The article in question is entitled “Cognitive Effects of Risperidone in Children with Autism and Irritable Behavior” (Aman et al., 2008). As the title implies, the goal of the research was to determine if using risperidone can affect cognitive performance in pre-adolescents and adolescents with autism. The researchers hypothesized that there would be a difference between placebo and risperidone, taking the lack of difference as a null hypothesis. The results showed that risperidone did not change the cognitive performance for the worse.
The study employed a parallel group design with random allocation of treatment, with the treatment group and placebo-assigned groups undergoing the preliminary cognitive function assessment and concurrent assessments at weeks 4 and 8. The mastery variable was constituted by the performance in the dot test, cancellation task, pegboard task, and classroom analogue task. In relation to the latter, the Wide Range Achievement Test was conducted to determine the participants’ math level; the results were not taken as a variable. Other variables included age in years, IQ, irritability, gender, and CGI severity (Aman et al., 2008). Considering that multiple variables were derived, ANOVA model was sufficient.
The limitations of the study include a small sample size (n = 38), which is unsuitable for predicting causality. On the other hand, the study was aimed at exploration and presenting a baseline for further research; besides, children with autism are a population that cannot be assessed easily. Consequently, the findings can prove important for further practice. Some corrections that can be made in this case would be devising a cognitive testing approach to include younger children and children with lower IQ for maximum heterogeneity.
The analysis involved only children with valid mastery levels. To analyze the problem statistically, the researchers first analyzed the children’s performance as a function of risperidone allocation as opposed to the placebo and the timespan. Secondly, they used the General Linear Models package, which is a statistical model describing relationships between more than one predictor. The results can be interpreted as follows.
The statistical significance does not always indicate the clinical significance of the findings. Considering the limitations of the study, the clinical significance is doubtful. The usage of risperidone did not cause any performance decline, but the positive outcomes were modest as well: F (1,17) = 3.18, p =.05, ηp2 = .16 on cancellation and F (1,13) = 4.42, p =.05, ηp2 = .20 on verbal learning, which are the only areas of improvement (Aman et al., 2008).
On the other hand, the significance of risperidone usage can be explained through behavioral performance: irritability and hyperactivity were reduced by ES = 1.20 and ES = 1.00, respectively. The application of risperidone, therefore, can be sufficient in behavior correction as opposed to cognitive performance. The authors mention the positive impact of risperidone on behavior but fail to inspect if such an outcome can be the cause of better cognitive performance.
Additionally, the research would benefit from conducting power calculation to reject the type II error: a minimum difference between the cognitive performance of the placebo-assigned group and the risperidone group could be calculated to prevent falsely identifying the null as false.
Overall, considering that there was little research exploring the cognitive performance of autistic pre-adolescents on prescribed antipsychotics back in 2008, the present research is relevant. The researchers performed a complex task of assessing children with autism and certainly provided a solid platform for subsequent research.
Reference
Aman, M. G., Hollway, J. A., McDougle, C. J., Scahill, L., Tierney, E., McCracken, J. T.,…Posey, D. J. (2008). Cognitive Effects of Risperidone in Children with Autism and Irritable Behavior. Journal of Child and Adolescent Psychopharmacology, 18(3), 227-236.