Fail #29: Max und Murat, diese beiden...

„Max“ und „Murat“ waren die beiden Namen, anhand derer zwei Psychologen von der Uni Mannheim testeten, ob angehende Lehrkräfte sich bei der Notengebung von ethnischen Stereotypen leiten ließen. Die Antwort ist Ja.

Bonefeld, M. & Dickhäuser, O. (2018). (Biased) Grading of Students’ Performance: Students’ Names, Performance Level, and Implicit Attitudes. Frontiers in Psychology 9. https://www.frontiersin.org/articles/10.3389/fpsyg.2018.00481/pdf.

Abstract: Stereotypen fließen in die Noten ein

Ein Diktat eines Drittklässlers wird von angehenden Lehrkräften anders bewertet, wenn man nur den Namen des angeblichen Verfassers ändert. „Murat“ wird dabei signifikant schlechter bewertet als „Max“.

Biases in pre-service teachers’ evaluations of students’ performance may arise due to stereotypes (e.g., the assumption that students with a migrant background have lower potential). This study examines the effects of a migrant background, performance level, and implicit attitudes toward individuals with a migrant background on performance assessment (assigned grades and number of errors counted in a dictation). Pre-service teachers (N = 203) graded the performance of a student who appeared to have a migrant background statistically significantly worse than that of a student without a migrant background. The differences were more pronounced when the performance level was low and when the pre-service teachers held relatively positive implicit attitudes toward individuals with a migrant background. Interestingly, only performance level had an effect on the number of counted errors. Our results support the assumption that pre-service teachers exhibit bias when grading students with a migrant background in a third-grade level dictation assignment. (Bonefeld und Dickhäuser 2018, S. 1)

Self fulfilling prophecy

Namen erzeugen in uns Lehrern ein Bild von einem Menschen. Dieses Bild trägt Eigenschaften in sich, die über den Namen hinausgehen. Aufgrund dieser angenommenen Eigenschaften erwarten wir bestimmte Leistungen oder ein bestimmtes Verhalten. Diese Erwartungen prägen unsere Interaktion mit den Schülern und transportieren sie auf die Schüler, die sie auch häufig übernehmen.

On a more subtle level, teachers can even change the way students perform because of the wide-known phenomenon of the self-fulfilling prophecy (Rosenthal and Jacobson, 1968; Jussim and Harber, 2005). Research on the self-fulfilling prophecy has demonstrated that teacher expectations (which may be based on students’ migrant background) can influence students’ behavior and their achievement. Besides these effects on students’ achievement, which may result from differences in the teacher–student interaction, information about a student’s background may also influence teachers’ perception of the student’s achievement, even though objectively there might be no differences between students. The present study is concerned with this latter effect and tests potential differences in assigned grades which depend on students’ migrant background. (Bonefeld und Dickhäuser 2018, S. 2)

Fehlurteile

Lehrer zeigen häufig die Einschätzung, dass Schüler mit Migrationshintergrund prinzipiell weniger leistungsfähig sind als „richtige“ Deutsche. Demgemäß geben sie ihnen auch bei gleicher Leistung schlechtere Noten.

Judgmental errors could arise as a result of specific teacher attitudes toward the future performance level of students from a specific group. Analyses of cultural stereotypes in Germany have shown that Germans with a migrant background are associated less strongly with performance and success-related attributes by other Germans compared to Germans without a migrant background (Kahraman and Knoblich, 2000; Froehlich et al., 2016). Furthermore, teachers tend to judge students with a migrant background as less academically able than students without a migrant background (Glock and Krolak-Schwerdt, 2013). This could be particularly important for the assessment of students with a migrant background because negative stereotypes about their performance could lead to negative attitudes about their performance and may influence the grading process. Due to negative stereotypes regarding the lower performance potential of students with a migrant background, teachers could have additional information about the student concerning characteristics that do not exist or, at least, that cannot be observed (Taylor, 1981; Walton and Spencer, 2009; van Ewijk, 2011). (Bonefeld und Dickhäuser 2018, S. 2)

Implizite Annahmen und Haltungen

Unsere Stereotypen beeinflussen unsere Informationsverarbeitung und lenken unsere Aufmerksamkeit. Oft werden sie automatisch aktiviert und entgehen unserer eigenen Aufmerksamkeit oder Selbstkontrolle.

Depending on the intensity of associations, attitudes are stable and important in terms of guiding behavior (Krosnick and Petty, 1995). Additionally, they can influence information processing and guide attention (Roskos-Ewoldsen and Fazio, 1992). In this case, implicit attitudes are possibly of great importance because they are automatic evaluations (Gawronski and Bodenhausen, 2006). They can be automatically activated and are spontaneous evaluations (van den Bergh et al., 2010). The automatic activation of implicit attitudes explains why they are often beyond our control and why persons are often not even consciously aware of them (Gawronski and Bodenhausen, 2006). (Bonefeld und Dickhäuser 2018, S. 3)

Folgerungen für die Lehrerausbildung

Man muss die angehenden Lehrkräfte dazu anleiten, sich objektive Maßstäbe zurecht zu legen, anhand derer sie Schülerarbeiten beurteilen und benoten.
Gerade weil die stereotypisierenden Prozesse in der Regel unbewusst verlaufen, ist es wichtig, die Mechanismen hinter dieser Dynamik zu verstehen. Dazu verweise ich auf den Aufsatz Menschenbild und Schulstruktur.

Our results have implications for teacher education. It should be kept in mind that, other than the variation of performance level, the dictation differed only with regard to the first name of the student. The information provided about the performance in the dictation was exactly the same for the different groups. Nevertheless, the dictation was graded differently depending on the background of the student. We can therefore infer that the biased judgments resulted from the variation in names. This highlights the importance of focusing on elucidation and teaching reasons for judgmental biases as well as strategies for avoiding biases in teacher education. Our results suggest that the difference in the assessment of students with and without a migrant background can be found in less rule-based judgments, whereas rule-based judgments are rather accurate, or at least do not differ depending on the background of the student. Based on this, we can clearly recommend incorporating fixed grading rules into teacher education.

Pre-service and in-service teachers should be encouraged to prepare grading schemes on the basis of rule-based judgments (which they can clearly deduce) and to award grades according to these fixed standards. For example, when grading a dictation, they could make grading dependent on the number of errors found. Broadly speaking, (pre-service) teachers should define which rule-based judgments compose their less rule-based judgment in terms of overall judgment. Then they should decide on which of these rule-based judgments their judgment (e.g., grade) is based. By means of this decision, an integrative rule for awarding certain grades could be made. In this way, biases could be reduced.
From a practical perspective, judgment errors are of great importance when one considers that school grades in elementary school influence future school careers and also have a great influence on later educational access, such as admission to different courses or different careers (Glock and Krolak-Schwerdt, 2013). Thus, it is necessary that teacher judgments are as accurate as possible to ensure the best possibilities for all students.
Apart from this, it is important to say that the processes that lead to judgmental biases, such as using stereotypes as information when assessing performance, are often unconscious and not intentional. The use of stereotypical attitudes can save cognitive capacity (Macrae and Bodenhausen, 2000) and is therefore efficient. Nevertheless, precisely because these processes are often unconscious and unintentional, it is important to determine the mechanisms behind these judgment processes which can have a considerable influence on students. (Bonefeld und Dickhäuser 2018, S. 11)

Die Unsicherheiten von Lehrkräften, bei Schülerarbeiten zu objektiven und nicht stereotypabhängigen Beurteilungen zu kommen, waren schon öfters Thema in diesem Blog:

Faktencheck #5: Diagnosekompetenz von Lehrkräften