Study attributes teacher bias to class and origin

Teacher bias measured by a report with 1,717 future teachers

In correcting an essay, future teachers would help migrant-origin students more than native ones. In their forecast about their future, they have less faith in them. This contradiction — more help today, less expectations tomorrow — is the finding that a study published by the European Commission attributes to teachers, and the trigger of a discussion where pedagogy has ended up mixed with the curriculum of the director of the medium that published it.

What the study measures and why the sample matters

The report — titled 'Teacher bias in assessments according to the status attributed to the student: a factorial experiment on discrimination and cultural reproduction' — relies on a survey of 1,717 students of teaching from 20 Spanish universities, 15 public and five private. Each participant was given the essay of an imaginary student and asked to evaluate spelling and grammar, as well as to anticipate the academic future of that student.

The authors chose students who do not yet practice so that the result was as conservative as possible: without years of classroom behind them, they have not yet absorbed the prejudices of the veterans. The sample is one of the largest assembled to experimentally measure teacher bias, and the conclusion does not allow nuances: the surname, the profession of the parents and the cultural capital that appears in a text change the grade.

The paradox: they help the migrant more and expect less from him

In the short term, future teachers would tend to help more the children of migrants — the study analyzes only children of Segarro born in Spain —, to girls and to those who exhibit a higher cultural capital. One of the authors, Irene Pañeda-Fernández, describes the mechanism: it is suspected that these students do not speak Spanish at home, so they are compensated in the classroom.

In the long term, the arrow turns. Being the child of a migrant or coming from a lower class makes the teacher put less hope in that student, and that expectation conditions the treatment, the involvement with the family and, ultimately, the performance. The report calls it cultural reproduction; in the jargon of the classroom it is known as the Pygmalion effect.

Monet scores higher than a TV reality show

There is the uncomfortable data. If in the essay the student mentions a Monet painting, he gets a better grade than if he mentions a reality TV program. 'This is worrying if we take into account that they were only asked to evaluate spelling and grammar,' says the researcher. Evaluators would confuse cultural capital with the objective quality of the test, and bad behavior penalizes equally even if it is not part of what is written: a student perceived as disruptive drags a lower grade in an exercise that only judged the form.

Is it racist to think that the son of a Segarro will have worse results?

It is the question that the research itself asks without closing it completely. 'It is difficult to affirm it, but we do see a prejudice. At a social level, that origin is associated with the fact of losing one's way and it is very negative,' admits Pañeda-Fernández. In Spain, a quarter of minors are now descendants of migrants, according to the Demographic Observatory of CEU, and the 32% of babies born in 2021 had a foreign father or mother.

Here opens the rift. One current holds that the differences in grades respond to causes that are not suspicious — the language spoken at home, the reinforcement classes that some parents can pay and others cannot, the educational level of the family — and that loading it all on bias is a way of not looking squarely at the problem. The reply is that the experiment dismantles precisely that: the evaluators did not see the students, only a text, and even so the grade changed.

The attack on the messenger and the defense of the classroom

Much of the noise was not against the report, but against who brought it to the front page. The director of the medium that published the article is reproached for having directed a newspaper without having finished his journalism degree, with the argument that his trajectory is explained more by his surname than by merit. The criticism has its twist: he is accused of being classist who is pointed out as an example of the classism he denounces.

On the other side is the teaching sector, which comes out to defend itself with classroom scenes. Teachers in practice assure that grades are taken by students, not them, and describe centers at the limit where the problem is not bias but conflict. There also appears the inverse suspicion — that historically grades have been inflated for certain female students so that they could enter the career they wanted — and a repeated complaint: the rule that allows adapting the curriculum to the student's abilities until cutting it leaves the door open to the given pass.

The gender that does not appear in headlines

Regarding girls, the study points to a favoritism that does not yet respond to the labor gap — those evaluated will be primary school teachers — and reminds us that they obtain better academic results than boys. There jumps the asymmetry that many underline: when the differential favors female students, it is called merit; when it harms boys, it is nothing. The research itself admits that the first gender oppressions are not yet translated into grade differences at that stage.

And a proposal that hovers over the conversation, simple and cheap: correct exams without name. Difficult to discuss, almost impossible to apply without breaking other things.

From the grade to the social destiny

The report is read in the light of a broader diagnosis: education as a broken social elevator, with lives marked by postal code from the cradle. The bias towards low expectations, according to the authors, is induced by popular beliefs, data on school failure that do not have to be perpetuated and stereotypes without more. Some remember that in the United States it has been responded with quotas and affirmative action, a path that is perceived as a transfer of the problem from correction to access.

With 1,717 future teachers, 20 universities and a controlled experiment, bias is measured. What is not measured is how much of that difference in grades explains prejudice and how much explains that some parents can pay reinforcement and others cannot. Does the school sustain that question without breaking from within?