Home > Research > Publications & Outputs > Attentional processes involved in the developme...

Electronic data

  • 2020iversenphd

    Final published version, 3.58 MB, PDF document

    Available under license: CC BY-NC-ND: Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License

Text available via DOI:

View graph of relations

Attentional processes involved in the development of set shifting and restricted and repetitive behaviours in young typically developing children and children with autism.

Research output: ThesisDoctoral Thesis

Publication date1/04/2020
Number of pages275
Awarding Institution
Award date12/02/2020
  • Lancaster University
<mark>Original language</mark>English


It is widely known that young typically developing (TD) children and many individuals with autism (ASD) perform poorly on executive function (EF) tasks. In pre- schoolers, these skills develop rapidly between the ages of 3 and 4 and are often measured through the Dimensional Change Card Sort (DCCS) task. This is also around the same time that restricted and repetitive behaviours (RRBs), a diagnostic characteristic for ASD, peak in typical development. These findings have led to an increasing interest in the relationship between EF skills and RRBs, but the studies have produced mixed findings. To our knowledge no meta-analyses have been carried out to examine the relationship between RRB scores and performance on EF measures. Moreover, no studies have yet pinpointed what it is about these skills or behaviours that make them associate so highly. This thesis therefore presents a series of experiments that firstly aim to examine the strength of the relationship between the behaviours and performance on EF tasks. Secondly, examine the relationship between different sub-groups of RRBs and various set shifting processes, such as the ability to shift away from dominant stimuli, and the ability to activate previously ignored stimuli. Finally, examine training implications for the skills by assessing if a short-term training program can improve the scores and possibly have an impact on the behaviours.

In chapter 1, we conduct three meta-analyses to examine the relationship between
RRB scores and performance on set shifting and inhibitory control tasks, as well as scores of EF parental report measures. We found significant correlations of medium strength in all three analyses. Moreover, whereas age and the type of RRB scale moderated the inhibitory control and parental report results; diagnosis, testing modality, and type of EF measure did not have an overall impact on the results. These findings suggest that the EF hypothesis may play a crucial role in the development of RRBs, or vice versa. Future research should focus on disentangling different EF measures to pinpoint what it is about the tasks that make them associate with the behaviours.

In chapter 2, the focus is on set shifting, the individual EF skill that showed the strongest association with RRBs. Our aim in this chapter is to uncover what causes the correlations between the behaviours, and performance on the Wisconsin Card Sort Task, (WCST) but not the much simpler DCCS. We review the main theoretical frameworks that have attempted to explain two types of errors; the ability to shift away from dominant stimuli and the ability to activate previously irrelevant stimuli. Whereas research on the DCCS suggests that children find both errors difficult, research on the WCST suggests that adults find it more difficult to activate previously irrelevant responses. We argue that the different findings are not evidence for different developmental trajectories in children and adults. Instead, the tasks differ crucially in a way that only the design in the adult task isolates the errors properly and is consequently a pure measure of the two shifting processes. Our review concludes that both the ability to shift away from dominant stimuli and activate previously irrelevant stimuli play key roles in set shifting development, yet only the ability to activate previously irrelevant stimuli may be able to explain the high levels of RRBs in young TD children and individuals with ASD.

In chapter 3 we assessed the two predictions in chapter 2 in more depth, through two
experiments that compared different variations on the standard DCCS with a new method in which the relevant response is no longer available. We found an age-related shift in which pre-schoolers learned to pass all task versions around the age of four, offering support for the
proposition that the ability to attend to previously irrelevant aspects of the environment play a key role in set shifting development. We also found support for the prediction that a child’s problems with activating a previously irrelevant cue (rule activation) may reveal biases of attention that explain the persistence of RRBs in typical and atypical development. We explain these through an attentional framework that suggests that the behaviours, and poor task performance is caused by difficulties with overriding automatic avoidance responses. These are responses that have been created over time as a person continuously ignores a response or an activity.

In chapter 4, we evaluated the training literature to address why there are a lack of training studies on the topic. We also made suggestions for future training interventions. More specifically, we stress that EF interventions can be challenging and expensive, as they often require a high level of resources, such as parent training, or supervision of adults or teachers. Moreover, it has been questioned if such interventions can offer long-term training effectiveness, and generalise to situations outside of the lab. Future research should therefore develop a brief and cost-effective EF training program that requires low resources, and can be easily implemented in schools to examine the long-term effectiveness of this type of intervention, as well as if training can have an overall impact on RRB scores.

In chapter 5, we examined the effectiveness of a brief training program to assess if
pre-schoolers and children with ASD can be trained on tasks that measure their ability to activate previously irrelevant rules, and if training has the potential to influence the frequency and nature of their reported RRBs. We found highly significant training effects, and no change in set shifting performance in the control condition. We also found a small, yet not significant, decline in the RRB scores for the TD children after training. These findings propose that a brief rule activation training program may aid set shifting development and thereby be useful in a school setting. The RRB findings are less positive however, perhaps suggesting that to see an effect on the RRBs a training program may need to involve more sessions and run over a longer period of time.

Overall, the results in this thesis provide evidence for the view that rule activation errors play a key role in the development of set shifting skills in pre-schoolers and individuals with ASD. Moreover, these errors may play a crucial role in the development of RRBs, or vice versa.