Faculty Bibliography

This article discusses single-case experimental designs (SCEDs) and their relevance to developmental-behavioral pediatrics. Information concerning SCEDs have not been described in the Journal of Developmental and Behavioral Pediatrics, despite its relevance to the field. General issues related to the underlying logic and applications of SCEDs are reviewed with examples selected from the literature to illustrate the strengths and weaknesses of different design strategies. It is suggested that SCEDs can be a useful alternative to traditional between-group designs for clinical and evaluation research because the unit of the analysis is the individual; therefore the feedback to clinicians and families is direct about the effectiveness of a behavioral intervention or medication for that individual. In the field of developmental-behavioral pediatrics, SCEDs can be especially useful in the management of vague symptoms or poorly defined diseases to improve the confidence in a treatment decision for an individual patient. This report is intended to facilitate the understanding and use of single-case methodology so that clinicians are aware that flexible, true experiential designs exist to fill the gap in knowledge and also "do the best for my patient."

This paper presents an argument for more extensive use of single-case experimental research designs in medical education research. Single-case experimental designs consist of a group of experimental techniques that are widely used in the social sciences but are just beginning to be utilized by medical researchers. The method emphasizes reliable observations of behavior, repeated measurements of outcome, and individualized tailoring of objectives for each subject; all of these occur within a system that allows an experimental analysis to be conducted. Single-case designs are particularly useful when only small numbers of participants are available for a relatively long period of time. Trends in medical education toward individualized instruction, adult-centered learning, and fine-grained analyses of medical skills and knowledge make this field especially amenable to single-case experimental designs. Issues of internal and external validity, generality, practicality, and ethics are discussed, and several typical designs are illustrated. While the emergence of qualitative research methods in medical education may prove useful, single-case designs can maintain experimental science's emphasis on methodologic rigor, while allowing the flexibility often needed to conduct research in applied settings.

The relationship of cardiac activity to stereotyped behaviour was studied in profoundly mentally retarded persons, using both inter-group comparisons. The first intra-group comparison involved examining the relative magnitude of an orienting response to an auditory stimulus during the occurrence or nonoccurrence of stereotyped body rocking. While no differences were found between periods of rocking and non-rocking when behavioural measures of orienting were used, smaller changes in post-tone heart rate were observed during periods of stereotyped rocking. A second intra-group comparison found that stereotyped behaviour was associated with a significant increase in heart rate but no change in heart rate variability when compared to periods of no stereotyped activity. Finally, using an inter-group comparison, individuals who displayed stereotyped body rocking exhibited higher tonic heart rates and lower heart rate variability during periods of no stereotyped behaviour than was observed in matched control subjects. This result suggests that stereotyped behaviour may be correlated with reduced vagal tone.

The pathological stereotypies frequently observed among severely mentally retarded and autistic persons are highly rhythmical in nature. Most attempts to quantify such behavior, however, have not analyzed stereotypy in terms of its cyclical properties. In the present paper we have detailed a method for electronically transducing stereotyped body-rocking and analyzing its frequency and amplitude characteristics with a standard polygraph and microprocessor. The relationship between stereotyped body-rocking and cardiac activity was also described using time-series analysis. The method described should provide a sensitive index of various experimental manipulations and treatment effects.