Functional analysis (FA) is an essential requirement within the practice of supporting individuals with a learning disability (intellectual disability or ID) with behaviours that challenge. The benefits of completing functional based assessments are outlined in current policy and best practice documents, including:
‘Challenging behaviour and learning disabilities: prevention and interventions for people with learning disabilities whose behaviour challenges’ NICE guideline, draft for consultation, December 2014 and ‘Positive and Proactive Care’ Dept of Health 2014.
This study aims to explore the validity of a variant of functional analysis assessment; trial-based functional assessment (TBFA). In the TBFA model, the subject is exposed to specific antecedents and consequences over time limited ‘trials’ within the context of ongoing routines and activities, with test and control settings.
So for example if the function of a behaviour is hypothesized as being ‘to escape’, the person is observed to see if they escape when presented with a particular antecedent or consequence. The aim is to systematically work out what triggers or maintains a behaviour and so verify its function.
The authors outline the prevalence of research into functional analysis procedures and its variants. They find there has been a growth in studies published, with the development of TBFA as a distinct variation of this assessment increasingly included in the published literature. The authors report the need to systematically review the evidence behind this and to identify future areas of research.
Completing Functional Analysis has clear benefits in planning responses to chalenging behaviour
Method
The study is set within the traditional style of a psychological University based review of an area of research. The methodology (the type of study) then focuses on detailing whether and how the selected studies will benefit research and practice.
The literature search follows a fairly standardised but rigorous practice, where electronic and manual searches based on key words are combined. There were a limited number of studies that met the requirements for inclusion, limited by the precise nature of the standardised TBFA procedure. The review identified and detailed 13 studies. The rationale for inclusion, what data was included for consideration and how agreement was reached about these decisions is then detailed in the review.
Findings
The difficulties of conducting systematic reviews of types of assessment are made apparent within the article. Even when comparing a small number of studies, it is still difficult to completely control for the variation of conditions.
The variations that would impact on the results of the study are: the selection of the studies for the review, the choice of subjects, behaviours across studies, settings of studies, who conducted the TBFA trials, the TBFA procedure itself and the outcomes of the trials – successful in identifying behaviour function or not.
These variations reflect the complex nature of the task at hand, that of defining the function of individual behaviours in different subjects across different settings. They also reflect the complexity of the procedure.
The authors believe that the data from the identified studies hold validity in that they enable behavioural function to be identified. This then leads onto ‘effective and efficient interventions’.
The identified implications of this study rest on its stated benefits for research and for practice.
This review identifies the need to complete further TBFA research. This is principally to address the difficulties encountered in validly generalising benefits, from the many variables contained between the 13 studies reviewed. For example TBFA is used predominantly in educational settings with children with a developmental/ intellectual disability.
The studies selected had small subject numbers, and studies were not wholly controlled for randomisation of procedure – (Control and test). These variations may limit how generalisable TBFA is to an adult population in community settings, prompting the authors to point future research to evaluate the ‘social validity’ of TBFAs.
Practice benefits are limited by the complex methodology designed to enable functional assessments to be shortened and so be less resource intense. This raises the issue of training and whether the process would remain valid ‘outside of the field of behaviour analysis’.
Although TBFA is not considered ‘evidence based practice’ yet, within a defined educational setting, it does promise to have less impact on disrupting classroom routines and be less resource dependent. Lambert & Bloom in 2012 identified that TBFA may be especially useful for early childhood settings because it can be conducted in a format that allows for brief assessment trials.
The reviewers identified researchers found it difficult to control for variation
Conclusion and Comment
Functional assessments are a means to an end in that they identify the function of behaviour so that interventions can be identified that support a reduction in ‘maintaining variables’ or triggers.
The authors find TBFA to be a promising assessment of challenging behaviour used in applied or educational settings with children, in that as an assessment it is less resource intense than traditional FA. Susan Bloom, an educational Psychologist and published proponent of TBFA, identifies that this assessment provides the benefits of experimental FA (that require extended resources and time) whilst remaining a practical and convenient way of obtaining valid results.
Other non-standard methods of FA were identified within this study. It was not stated whether they have been found to be as valid or not, although the need seek out ‘acceptable and feasible’ FA models is a recommendation.
Whilst the study adds to the research literature within this restricted sphere of study, it may also distract from the wider practical benefits of ensuring individuals receive thorough functional assessments paired with effective interventions in non-applied community settings.
Although it has been argued that the rigorous selection of cases, with limited subject numbers in distinct settings has limited the generalisability of findings. It can be equally argued that a strength of this study was that TBFA did successfully identify behavioural function, across 13 studies and their wide range of variabilities.
However, only 4 of the 13 studies went on to test for validity of these findings using function-based treatments. In addition, the rigorous selection of case studies may have excluded information and data salient to identifying alternative brief FA methods. It may be there are other brief FA methods with equal or increased validity.
A number of ethical issues have been raised by the use of this (as well as analog FA) methods due to the fact that the procedure directly reinforces the very behaviour being targeted for reduction.
It is recognised that the authors identify there is a need for further research particularly to ascertain if the process has social validity in terms of acceptability, feasibility and suitability.
Need for further research in acceptability, feasibility and suitability
Link
A Systematic Review of Trial-Based Functional Analysis of Challenging Behavior Mandy Rispoli, Jennifer Ninci, Leslie Neely, Samar Zaini Journal of Developmental and Physical Disabilities
26, 3, 271-283 [abstract]
RT @LearningDisElf: Trial-Based Functional Analysis has limited validity outside of defined clinical settings http://t.co/SgCZ46ja7o
The Learning Disabilities Elf liked this on Facebook.
Tracey Rose liked this on Facebook.
Completing Functional Analysis has clear benefits in planning responses to chalenging behaviour http://t.co/SgCZ46ALvY
Hampshire Healthcare Library Service liked this on Facebook.
“@LearningDisElf: Completing Functional Analysis has clear benefits in planning responses to challenging behaviour http://t.co/5yeCun8Q30“
Don’t miss: Trial-Based Functional Analysis has limited validity outside of defined clinical settings http://t.co/SgCZ46ALvY #EBP
Trial-Based Functional Analysis has limited validity outside defined clinica.. http://t.co/XeQH4HaPvD @LearningDisElf http://t.co/b9lf6D9K6c
Trial-Based Functional Analysis has limited validity outside of defined clinical settings. http://t.co/zzg9IsgCsJ
I appreciate the intentions of this author (Russell Woolgar) and this site in general. It is important to make research findings accessible to a wider audience, and toward that end both should be commended. However, I noticed some obvious errors in this blog post and that makes me concerned about the presence of less-obvious errors in this and other posts. Errors at this level could be a big problem because correcting a mis-informed public is likely to compete with the care-taking responsibilities of well-educated therapists. What kinds of editorial and fact-checking work goes into these posts before they are made available online?
Thanks to Andrew for his comment. I asked for some specific questions. I have posted his email below along with Russell’s response/
Andrew Samaha’s email of 15th April 2015
“Hi John,
Thank you for contacting me. I do sincerely appreciate the mission of your blog and I do not think your writer (Russell Woolgar) had any malicious intent. I happened to be doing a google search on trial-based FA’s and came upon it. With respect to the obvious errors I mentioned, I do not know anyone named Susan Bloom doing work in trial-based FA. But I think Mr. Woolgar might actually be referring to Sarah Bloom, who is a Behavior Analyst, not an educational psychologist, and has conducted a number of the studies on trial-based functional analysis.
The reason the post caught my eye was that I was surprised by its title. Specifically, it is not clear if the title reflects Mr. Woolgar’s opinion, or a summary of Rispoli’s systematic review. In my reading of the latter, I do not believe Rispoli comes to that conclusion. If it is just Mr. Woolgar’s opinion, then perhaps that could be made more clear. The trouble for me is that the post reads as a summary of Rispoli’s paper, and as such, it comes across as though Rispoli says that “Trial-based functional analysis has limited validity outside of define clinical settings”.
In the only section of Rispoli’s review that seems relevant to this question (the first full paragraph on page 281), the following text can be found:
“The studies reviewed suggest that experimental analysis of challenging behavior can be moved out of highly controlled clinical settings into classrooms and other applied settings, yet these applied settings were quite clinical and non-inclusive. It is unclear why TBFA’s have been primarily conducted in self-contained settings.”
My reading of this passage is that the procedure has been examined in classrooms and other applied settings (these are both clearly facts and can be supported by examining the literature), but that more work can be done. Is a special education classroom a defined clinical settings? The title of Mr. Woolgar’s post would suggest it is, but I would be surprised if he actually believes that, or if the average reader would make that distinction. I think it’s more likely that at the time Mr. Woolgar wrote the post, he thought the sentence I quoted from Rispoli’s systematic review and the title he chose were functionally equivalent.
On the other hand, perhaps Mr. Woolgar is claiming that inclusive settings and special education classrooms are so different as to likely affect the generality and effectiveness of trial-based functional analyses in the former. Of course, this is an empirical question, but it is a different thing to claim that the procedure will not work there.
This brings me to the question of what the phrase, “limited validity” is intended to mean. I could not find it in Rispoli’s review, which suggests to me it is Mr. Woolgar’s opinion. Literally, I think it would mean that evidence exists to suggest the assessment or its findings do not apply outside of defined clinical settings. But in reality, a good body of evidence suggests it works in clinical and special education classroom settings, and I know of no studies that show the procedure does not work outside of those studies. So, the statement is technically incorrect.
Further, trial-based functional analyses are based on, and borrow heavily both conceptually and methodologically from, the standard functional analysis, which has a substantial body of evidence supporting its effectiveness with a wide variety of populations, including adults, and typically functioning individuals. So, it would be difficult to propose an argument for why one would expect the trial-based functional analysis to not work under some context.
I think, at the very least, it would be helpful if Mr. Woolgar clarified whether the post documents his opinion, or is simply a summary of Rispoli’s systematic review.
Thank you for taking the time to read my concerns.
Regards,
Andrew Samaha”
Russell Woolgar’s response
“Thank you for copying me into the e-mail from Andrew Samaha. (Assistant Professor in the Department of Child and Family Studies within the College of Behavioural and Community Sciences at the University of South Florida) and co-author of TBFA articles with Sarah Bloom. Mr Samaha clearly has an extensive knowledge of behaviour analysis and its use in US settings.
I appreciate his comments, and feel a response needs to include the following:
I agree it is Sarah and not Susan Bloom (apologies) Her main work has been in educational settings and she graduated in Psychology so again apologies, her true title is Behaviour Analyst, although is currently listed as Assistant Professor, may be some confusion mixing UK/US terminology..
Looking at Mr Samaha’s specific points: Yes there are views within the blog which I feel reflect current clinical practice in supporting adults with a learning disability, limiting how generalisable TBFA is to an adult population in community settings.
I believe a classroom can be viewed as a defined clinical setting, and as the article states ‘It may be that those settings offer a bit more control and predictability than inclusive or community-based settings and thus allow for tighter control over TBFA conditions’.
The Point I make is that as yet TBFA has not yet extensively been used in other inclusive or community based settings and so currently has limited validity outside of the settings it has been mostly used in.
I am not claiming TBFA will not work outside of the ‘defined clinical settings’ just that at present there are limited findings, as the review itself states; ‘more research is needed before TBFAs should be common practice’. I do state that the review finds TBFA to be a promising assessment, but that further research is required given the small numbers of subjects/ complexity of factors needed to be controlled for within the studies reviewed.”