All posts by Forging Connections

Why should we take a dimensional approach to studying developmental disorders?

Developmental disorders like attention deficit disorder, attention deficit hyperactivity disorder (ADD/ADHD), autism spectrum disorder (ASD), language, learning and movement disorders are relatively common, more common then we might think. Furthermore, these disorders have a considerable impact upon the daily lives of those who struggle with them. Because some of these disorders are more apparent in some contexts than others, and their severity is highly variable, they may go unrecognised for some time. Indeed, in many cases these disorders are only formally recognised when a child has already progressed through years of formal schooling. This means that they may already have had a largely negative experience of learning, may have lost motivation, and may already have fallen far behind their peers.

When trying to study these disorders, researchers normally use a case-control approach. It’s an observational study in which two groups differing in outcome are identified and compared on the basis of some presumed causal attribute. Researchers use this method to identify factors that may contribute to a medical condition with the help of comparing subjects who have that condition (“cases”) with patients who do not have the condition but are otherwise similar (“controls”).

Case-control studies are relatively cheap and they are a frequently used type of epidemiological study that can be carried out by small teams or individual researchers in single facilities in a way that more complex experimental studies often cannot be. This design is often used in the study of rare diseases or as an exploratory study where little is known about the connection between the risk factor and the disease. In several cases they have bigger statistical power than cohort studies. This approach has largely been translated from the clinical sphere to study developmental disorders.

Well-designed observational studies, like case-control designs, can provide valuable evidence. It is however worth noting that they are quasi-experimental in nature and thus do not bring the same level of evidence as randomized controlled experiments. That said, case-control designs can sit well alongside complementary randomized controlled experiments. There are however other problematic features of case-control designs, which are particularly highlighted when studying developmental disorders. Selecting an outcome of choice, indeed the basis for choosing one particular group, may produce unintended biases that can have a strong effect in overall findings.

One such example is the exclusion of children with any comorbid symptoms – this is routine practice when using a case-control design to study developmental disorders. The most important disadvantage in case-control studies relates to the problem of acquiring reliable information about an individual’s status over time, and then using this as a basis for choosing some children whilst excluding others. The children actually included may be atypical of those with a particular disorder. This exemplifies why this design does not always translate well into the study of developmental disorders; in developmental disorder comorbidity is more the rule than the exception. This approach can also give a false impression of the nature of these disorders, which can be graded rather than discrete.

However, it also is possible to use a dimensional approach instead of the case-control approach. A dimensional approach puts focus on the kind of problem a person is experiencing and on the extent to which that aspect of cognition is impaired. It doesn’t place people into diagnostic categories but along dimensions. Diagnosis then becomes not a process of deciding the presence or absence of a symptom or disorder, but the degree to which particular characteristic is present. This is entirely the approach taken by the Centre for Attention Learning & Memory (CALM; Children are referred not on the basis of any discrete disorder or diagnosis, but because they are experiencing problems in the areas of attention, learning and memory. The researchers are taking a dimensional approach to exploring the nature of these impairments.

Instead of making judgements of “present or not?” the dimensional approach asks the question “how much?”. It ranks disorder on a continuum based upon multiple domains of cognition, assessed using standardised materials. A dimension is viewed as a cluster of related psychological/behavioural characteristics that occur together. This approach generates profiles, rather than discrete diagnostic categories. Of course, one could argue that this approach is far ‘messier’ than a simple case-control approach. However, one might also argue that this unique profiling is far more informative about the nature and extent of the impairments themselves, and provides a far clearer picture about the pattern of deficits actually present in a population of children with problems of learning.

Whilst working at CALM I have been trying to understand how children attend, listen and remember and how these skills impact on learning. These include difficulties in language, literacy and maths. By improving our understanding of the cognitive and brain processes involved in learning, we hope to develop ways of identifying and overcoming problems that might appear during childhood. We also hope to provide an information hub for researchers and professionals in children’s services, and to run regular workshops.

A child visiting the CALM clinic is profiled by grading the severity of symptoms from a number of dimensions using standardised tests. For example these dimensions include working memory, attentional control, short-term memory, phonological skills, the ability to inhibit and control responses, to initiate, plan, organise and set goals, inattention, hyperactivity, aggression, conduct problems, emotional symptoms, peer relations, prosocial behaviour, as well as aspects of communication (speech, syntax, semantics, coherence, initiation, use of context and non-verbal communication). This information can be fed back to the referrer and can then be used to help guide the support that the child receives. In parallel to this we are building a large and rich dataset. A dimensional approach is better able to capture the complexities that a categorical approach may miss. Of course, this approach is not without its challenges also. How does one define a dimension? What statistical approaches ought we to use? And what kind of scores would warrant some form of intervention?

Despite these challenges, we think that the dimensional approach will provide a way of capturing the rich complexities of these data. Whilst other disciplines may strongly favour an approach with rigid categorical boundaries, this approach is not always appropriate for studying developmental disorders. Whilst strict case-control studies can be valuable, reliance on these designs alone can provide a biased and unrealistic view of the children with problems of learning.


Gelder M, Harrison P, Cowen P. Classification and diagnosis. In: Shorter oxford textbook of psychiatry. 5 th ed. Oxford: Oxford University Press; 2006. p. 21-34.

Helzer, J. E., Kraemer, H. C., & Krueger, R. F. (2006). The feasibility and need for dimensional psychiatric diagnoses. Psychological Medicine(36), 1671–1680.

Kendell RE. Five criteria for an improved taxonomy of mental. In: Helzer JE, Hudziak J, editors. Defining psychopathology in the 21 st century: DSM-V and beyond. Washington DC: American Psychiatric Publishing; 2002. p. 3-18.

Lewallen, S., & Courtright, P. (1998). Epidemiology in Practice: Case-Control Studies. Community Eye Health(11(28)), 57–58.


Does working memory training change neurophysiology in childhood?

The short answer to that question is ‘yes’.

We have known for some time that training particular cognitive skills, like working memory, can produce improvements in cognition. These improvements transfer to other untrained tasks, provided that they are similarly structured. However, we know very little about how these kinds of intensive cognitive training programmes change children’s performance.

The study has been under embargo whilst it awaits publication in the Journal of Neuroscience, but the embargo has just been lifted, and we can now tell you all about it (the paper itself should be published very soon – open access, naturally). We used magnetoencephalography – a technique for measuring electrical brain activity – to explore patterns of brain activity as children rested in the scanner with their eyes closed. We repeated this procedure before and after the children underwent working memory training. Importantly, only half of the children underwent an intensive version of the training, with the other half doing a low intensity version. This latter group acted as a control, and the children were randomly allocated to the high or low intensity conditions.

After the training, children’s working memory performance improved. We used standardised assessments of working memory to show this. Importantly, these improvements were specific to the group of children who had undertaken the intensive training programme, with the control group showing little or no improvement. This pattern could not have resulted from us expecting that the children in the intensive group ought to show bigger improvements than the controls, because the researcher doing the assessments did not know which group the children were assigned to.

The magnetoencephalography data allowed us to explore whether there were any significant changes in children’s brains, and whether these changes mirrored the improvements in performance that we observed. We used the spontaneous electrical activity in the children’s brains to explore network connectivity – that is, how different brain areas are coordinated. After the training, connectivity within networks involved in attentional control were significantly enhanced. Furthermore, the bigger the change in connectivity, the bigger the improvement in the child’s working memory.

We have a lot more data on these children, which we are slowly crunching our way through. So there will be more to come!

Coder’s Little Time Saver

We all know the problem; it’s getting late in the office, all your colleagues left hours ago, and your eyes are watering from staring at the analysis output of a script that should have finished running ages ago. Yet for some inexplicable reason, it’s still not done. Wouldn’t it be great if you could just nip out to get some fresh air and be informed when the script is finally done? Well actually, you can! Here’s a handy little tip explaining how to embed email alerts in MATLAB and Python scripts:

MATLAB comes with a handy function that supports sending emails within scripts. But before we can actually get to the email sending, we need to configure some server information. Here is an example for a gmail account:

mail = ''; %Your GMail email address
password = 'secret'; %Your GMail password

props = java.lang.System.getProperties;
props.setProperty('mail.smtp.socketFactory.class', '');

Now, we are ready to send an email:

sendmail(‘’,’Hello there!’);

The second argument in the sendmail function corresponds to the subject line of the email. If you are keen to let MATLAB send a more elaborate email, you can also include a text body:

sendmail('','Hello there!','Have you seen that great post on the Forging Connections Blog?');

It is even possible to send attachments:

sendmail('','Hello there!','Have you seen that great post on the Forging Connections Blog?',{'/Users/Fred/image.jpeg’});

By including these few lines of code in your time-consuming MATLAB script, you can now get notified when it is time to go back to the office for the results.

Python offers a simple solution to send emails from within scripts via the smtplib module. Here is a function that provides the configuration for gmail:
def send_email():
import smtplib

gmail_user = ""
gmail_pwd = "Password"
FROM = ''
TO = ['']
SUBJECT = "Meeting at 3pm"
TEXT = "Are you coming to the meeting?"

# Prepare actual message
message = """\From: %s\nTo: %s\nSubject: %s\n\n%s
""" % (FROM, ", ".join(TO), SUBJECT, TEXT)
server = smtplib.SMTP("", 587)
server.login(gmail_user, gmail_pwd)
server.sendmail(FROM, TO, message)
print 'successfully sent the mail'
print "failed to send mail"

For more information see and