PhD defence S.J.M. (Suzan) van Brussel

Dissertation in short:

The preference for confirmation can have far-reaching consequences in education. For example, it can promote inequality of opportunities in the classroom. This bias also makes it difficult to substantiate choices afterwards. Moreover, it is not easy to teach children to be open to other perspectives when teachers themselves fall into the trap of the so-called 'confirmation bias'. Can we educate student teachers to reduce their confirmation bias and thereby support unbiased thinking?
To reduce confirmation bias, people should consciously use opposing information. This thinking technique is called “consider the opposite”. It is a thinking technique that forces you to actively take a different perspective when making a decision or a judgment than you are first inclined. In several experiments, I investigated whether confirmation bias decreases when students are given both instruction and practice tasks that target this bias. The instruction consists of an explanation of what 'confirmation bias’ is, why it is so important to reduce it and what the thinking technique 'consider the opposite' means. I started with the question whether both these instruction and practice tasks within an educational context would lead to a decrease in confirmation bias. In addition, I investigated the role of feedback on the practice tasks in reducing this type of bias. Finally, I investigated how students best process the instruction and whether this leads to a decrease in confirmation bias. But also whether they can impart a so-called 'citizenship lesson' in an unbiased way, for example about migration or religion.

The research shows that explanation and practice with confirmation bias tasks is necessary to reduce this bias. Compared to the pre-tests, all students on average performed better on the post-tests after they had received instruction and practised tasks. It did not matter under what conditions they were given these instructions and practice tasks. The results of the study therefore provided no evidence that context-specific instructions and extensive feedback lead to an additional reduction in confirmation bias compared to the control conditions. And although expected, a specific processing of the instruction through teaching on video also did not lead to that result. I did find an interesting result regarding the role of feedback. Students who received feedback on their performance on confirmation bias tasks in the practice phase performed better on the post-test than on the pre-test. This was not the case for students who received no feedback. Thus, this shows that receiving feedback has an effect on reducing confirmation bias, regardless of whether the feedback is concise or extensive. There are several reasons why I did not find any pronounced differences between the conditions in the other experiments. One of those reasons may be related to the way confirmation bias has been measured. I used abstract task. The question is to what extent such a task calls upon critical thinking skills that are also needed in everyday life. A recommendation is therefore to make use of tasks that are more in line with the educational context in future research.