysis of the validity and reliability properties of the survey showed an acceptable item fit for a unidimensional scaling approach. The person separation reliability (r = .79) was sufficient for scale calibration.
Significance
Our study addresses the need for understanding the perceptions, beliefs, and attitudes preservice teachers have related to “grades as feedback.” Working with novice teachers on improving their classroom assessment practice requires addressing their prior knowledge, including potential misconceptions. Findings can aid teacher educators in seeking diagnostic information on novice beliefs about grading and feedback.

Authors

• Brent Duckor, San José State University

• Carrie Holmberg, San José State University