The verbal information pathway to fear and subsequent causal learning in children

Abstract

Recent research has shown that verbal threat information creates long-term fear cognitions and can create cognitive biases and avoidance in children. However, the impact on future learning is untested. This experiment exposed a non-clinical sample of children (aged 7–9 years) to threat, positive or no information about three novel animals to see the impact on their subsequent causal learning. In this causal learning task, children saw a series of pictures of animals and had to predict on each trial whether there would be a good or bad outcome. They then saw a picture to indicate whether the outcome was good or bad. The probability of each outcome was either .2 or .8. At the end of a block of trials children were also asked to estimate how many trials they thought had concluded with a negative outcome. Results showed that verbal information directly affected the estimate of associative strength between animals and positive and negative outcomes in a causal learning task. These results support theories of fear acquisition that suppose that verbal information affects components of the fear emotion, and suggest possibilities for using information to protect children from acquiring animal fears.

Acknowledgements

This research was supported by ESRC grant R000239591 to AF and Robin Banerjee.

We would like to thank the staff and pupils of Southover C. E. Primary School, Lewes, East Sussex and Uckfield Community Technology College, Uckfield, East Sussex for their co-operation.

Notes

¹The data were analysed in this way because we wanted to track learning throughout the 80 trials. There is some case for analysing the data as a 4 (block: first 10, second 10, third 10, fourth 10 trials)×2 (contingency: .8 or .2)×2 (animal type: negative or no information) mixed ANOVA. However, this is inappropriate because it assumes that each block of 10 trials in the .8 contingency is comparable to the equivalent block for the .2 contingency and this is not the case: in the .2 contingency blocks there has been greater prior experience with the task.