What does the research say about bias and stereotyping in the classroom?
17 February 2017
Author: Jonathan Haslam
“Atticus, you must be wrong….”
“Well, most folks seem to think they’re right and you’re wrong….”
Harper Lee, To Kill A Mockingbird
A recent study looking at when gender stereotypes develop received a lot of coverage in the press. It’s a reminder that, no matter how impartial or independent we think we are, all of us have biases and prejudices. Even the widespread coverage of that study was accused of bias.
There are many studies that show that there is bias and stereotyping in the classroom. To take a few examples:
- In the US, non-African-American teachers had lower expectations for African-American students, particularly for male students and maths teachers.
- Schools in the UK that send relatively more girls to do A-level physics also had a smaller gender imbalance in other subjects.
- Teachers in Germany gave students a higher rating if their personality was similar to the teacher’s, although this relationship broke down when teachers were asked to make task-specific judgements.
- Teachers in the Netherlands had higher expectations for pupils in high-performing classes or classes with only a small proportion of pupils from underprivileged families. Teacher expectations of Turkish, Moroccan, and other ethnic minority pupils with low-educated parents were larger than the average bias found when comparing teacher expectations with pupil performance.
- In Israel, blind assessment of tests showed that teachers over-marked boys, and this led to a significant positive effect on male academic achievement over girls.
- In the UK, children from low-income families, boys, SEN pupils, and EAL pupils were all less likely to be judged “above average” at reading by their teacher, despite performing equivalently to their peers in tests. For maths, girls and Black Caribbean pupils less likely to be rated highly despite their test performance.
Perhaps I have shown bias in collecting these examples, but they do seem to be a regular feature of the research, and point consistently in one direction.
So how can such biases be avoided? Discussion of this topic and its possible solutions often includes ideas such as:
- Awareness – being aware, and making colleagues aware, of research such as this, and individuals’ tendencies towards bias
- Avoiding stereotypes – don’t use language or behaviour that stereotypes children by gender, race, etc, etc.
- Promote integration between boys, girls, etc, mixing up groups and encouraging them to sit and work together.
However, while these strategies are no doubt worthwhile, I wonder whether they are sufficient. After all, very few teachers would say that they are intentionally biased against individual students, or groups of students. The bias that takes place is mostly unconscious, and therefore more difficult to shift. I wonder if it’s possible to take some lessons from the way that researchers attempt to avoid bias in research studies. This would suggest ideas such as:
- Randomisation – use a random method (like the lollipop technique) to choose pupils for tasks, questions, etc.
- Blinding – for assessments/tests, is it possible to blind yourself to the pupils that have completed the assessment (clearly easier for some subjects than others)? Alternatively, are there third parties who could mark anonymised work?
- Standardised assessments – as in some of these studies, use standardised, external assessments to measure student achievement. Then compare/moderate with teacher assessments.
How would you address issues of bias and stereotyping in the classroom?
What support might be needed across school to address these issues?
How might you measure whether you were successful in reducing bias?Posted on 17 February 2017
Posted in: Blog