Investigating the Link Between Inhibitory Control and Grammatical Ability

The Challenge

Select worked with Dr Paul Ibbotson, lecturer in developmental psychology at the Open University, to explore the link between grammatical ability and inhibitory control within the brain.

Learning the grammar of the English language involves learning a series of basic rules together with the, perhaps surprisingly, many exceptions to those rules. For example, the usual pattern for forming the past tense is to add –ed to a verb (e.g., jump → jumped, show→ showed). However, there are many irregular verbs that do not follow this rule, such as draw → drew and seek → sought. Dr Ibbotson was curious to find out whether a child’s ability to form the correct past tense in irregular verbs was related to a type of inhibitory control that is required to be successful in the well-known Stroop test. In this version of the Stroop test the child is shown a series of pictures of the sun and the moon and instructed to say “moon” when presented with a picture of a sun and vice-versa. The number of “correct” responses given in a set period of time provides a measure of the participant’s ability to inhibit a tendency to say what they see in the picture.

The Approach

In order to test the hypothesis, 81 five-year-olds were given the Stroop test as well as being tested on their irregular past tense formation. Select then helped Dr Ibbotson to analyse the results by constructing a mixed effects model for the probability of making an error in the formation of the past tense of an irregular verb. To account for the possibility that children who are good at the Stroop test simply knew more words or were a little older, which might make them better at the grammatical task, we included age (in months) and a vocabulary score, as well as the Stroop score, as fixed effects in the model. Each participant was tested on past tense formation for a number of different verbs so, in order to account for errors being correlated within participants and within verbs, we included both participant and verb as random effects in the model.

The resulting model demonstrated that, accounting for age and vocabulary score, Stroop score is a statistically significant predictor of verb errors. Further, the model quantifies this effect and provides confidence intervals that indicate the uncertainty in the estimates. For example, the model showed that an increase of 1 in the Stroop score is associated with the odds of a verb error increasing by 25%.

The Value

This study provides evidence that grammatical errors involving past tense formation and errors on the Stroop test are related and suggests that both tests might require a common capacity within the brain to inhibit tempting but incorrect responses.

The study has been published in the peer-reviewed scientific journal PLOS ONE.

The consultant from Select was patient and understanding throughout. The project couldn’t have been completed without their input and it was a pleasure working with them.Dr Paul Ibbotson – Lecturer in Developmental Psychology, The Open University