TEACCH and Autism Ranking: Strong positive evidence

Status Research

There are a number of limitations to all of the research studies published to date. For example

  • Many of the studies used single-case designs (such as multiple baseline or case series) and most included 20 or fewer participants. For example, the study by Probst et al, 2010 was a case series of only 3 participants.
  • Some of the group studies (such as Welterlin et al, 2012) had fewer than 30 participants.
  • Most of the group studies (such as Tsang et al, 2007) were non-randomised and non-blinded. 
  • Some of the studies (such as Short, 1984) did not independently verify the diagnosis of autism using established diagnostic tools like the ADOS or ADI-R (which were not available when Short wrote his study). 
  • Some of the studies (such as D'Elia et al, 2014) did not provide fidelity measures, meaning it is not clear how closely the intervention did or did not follow the TEACCH model. 
  • Some of the studies (such as Siperas et al, 2007) did not use standardised outcome measure but relied on simple observational measures (such as a frequency check-sheet designed to summarize the frequency of certain challenging behaviours over a month).
  • Most of the studies (such as such as Ozonoff and Cathcart) did not provide long term outcome data, meaning it is not clear how long any benefits of the intervention lasted.
  • According to its authors, the study by Boyd et al (2014) had a number of limitations. These included the fact that the researchers used raw versus standard scores for some measures; the researchers were reliant on school officials to nominate classrooms/teachers to participate in the study; there were some pre-treatment differences between groups; the assessors were not blind to the children’s group assignment; and the control condition used may represent the ‘‘best’’ of standard practice and may differ substantially from the modal level of quality that reflects ‘‘business-as-usual’’ classroom practices.
  • According to its authors, the study by Mandell et al (2013) had a number of limitations. These included the fact that the randomisation was not successful with regard to ethnicity; fidelity was measured through video at monthly intervals; the comparison group comprised a rigorous and promising intervention, rather than “teaching as usual; the study was overly reliant on the Differential Ability Scales as an outcome measure, which may miss important components of academic achievement.
  • Many of the studies (such as Van Bourgondien et al, 2003) were undertaken by researchers who were not independent of the interventions being studied.  Those researchers may therefore have been biased towards the intervention, however unconsciously.
  • Very few of the studies appeared to involve people on the autism spectrum or their parents and carers in the design and development of the studies, although some did involve parents in the evaluations.

For a comprehensive list of potential flaws in research studies, please see ‘Why some autism research studies are flawed’

12 Dec 2017
Last Review
01 Nov 2017
Next Review
01 Nov 2020