Application of Alternating Treatment Designs Single Subject Research

Application of Alternating Treatment Designs Single Subject Research (Richards et al. ) Chapter 10

Alternating Treatments with No Baseline Design Caldwell, M. L. , Taylor, R. L. , & Bloom, S. R. (1986). An investigation of the use of high- and low-preference food as a reinforcer for increased activity of individuals with Prader-Willi syndrome. Journal of Mental Deficiency Research, 9, 347 -354.

Feature description Type of Design Alternating treatments with no baseline Goal of study Determine effectiveness of using food reinforcers to increase activity level Subject Eleven adolescents and young adults with Prader-Willi syndrome Setting Residential summer program on university campus Dependent variable Activity units (20 min of exercise = one activity unit) Independent variable High- and low-preference foods Results and outcomes High-preference foods were effective in increasing activity in 5 of the 11 subjects; no or unclear effects were found in 4 subjects; 2 subjects showed increases under high- and lowpreference food conditions

No Baseline Design Why use an Alternating Treatments with no Baseline design for this study? �There was a limited number of days to conduct this study (It was a summer camp). �The clinical importance of having the subjects increase their activity level was evident. So the design permits researchers to begin their treatments immediately. �However, a no-treatment condition was included as one of the alternating treatments to help determine the relative efficacy of the two food reinforcement conditions, due to conflicting reports about the food preferences in this population.

Limitations of the Study �This study, using multiple subjects confounded the overall results because there was not a consistent pattern across all subjects (three of the subjects did not respond to any of the three treatment conditions, attributed to lower IQs and lack of discrimination among treatments).

Baseline Followed by Alternating Treatments Design Weismer, S. E. , Murray-Branch, J. , & Miller, J. (1993). Comparison of two methods for promoting productive vocabulary in late talkers. Journal of Speech and Hearing Research, 36, 1037 -1050.

Feature description Type of Design Baseline followed by alternating treatments Goal of study Determine effectiveness of modeling procedures to increase vocabulary in young children Subject Two boys and one girl ranging in age from 27 to 28 months who had very limited expressive vocabularies Setting Unspecified Dependent variable I) Frequency of targeted words; (2) number of different words produced; (3) number of targeted words acquired Independent variable Modeling alone and modeling plus evoked production; a no-treatment condition was also used Results and outcomes Equivocal results, with each subject responding differently

Baseline Followed by Alternating Treatments � The three conditions (modeling, modeling plus evoked production, and approximation) were presented in semi-random order, making sure that no more than three sessions of any one type occurred consecutively. � The initial order of treatments was also counterbalanced across subjects. � No baseline data were collected for the group instruction. Why use a baseline followed by treatments design for this study? � Although baseline data are not required in an alternating treatments design, they collected them for the individual sessions to “further document the lack of target vocabulary in the child’s repertoire before teaching” (p. 1040). � Withdrawal design hasn’t been used, because it s anticipated and educationally desirable that the subjects retain the vocabulary words once they learn them.

Limitations of the Study �The results were inconsistent with those in previous research. �The researchers did not make specific recommendations. �But the authors did provide cogent arguments for the presence of specific subject characteristics, such as learning style and personality factors, that might differentially affect response to treatments.

Baseline Followed by Alternating Treatments and a Final Treatment Phase Design Singh, N. , & Winton, A. (1985). Controlling pica by components of an overcorrection procedure. American Journal of Mental Deficiency, 90, 40 -45.

Feature description Type of Design Baseline followed by alternating treatments with a final treatment phase Goal of study Determine if individual components of a multi-component intervention procedure would reduce the frequency of pica Subject Two adults with profound mental retardation with a history of pica Setting Three separate rooms in an institution Dependent Variables Frequency of pica (# of incidents per minute) Independent Variables Three components of an overcorrection procedure—tidying, oral hygiene, personal hygiene Results and outcomes oral hygiene was the most effective; its use was effective across settings and therapists

Baseline-Alternating Treatments. A Final Treatment Phase � In the alternating treatments phase each of the three treatments was randomly assigned to each setting on a daily basis. � The final phase involved using only the most effective treatment but with different therapists. Why use a baseline followed by alternating treatments and a final treatment phase design for this study? � Baseline data were collected to strengthen the study by showing the rates before treatment sessions began. � The use of alternating treatments, particularly across the three settings, allowed each component to be evaluated with a minimum of sequence effects. � The last two phases were included both to test for maintenance and generalization and to leave the subjects with a reduced rate of pica. � The single-component/most effective treatment phase also helped demonstrate that the results were not affected by multiple treatment interference.

Limitations of the Study �The design did not allow for the demonstration that the subjects would return to baseline levels if the treatments were withdrawn. Suggestion: �No-treatment condition in the alternating treatments phase would have addressed that concern. �However, it was more clinically appropriate to focus on the actual treatment conditions.