Introduction to Music Therapy Research by Wheeler Barbara L.;Murphy Kathleen M.;

Author:Wheeler, Barbara L.;Murphy, Kathleen M.;
Language: eng
Format: epub
Tags: Music Therapy
Publisher: Barcelona Publishers
Published: 2016-07-15T00:00:00+00:00

Problem: Attributing Causality Without a Control Group. The regression-discontinuity design is used when random assignment to a control group is not practical, is not desirable, or is unethical. A review of the capabilities and limitations of the regression-discontinuity design is presented by Imbens and Lemieux (2007). This design can be used when the participants are screened for severity of symptoms so that those with the most severe symptoms, for example, in the worst pain, will receive the treatment first, while those with less severe symptoms will be placed on a wait list and receive treatment later. The screening has a cutoff point for severity. Those very near the cutoff point on either side of it have very similar symptom severity, and the probability is high that they fell on one or other side of the cutoff by chance arising from measurement error. The essence of this design is in comparing the near-the-cutoff participants who were barely included for treatment (those just above the cutoff) to those who were barely excluded from treatment (just below the cutoff point).

Statistically, this subgroup of patients near the cutoff can be treated as if they were randomly assigned to treatment or control groups, and the ordinary least-squares procedures such as ANOVA, ANCOVA, and the whole range of linear and multiple regression methods may be used. The conclusions are causal provided that the researcher did not disturb the cutoff process, for example, by compassionate inclusions of patients in the treatment group who fell below the cutoff but whom the researcher felt should get treatment without delay. The group assignment to experimental and control was not perfectly random but rather as good as random and further disturbance of the process may invalidate it.

The drawback to this design is that only a fraction of the patients treated can be viewed as near the cutoff point and included in the study, so a number of cycles may have to be run before a sufficient sample size is attained.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.