Teachers' assessment of linguistic difficulty of science tasks for students at risk (TAL)

• Applicant: Professor Dr. Jan Retelsdorf
• Subject Area: General and Domain-Specific Teaching and Learning ; Developmental and Educational Psychology
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 543598930

In the project, two previously unconnected strands of research will be brought together: first, research on judgment bias when teachers assess students and, second, research on the linguistic design of tasks. The linguistic design of tasks will be investigated as one factor that contributes to task difficulty and can thus provide teachers with important information for assessing students according to their performance in these tasks. The planned project will look at students who, according to current large-scale assessments, can be considered as students at risk and have so far been little studied in connection with judgment bias. Based on studies on judgment bias in students of negatively stereotyped groups, it can be assumed that students of this group are judged in a stereotypically negatively biased way compared to students who do not belong to this group. However, following the shifting standards model, it can also be assumed that teachers adjust their assessment standard according to task difficulty (here linguistically manipulated) and do not assess students at risk more negatively or even more positively on difficult tasks. In the project, this will be investigated in the subjects of physics and chemistry in a total of three research questions. First, the question arises as to what extent teachers assess physics and chemistry tasks with varying linguistic complexity as differently difficult. Second, the role of students' belonging to the risk group in teachers' assessment of students in physics and chemistry tasks of varying linguistic complexity will be investigated. Third, it will be investigated whether explicit references to the linguistic complexity of the tasks lead to changes in the performance assessments of students by teachers. A total of six experimental studies will be conducted in the project, which will include conceptual replications of the findings for each research question in a second subject. The first research question will be investigated in a within-subjects design in which tasks of varying linguistic complexity will be randomly assigned and assessed in terms of their difficulty. The second and third research questions will each be implemented with experimental vignette studies in which teachers are asked to assess the performance of experimentally manipulated students regarding their belonging to the risk group. The results can provide important insights into models for the development of teachers' diagnostic judgments and illustrate possible effects of the linguistic design of tasks for subject performance assessments.

• DFG Programme: Research Grants
• Co-Investigators: Privatdozent Dr. Sascha Bernholt; Professor Dr. Hendrik Härtig