Early Intervention in ASD: Continuing Thoughts



This set of notes describes the outcome of a survey of parents’ perceptions of the ABA approach, highlighting the usefulness of direct parental involvement, but also noting the significance of training and experience of those organising the programme.

The SE Regional Group Study is then described with reference to initial findings concerning the apparent effectiveness of ABA and of nursery attendance, with discussion of the non-direct relationship between time on input and outcomes as well as confirmation of the significance of training and qualification among tutors.

The final section simply highlights how themes and emphases change over time, with the corresponding need to examine cases in the here and now when seeking to unravel the significant factors within the interaction of child, circumstances, and intervention most predictive of success.



Parental Evaluations of the Applied Behavioural Approach   


There is converging evidence that ABA can be an effective intervention for many children with autistic spectrum disorders, albeit recognising the variation among studies in terms of the extent of progress and the interaction effects among child, facilitator, and environmental characteristics.


Dillenburger et al (2004) completed a review of studies and were able to highlight a number of examples of positive outcomes from intensive early behavioural programmes in terms of cognitive, social, and behavioural features which can be generalised to different settings and maintained over the long term.  They suggest that the best known of the programmes are those established by Lovaas and his colleagues and which involve discrete trial methodology, but that there are many variants complicated by considerable differences in the extent of training and expertise of the programme facilitators/therapists (and the present writer would suggest that many programmes described as following the Lovaas 1987 principles may actually have diverged from the original methodology as a result of therapy drift).


However, these authors suggest that there has been little information elicited about parental experiences and perceptions of ABA programmes.


One of the conclusions from their literature search is that the quality of ABA training for those who will be organising the programmes is crucial in predicting the likely outcomes; and reference is made to Smith et al (2000) who found that parent-employed therapists who had received short term training were less effective than college trained therapists. 


Further, the style of training for parents in ABA differs among programmes.  Some may focus upon specific and distinctive types of intervention, while others may emphasise more general behavioural principles. 

The former have been subject to criticism with, for example, Bronson and Thyer (2001) describing such training as a “cookbook approach to treatment” or as implying inappropriately that children may benefit equally from any one of a number of approaches.

The latter are seen as more viable as a means of providing a clear understanding of a wide range of behavioural principles with the opportunity afforded for parents to adapt elements of interventions to fit the particular and ever-shifting needs of their children with autism.


In this current study by Dillenburger et al (op.cit), the focus was upon an 18-week introductory course for parents in ABA, with its orientation towards the general principles’ approach, covering topics such as reinforcement, extinction, increasing adaptive and decreasing maladaptive behaviours, shaping and chaining alternative responses, stimulus discrimination, observation, and data collection. 

Following each session, the parents implemented the practical skills developed, and gradually developed an individualised programme to meet the style and needs of their own children. 

An ABA professional visited the families every week for a 2-hour session in which to supervise and modify the programmes in consultation with the parents.


The authors are aware of only one study specifically of parental satisfaction with intensive early behavioural intervention. 

This was completed by Boyd and Corley (2001) and involved Lovaas type therapy delivered to 22 children with autism by tutors trained in discrete trial methods and monitored by supervisors with expertise in this field.

The mean length of treatment was 23 months and the mean age of the children at the start was 41 months. 2 children were on home programmes, the rest received the discrete trial training in a school setting.

Weekly meetings were held among tutors, supervisors, and parents (but it is not clear how involved, if at all, the parents were in the implementation of the intervention.)


The conclusions drawn were that all children had made progress, but no child had approached “recovery”.  Meanwhile, parents were described as generally satisfied or very satisfied with the implementation of the programmes and the outcomes achieved.


The Dillenburger et al study differed in that the training did not focus upon one specific approach, but upon wider principles enabling them to modify interventions to the particular needs of the children; older children were included; all the children were on home programmes; and the effect of length of intervention was assessed. Further, it was the parents, rather than employed tutors, who carried out the applied practice of the skills they had learned as frequently as possible during the day.


10 children for whom evaluation data were available had been involved in a programme for an average of 6 months; a longer term group of 12 children for whom evaluation data were available had been involved for an average of 35 months. 

The age range of 3 to 12 years was similar in the two groups, although more of the older children were in the long-term group.  The mean age of children at the start of the intervention was 91 months for the long-term group (10 boys and 2 girls), and 46 months for the short-term group (9 boys and 1 girl).  

The parents in the two groups were seen as largely similar in terms of employment status and socio-economic status, and in terms of the ratio of mothers to fathers.  


Evaluation on the part of parents was organised through questionnaires exploring the perceived validity of ABA, the significance of the goals selected, the appropriateness of the strategies, and the importance of the outcomes.


The results indicated that parents saw ABA as having made a difference to a number of mutually agreed goals, such as independence, skills development and maintenance, and interaction. 

Only one parent was unsure whether ABA had made a difference to the quality of life of her 10 year old severely disabled child.

There was consensus over the effectiveness of the strategies adopted, the positive impact upon family life, and the self image of the parents.


It was noted that prior to the introduction of ABA, there had only been one intervention approach (TEACCH) available in the area of Northern Ireland where this study was sited, and the availability of a choice seems to have been very well regarded, and the outcomes of the use of ABA perceived positively in terms of facilitating the development of the children and in enhancing parental feelings of confidence and empowerment.


The authors speculated whether the training in the wider and general principles of a behavioural approach was valuable in enabling parental interactions with their children and strategies adopted to be modified according to observed performance and to the changing needs of the children over time. 

It is recognised that treatment intensity has some impact (albeit a complex one interacting with child and tutor characteristics) and it was further speculated whether educating the parents in ABA principles enables the treatment intensity to be maintained at a high level since all interactions between child and parent-tutor will be informed by the training and ongoing observations.


The suggestion is for a future study to compare the outcomes of programmes which involve parents as tutors and those where the parents have received training in the principles involved but do not have the responsibility implementing the programme. 


[Meanwhile, the present writer - MJC – would point out the lack of data concerning the actual application of the behavioural approaches at home.  “As frequently as possible throughout the day” was the description of events, and one might have preferred more specific data, albeit recognising the difficulty of quantifying intensity if all interactions are informed by the gradually increasing awareness of the behavioural principles.


However, one might argue that the positive feelings of the parents are all the more notable if the interventions were not at a consistently and highly intense level.


On the other hand, the evaluative data were received from only 12 of 25 families in the long-term group, and from 10 of 25 families in the short-term group.  One might ask whether responses are the more likely where positive outcomes have been perceived such that the observed data could be skewed. 

The problem is that one cannot determine the answer to such a question, and one can only ponder over the likely correlation between the extent of parental involvement in the programmes and the probability of positive perceptions of their usefulness.]         


Meanwhile, there is evidence form this present study that, while there may be an optimal and early age for starting on a programme, there can still be benefits reported by parents from later onset programmes.

Similarly, short-term benefits are also reported despite the common view (eg Smith 1999) that length of intervention is a significant factor.


In their conclusion, the authors restate the indications that parents are satisfied with the outcomes achieved through basic ABA training, both in terms of developmental progress in the children and in their own feelings of confidence or the enhanced quality of family life.


However, one notes further their reference to the need for ongoing studies by which to gain more data by which to clarify exactly what are the variables responsible for the achievement of treatment goals (and, one might add, where and when they interact).


South East Regional Study : Early Interventions


The work completed by Reed et al (2004) involved following the progress of groups of pre-school children, all diagnosed with autism.


There were 4 groups ….. The ABA group (N=12); The Nursery group (mainstream and specialist, N=13); The Portage group (N=13); and the PACTS group (N =9).

Data are still being collected on a number of children; and several children in the ABA and Portage groups require further visits.


(PACTS is the early intervention programme for young children with autism introduced in Bexley.  It is home-based and designed for children between 2 and 4 years of age, and provides parents with training to gain skills by which to understand autistic difficulties and to deliver teaching and intervention to foster the child’s development.  The emphasis is upon social cooperation, communication, self-help, basic skills, and play.)


The available findings so far …. and one would stress that the analysis is at a very early stage …. suggest that IQ gains over the nine month period, according to the Psycho-Educational Profile, are very good for the nursery group (13 points), and moderately good for the ABA and PACTS groups (8 points).  The Portage group was described as showing no noticeable gains.

In terms of socialization, according to the Vineland Adaptive Behaviour Scale, the nursery group made moderately good gains (7 points), and the ABA and PACTS groups made some gains (4 points).  No gains were observed in the Portage group.


The writers acknowledge that they now need to examine closely the precise input in the groups, and seek to unravel the interactions between child characteristics and the nature and intensity of the interventions, the impact upon the families, and the training issues.


However, two matters have been noted as worthy of comment.

Firstly, there appears not to be a direct relationship between the time input and the extent of gains.  The relationship is likely to prove complex, and reflect the unique interactions between the characteristics specific to the child, and those relating to the nature of the intervention, the expertise of the tutor, and the family.

Secondly, the level of staff training appears to be significant.  As noted in the previous section, there is emerging evidence that interventions organised by well- qualified ABA or teaching staff may well bring about good outcomes; while those organised by less qualified staff appear not to produce gains.


While these and other issues will be explored more fully, one might still suggest that they provide helpful pointers for LEAs who are faced with decisions whether or not to support early intervention programmes, even if it is recognised that these current samples are quite small and that there will be scope for debate whether the findings can be generalised from these children in these particular interventions involving these tutors at this time.




“Historical” Features


The present writer – MJC – has noted how intervention programmes may gradually evolve in their general style, and how a programme established for a specific child may shift in its emphasis or style with time and with changes in the child’s needs.

It was interesting, therefore, to look back at some of the much earlier conclusions or advice relating to ABA (Lovaas) approaches.


For example, Lovaas et al (1973) evaluated the outcomes of their intervention with 20 autistic children with apparently severe difficulties, and the summary provided by O’Leary and Wilson (1978) highlights the “heavy reliance” upon a number of operations …..


This study caught the attention because of the debate over the relevance of aversives in Lovaas programmes, with the current view from Lovaas and his associates that aversives are not and should not be used.


Meanwhile, there is the recognition that “the programme did not give everything to every child.  Sometimes it gave very little to a particular child, but it did give something to each child ”. 

However, the significance of parents as tutors was also recognised at this early stage, and the advice highlighted the likely effectiveness of teaching parents behaviour modification skills to ensure the continuation of a specific form of intervention until peers’ or adults’ behaviour become sufficiently rewarding to maintain the child’s newly acquired skills. 

(It was also noted at this time that parents seen as most successful were the ones who “showed a willingness to use strong consequences, such as food or spankings, to be emotionally responsive, and to show their anger as well as their love ” !)


In other words, one might still argue that it may be dangerous to cite events and outcomes cited in studies which relate to a particular combination of strategies and a particular sample of children and which may reflect practices which are no longer common or which have undergone shifts of emphasis. Instead, there is a need to examine cases in the here and now in trying to determine which are the most effective components of an intervention. 

The goal remains that of identifying the characteristics in the child or in his/her circumstances which are most predictive of success of a given intervention, of gaining valid data including a base-line measure of performance such that monitoring can be meaningful, and of being realistic in recognising when the intensity or other costs of an intervention are not being matched by observable progress.


M.J.Connor                                                                                         February 2004





Boyd R. and Corley M.  2001   Outcome survey of intensive behavioural intervention for young children with autism in a community setting.   Autism  5  430-441


Bronson D., and Thyer B.  2001   Behavioural social work : where has it been and where is it going ?   The Behaviour Analyst Today  3  192-195


Dillenburger K., Keenan M., Gallagher S., and McElhinney M.  2004   In Press

Journal of Intellectual and Developmental Disabilities


O’Leary D. and Wilson G.  1978   Autism : behavioural approaches to treatment.  In Readings in Autism.  Special Learning Corporation : Guilford, Connecticut.  (First printed in Behaviour Therapy Application and Outcome 1975)


Lovaas, Koegel, Simmons and Long  1973   Cited in O’Leary and Wilson (op.cit)


Reed P., Osborne L., and Corness M.  (University of Swansea) 2004    SE Regional SEN Partnership : Early Intervention for Autism.  Presentation to The SE Partnership Working Group, January 2004.


Smith T. 1999   Outcome of early intervention for children with autism.   Clinical Psychology : Science and Practice  6  33-49


Smith T., Buch G., and Gamby T.  2000   Parent-directed intensive early intervention for children with pervasive developmental disorder.   Research in Developmental Disabilities  21  297-309

This article is reproduced by kind permission of the author.

© Mike Connor 2004.

Back to NAS Surrey Branch Welcome Page