Teacher assessments: cognitive biases to keep in mind


Teacher assessments: cognitive biases to keep in mind

Earlier this year, the government announced that, with GCSE and A Level exams cancelled for 2021, students’ grades will be determined by teachers. Schools and teachers will be able to choose how they decide on student grades and what evidence they use for this, such as optional exam board papers, mock exams, coursework and essays. The government stressed wanting to find a system that was ‘fair to every student’.

However, we are all human, and as such are subject to a number of subtle yet powerful thinking biases. These could impact on how we assess student performance. By being aware of them, hopefully we can help avoid some of their pitfalls.

 

Thinking biases to be aware of: Student Characteristics

Some biases that can impact on teachers’ thought processes when grading students are based on characteristics of the students themselves. Here are some examples…

Gender Bias - Unfortunately, traditional gender roles still have an unconscious impact on our views of people's abilities. This study looked at students who achieved similar results in a range of tests. Without knowing the test results, when asked about perceptions of their students’ abilities, teachers rated boys as being less able in reading and girls as being less able in maths.

Racial Bias - This research found that on average, pupils of Black, Caribbean, Pakistani or Bangladeshi ethnicity were more likely to receive lower scores in teacher assessment than white students. Meanwhile, Chinese, Indian, and white-Asian students were more likely to receive higher scores in teacher assessments than white students. So, students’ ethnic origins may also have an implicit impact on teachers’ grading.

Status Bias - Just as a student’s gender or race can impact a teacher’s opinion of their ability, so can their socio-economic status. It’s possible that teachers may unconsciously judge students who come from disadvantaged backgrounds as lower performing at school.

Weight Bias - Research even suggests that teachers may rate students differently depending on their weight. For example, this study found that teachers were more likely to give shorter and heavier students lower grades.

Talent Bias - Finally, whether teachers believe students to be ‘naturally gifted’ or not could bias the grades they give. This research played students music from a musician who was described as a ‘natural’ and a musician who was described as a ‘striver’. The ‘natural’ showed innate ability from an early age, while the ‘striver’ showed motivation and determination. Students rated the ‘natural’ as more talented and more likely to be successful than the ‘striver’. In fact, both pieces of music were from the same musician. This just goes to show that we may be biased towards those with a ‘natural talent’ over those who work hard.

It’s worth noting that in many of the above studies, when teachers were blind to student characteristics, or work was marked by external examiners, these grade differences dependent on gender, race, weight etc. did not occur.

 

Thinking biases to be aware of: Time-based biases

Time-based biases show how we can form our whole opinion of someone on one performance, based on when it occurs. They involve sweeping conclusions or ‘overgeneralisation’ based on only one piece of evidence.

The Halo Effect - The Halo Effect occurs when your overall impression of someone is influenced by either one part of their character or your first opinion of them. For example, if a student gets an A on their first essay, it is easy for teachers to assume that they are high-achieving, and therefore would expect them to achieve similar success in the future. However, this is a sample size of one. There is not enough information to make a reliable prediction.

The Recency Effect - Contrary to the Halo Effect, this describes how your overall impression of someone can be disproportionately affected by their most recent performance. For example, if a student’s most recent performance has been rather low, or rather high, this is more likely to stick in a teacher’s memory and influence the overall grade assigned. As with the Halo effect, it’s not fair at all to base a student’s overall grade for the year based on a single test or performance. And after all the stresses that students have been through recently, we can hardly expect their most recent performance to be a true representation of their overall abilities.

Confirmation Bias - This refers to the idea that people pay more attention to evidence, and perhaps even seek out evidence, that supports ideas they’ve previously agreed with. Paired with other biases, this can be even more dangerous. For example, think of that student who got an A in their first test and whose teacher expects them to continue to do well due to the Halo Effect. With Confirmation Bias, their teacher may be more likely to notice future occasions where they work well, and ignore times when they don’t do so well. This can be quite damaging for students who start off at a lower level of performance, even if they have the potential to make lots of progress. 

 

And there’s more…

Here are some other, quirky thinking biases that teachers should be aware of when deciding on grades…

The Ikea Effect - People tend to place a disproportionately high value on the things they personally create. Translated to teaching, teachers may subconsciously place a higher value on, and thus grade higher, those students who they have spent more time supporting and helping. However, there may be students equally deserving of a high grade, who work well independently and did not need or ask for so much support, but who may get overlooked because of this.

Law of the Instrument - This describes how we often favour a tool or strategy we’re familiar with over other approaches that could work better. This means teachers could get into the habit of relying on a certain method to assess their students. However, different types of assessment suit different students (the obvious one being that some students do better in coursework, while others favour exams). All this means is that it might be beneficial for teachers to think outside the box when deciding on the evidence to use for grading students, exploring all the options rather than defaulting to a particular method.

The unintentional arms race - Let’s picture a scenario. Teachers A, B and C are discussing the grades of their classes. Teachers A and B say that they’ve already graded their students and the class has done well. An A average, for example. Teacher C is still grading their students. Is Teacher C likely to go away and give their students lower grades? Not really. The likelihood is, they’ll be swayed to push their students’ grades higher too, using their colleagues’ grades as an anchor. In this way, grading can become an unintentional arms race, between teachers trying to achieve the highest grades, and between schools too.

 

FINAL THOUGHTS

All we can do is grades students as fairly and honestly as we can. Hopefully, being aware of these sort of biases can help do that. And if you think none of these biases apply to you, then chances are you are suffering from the Bias Blind Spot, which refers to the belief that you are less likely to have these biases than others.

The truth is that everyone is susceptible to thinking biases to some extent. It’s just important that teachers are aware of the implicit biases they might have, to ensure these don’t sway student grades, and to make sure the grading system this year is as fair as possible.

Rosenshine's 10 Principles of Instruction online teacher CPD on the InnerDrive Online Academy