Despite years of intricate variable studies and thorough practicing, the analysis of behaviour was seldom demonstrated in detail; accounts offered regarding existing mechanisms were often general and lacked nuance. It was crucial to have a strong philosophical basis for behaviour analysis as it has applications in a multitude of settings, including intellectual disabilities, crime, and education; behaviours involving social significance. Applied behaviour analysis involves applying behavioural principles to improve socially significant behaviours while evaluating whether behavioural change can be attributed to the intervention. This also includes evaluating the efficacy of said principles in evoking significant improvements, thus allowing the application to be self-examining, self-evaluating, and discovery-oriented. The core difference between this and basic research is that this type of research aims at studying behaviours that have a scope for socially significant improvement. For instance, general research might study the oddities in the chromosomal structure of an individual with intellectual disabilities. However, as chromosomal oddities are not lent to experimental manipulation, applied research instead looks at reinforcement schedules and how they can be manipulated to produce socially improved behaviour. In addition, conducting applied behaviour analysis in the laboratory is seldom feasible as variables are difficult to manipulate if socially significant behaviours are naturally occurring behaviours and part of everyday social settings. Consequently, the authors proposed and articulated seven dimensions – namely applied, behavioural, analytical, technological, systematic, effective, and generalizable – as defining criteria against which application of behaviour analysis can be evaluated. These dimensions aren’t merely descriptions of but rather guidelines for ABA's future.

​

The applied dimension of ABA posits that the behaviour under investigation must be socially significant and not merely based on the researcher’s interests. This separates true applied behaviour analysis from basic research as for something to qualify as “applied", it necessarily has to benefit the individual and/or the community. Thus, what is deemed “socially significant” is not determined by the researcher but rather by society and how a change in behaviour could improve the quality of someone’s life. A key element in this dimension is to address the issues that have real-world implications. For instance, teaching effective communication and reducing self-injurious behaviours fulfill the requirement that the target behaviour is socially significant. ABA is also not only restricted to reducing behaviours of concern but also promotes positive skills such as self-care and other skills prominent in real life. The authors suggest that the best way to identify if an intervention is applied is by asking how important this would be to the subject under study. 

​

The behavioural dimension emphasizes that the behaviour being studied must be measurable and observable. Internal states, such as thoughts, feelings, traits, interpretations, and other abstract constructs cannot be operationalized and thus a socially significant change cannot be meaningfully measured. Behaviour inherently involves directly observable and measurable occurrences of physical events . In contrast to non-applied research, which often has the option to study a measure that is easily quantifiable, applied research is afforded behaviours that cannot be easily quantified. Baer et al. suggest that defining the behaviour on the basis of its origin during the study can help make evaluating target behaviours easier. Although it might appear as though the questions may be trivial, a clear distinction can be noticed when observing recorded change vs behavioural change. To elucidate, the authors consider the example of an individual who was often dressed up by their attendant in the morning. The experimenter then imposed a stipulation on the subject, stating that the subject must dress up within thirty minutes of awakening, which, eventually, led to an increase in the recorded probability of the subject dressing up. This change could either indicate the effectiveness of the penalty in changing behaviour or suggest that the subject could dress up within thirty minutes, given that he was now afforded the time before his attendant intervened. This highlights the fact that the measured data may not always reflect the desired behaviour change metric in the subject. 

​

The analytical dimension requires the demonstration of a functional relationship between environmental variables and the target behaviour. The authors then elaborate on two designs that can be used to elucidate functional control. The reversal technique involves measuring the baseline of a behaviour and examining if an independent variable introduced produces a meaningful change. If so, the variable is removed to measure the diminishing effect of removing the variable. The variable is then reintroduced to measure the recovery of said behaviour. Despite its effectiveness, reversal techniques pose ethical and feasibility issues. For instance, returning to the baseline of the behaviour can be harmful in the case of interventions that target self-injurious behaviour such as self-rumination and self-harm. In addition, if the subject learns the behaviour through this design, it is difficult to unlearn said behaviour and returning to baseline is not always possible, putting the efficacy of the design into question. Thus, an alternative, named multiple baseline, is preferred in situations where reversal of a behaviour is either not possible or undesirable. Multiple baseline involves measuring the baselines of multiple behaviours, followed by implementing an intervention to one of the behaviours. If this intervention produces a change, the experimenter will then apply the same intervention to a different behaviour without withdrawing it from the initial behaviour. Once a change is measured in the second behaviour, the experimenter can point to it as evidence that the initial change was not a matter of coincidence. The authors emphasize the importance of replication in increasing the reliability. They mention that these techniques are exemplifications of replication, which is the essence of reliability. If a technique is reliable, it is more likely to be used and recommended, a quality which is particularly necessary in the field of Behaviour Analysis, as it is in its infancy and often opposed. Once reliability of a technique is established, analysis of said technique is just as necessary to determine which component of the intervention, and how much of it, is responsible for behavioural change. The authors simplified this by considering the example of a teacher using candy to change the behaviour of a student. Component analysis involves determining if the attention of the teacher or the candy is what caused the change in behaviour while the parametric analysis involves determining how much attention/candy was needed to cause the change in behaviour. 

​

Technological dimension states that simply labelling an intervention is not enough; it involves the complete description of the technique implemented. One way to verify a description is technological is to determine if a formally trained interventionist can implement a program and produce replicable results simply based on reading the description of said program. The description must also provide a detailed account of the procedure, specifying all contingencies associated with the response in every form. It is not enough to merely specify what is to be done in response to a particular behaviour but also to possible alternative behaviours. 

​

In addition to being technological, it is also crucial to define and label procedures and base them upon basic principles of ABA. The authors mention that although a thorough description is adequate to produce replicable results, relating them to the basic principles of ABA allows them to maintain their position as part of the discipline. If not, they lose conceptual coherence, making them difficult to learn, teach, and upon which to expand. 

​

The essence of the practical application of an intervention program is to produce socially significant behavioural change. Thus, a program that fails to do so can be considered ineffective. The degree to which a behaviour change occurs is the primary indicator of a program’s effectiveness and to evaluate if said change was meaningful, one might ask how much of the behaviour needed to be changed. This question, although unscientific, can provide insight into whether the change is considered significant by the subject themself or those familiar with the behaviour.

​

The ability of a program to be effective across various environments, behaviours, and individuals, and maintained over time is referred to as generality. If a program is effective in producing a change in only one setting has low generality and thus considered a failure as the same behaviour can emerge in a different setting. Similarly, if a program successfully changes a behaviour, which then represents itself in the future has low generality and consequently, also a failure. Lastly, generality does not present itself automatically. It might be necessary for a program to provide intervention to a behaviour across more than one setting for it to show its effect in settings beyond the ones taught. In essence, generality often requires explicit effort to program without the expectation of its manifestation naturally.

​

In summation, ABA programs should aim to produce socially significant changes in observable and measurable behaviours through clearly defined procedures based on the principles of ABA. Changes must also be significant enough to affect the individual and those around them, while being stable across various settings and maintained over time. Baer, Wolf, and Risley (1968) established a standardized set of principles against which the interventions even today are evaluated for their social significance, scientific rigor, and overall generalizability.

​
 

Citation:-

​

Baer, D.M., Wolf, M.M. and Risley, T.R. (1968), SOME CURRENT DIMENSIONS OF APPLIED BEHAVIOR ANALYSIS. Journal of Applied Behavior Analysis, 1: 91-97. https://doi.org/10.1901/jaba.1968.1-91