If you have any feedback for the author, E-mail me
REQUIRED TEXT: Experimental Methodology eighth edition by Larry B. Christensen, Allyn and Bacon, 2001 ISBN #0-205-30832-5 ------------------------------------------------------------ Chapter 11 Single-Case Research Designs, e.g. Biofeedback form of Time-Series, repeated measuring, within-subject design so no control grp possible Compare participant's own pre-treatment responses to his/her post- treatment responses Pre-treatment measurements until they stabilize -> avoid maturation errors but assumes that, if there had been no treatment, this same pattern of responding would continue -> History error - Pre predicts Post A-B-A Designs: A => Baseline (measurement without treatment): B => Treatment: After removal of B, if behavior reverts back to pre-A, then alternate hypotheses (e.g. history, carry-over, relatively permanent changes in behavior) are very unlikely. But if B is a therapy or learning procedure, post-A hopefully will NOT return to pre-A levels. Then the use of placebo controls, waiting list groups, 'physiological status' conditions help eliminate alternate hypotheses. ABAB design allows comparisons of rates of change in behavior across conditions and sends the client home after treatment. Interaction Design: combined influence of 2 or more specific levels of 2 or more different IVs. :. Isolate the interactive effect of levels of 2 IVs from the effects of only one, e.g. A - B - A - B - BC - B - BC, A - C - A - C - BC - C - BC, B = feedback; C = praise or different feedback. :. At least 2 people and neither variable individually produce maximum (ceiling) effect. Multiple-Baseline Design: multiple behaviors in same individual, same behavior for multiple individuals, same behavior across multiple situations for same individual, e.g. Systematic Desensitization by Wolpe, Biofeedback to one mode while recording 4 modes. Target behaviors must be interrelated, e.g. inappropriate motor, verbal, task behavior, may need 4 or more baselines, and one should reverse. Changing-Criterion Design: once stable behavior is maintained on one criterion, the target criterion is raised, e.g. Learning, Biofeedback. a) the length of baseline and treatment phases must be long enough and variable to insure that increases in behavior due to treatment criterion and not time; b) small criterion steps that are achievable; c) two or more criterion changes, more the better. Warnings: Baseline must be stable Change one variable at a time Time must be somewhat short to avoid cyclic changes from influencing results, e.g. SAD For establishing causal relationships: May need several participants Reversal and Multiple-Baseline designs are strongest. ------------------------------------------------------------ My office is in room 116C of HGH (826-4571) and I invite you to consult with me on any matter during my office hours or any time that is agreeable to you and me.Go back to the beginning
Copyright © 2002, Dr. John M. Morgan, All rights
reserved -
This page last edited Nov 25, 2002
If you have any feedback for the author, E-mail
me