How Teacher Expectations Shape Students’ Experience in the Classroom

More recently, in 1968, the Pygmalion effect also came to be known as the Rosenthal effect, named after Robert Rosenthal, one of the researchers who demonstrated the impact of this psychological phenomenon in education. In the classroom, this phenomenon often operates through teacher expectations, affecting student performance.1 

According to research by Rosenthal and his colleague Lenore Jacobson, there are four pathways through which teachers’ behavior can impact student performance: 

• Climate (whether the classroom environment is emotionally cold or warm and welcoming);
• Input (whether students have access to low- or high-quality study materials);
• Output (whether students have opportunities to respond and engage in the classroom); and 
• Feedback (the extent to which students received detailed, personalized advice on how to improve). 

The higher the expectations a teacher has for a student, the warmer their approach will be (e.g. more nodding and smiling), the higher the quality of materials they’ll provide, the more opportunities they’ll give to the student to participate, and the more specific and personalized their feedback will be. 

Since Rosenthal and Jacobson published their landmark study, research in this area has flourished.2 One notable review3 by the Center for American Progress analyzed data from the Educational Longitudinal Study, which tracked the progression of a group of American 10th-graders from 2002 to 2012. The review found that all else being equal, students who had teachers with higher expectations were more than three times more likely to graduate from college, compared to students whose teachers had low expectations for them. In other words, teacher expectations were highly predictive of college graduation rates—more so than other factors such as race and academic efforts.

It is important to note that this kind of research has methodological limitations. Even though the above studies have controlled for variables like race and socioeconomic status, it is difficult to entirely separate the impact of teacher expectations from that of societal factors that make certain students less likely to, say, go to college. If a teacher works mostly with kids from disadvantaged backgrounds, who face more challenges and have been historically less likely to achieve academic success, they will probably approach their students with lower expectations, but that doesn’t mean those expectations are the only thing holding those students back. 

Still, there is good evidence that there is a relationship between teacher expectations and student performance. Seeking to refine these associations, researchers have delved into exploring the effects of teacher stereotypes and expectations. For example, a recent World Bank study4 found strong evidence that the stereotypes teachers hold related to social class have an influence on their assessment of their students’ scholastic aptitude and behavior. Such studies show that biases and stereotypes are critical in forming teacher expectations.

Expectations are the lenses through which we see the world, reflecting our biases and mental models. If we consistently see men in STEM careers and leadership positions, then our brains, ceaselessly looking for patterns to help us interpret the world, implicitly associate the two. If we see an attractive person enter the room to give a presentation, we are likely to judge that person and their presentation more favorably, merely because of the halo of their external appearance. Such biases are problematic because they impact our judgments of and behaviors toward different social groups, and therefore can lead to real-world harm. 

In the context of education and training, these biases are likely to predispose the teacher to act (un)favorably towards their students. 

The bad news: research has shown that becoming aware of one’s own biases is not sufficient to overcome them. This phenomenon has been referred to as the “G. I. Joe fallacy,”5 courtesy of the famous cartoon slogan “Now you know, and knowing is half the battle!”

A recent Harvard Business School working paper6 took up the challenge of categorizing biases based on their cognitive architecture—that is, the extent to which they are insulated from information that resides in other cognitive domains. They found that some of the most impenetrable biases (i.e. those that are difficult or impossible to break through conscious reflection) are visual illusions, implicit associations (such as gender and racial bias), halo effects (e.g. favoring someone based on their outward attractiveness), strong emotional states (due to the hot/cold empathy gap), the effects of situational factors (such as the weather or finding a coin on the street), and loss aversion. 

The authors call this category of biases “encapsulated” biases. Much like with visual illusions, where there is a persistent miscommunication between the eye and the brain, encapsulated biases prevent us from seeing the objective picture even after engaging our System 2 thinking. Encapsulated biases cannot be eliminated by knowing about them, because the representations or feelings that automatically affect our choices in these cases are invisible to us during the moment of decision-making. 

Implicit associations are a form of encapsulated bias. If a teacher implicitly associates some groups with higher or lower chances of academic success, this association may surface in their behavior, without their realizing it. This could end up contributing to unequal outcomes for their students, ultimately reinforcing the original stereotype and triggering the Pygmalion effect.

Consider an introductory computer science course taught by a male professor who has no overt prejudices and truly believes (on a conscious level) that all students, regardless of socio-demographic background, are capable of mastering the material. Without realizing it, however, the professor may behave in ways that trigger feelings of self-doubt among the female students. For example, he may signal lower expectations for female students by excessively praising females for giving correct answers in class.7

Despite the professor’s good intentions, this kind of reception can be perceived as demeaning, and the resultant stress may cause female students to lose interest in the course, put forth less effort, and ultimately choose not to major in computer science. (This also relates to the phenomenon known as stereotype threat, wherein people are more likely to perform poorly if they’re aware of task-relevant negative stereotypes about their social group.)

The good news is that humans are not born with an innate bias towards any particular group. Rather, we gain these prejudices through exposure to certain environments and experiences, which are both consciously and unconsciously stored in our brains and later influence our automatic decisions.8

That being said, since the environment in which we live and operate is rife with bias, we need to work purposefully in order to overcome such encapsulated biases by altering the availability of certain representations.

Teachers can work on four levels to fight these biases.

1. Self-reflection

This level starts with increasing self-awareness—for instance, through taking the Implicit Association Test (IAT) for various biases. Although this test is not a perfect instrument, the average result of multiple IATs can provide a measure of unconscious biases for an individual, and can detect attitudes that participants would consciously try to hide or might not even be aware of. After taking these tests, the teacher can then reflect on counter-stereotypical examples9 and use the inversion technique for each of these examples. For instance, if the teacher receives a result that suggests they hold some biases about gender, they can consciously try to think of some examples of women who became successful in male-dominated fields, and the traits that helped them to succeed. As an exercise performed regularly over time, it can help the brain to recreate representations in the environment and balance existing biases.

Along the same lines, teachers can purposefully reflect on common examples of biases found in studies,10 such as:

• Assuming that students from certain backgrounds or social groups have different levels of intellectual ability and/or ambition, and are satisfied with lower achievement levels. 
• Assuming that students who are affiliated with a particular identity group want to act as experts on issues related to that group.

2. Inclusivity

Teachers can work deliberately to develop an inclusive classroom climate and inclusive teaching practices, by getting to know all students equally and preparing teaching materials that include diverse examples for all students to relate to. 

3. Feedback

Teachers can solicit feedback from outside observers such as fellow colleagues, as well as from their students.

4. Personal growth

As research grows, teachers can make time to keep up with the latest findings and good practices. They can also participate in professional development programs that decrease prejudice and encourage empathy-centered discipline.

Of course, fighting bias in the classroom is not up to teachers alone. The entire educational system and associated institutions can help by lending support, allocating the necessary time and resources to teachers, and investing in staff diversity. It’s a long road, but the stakes are high.

From a bedtime story about a sculptor to a life lesson. What’s your favorite story and what could it teach us?