Skinner on Averaging: Was He Right? Should We Keep the Faith?
|Sunday, May 28, 2017|
|6:00 PM–6:50 PM |
|Hyatt Regency, Centennial Ballroom D|
|Area: PCH; Domain: Theory|
|Instruction Level: Advanced|
|CE Instructor: Neville Morris Blampied, Ph.D.|
|Chair: Darlene E. Crone-Todd (Salem State University)|
|NEVILLE MORRIS BLAMPIED (University of Canterbury)|
|Neville M. Blampied graduated from the University of Auckland in 1970. That year he moved to a faculty position in the Department of Psychology, University of Canterbury, Christchurch. In 40+ years at Canterbury he morphed from a physiological psychologist to a behavioural pharmacologist to an experimental behaviour analyst and then (finally) to an applied behaviour analyst. His major research area for the past 20 years has been in applied family psychology, notably pediatric sleep disturbances. Recently he has become concerned with methodological issues in research and with attempting to position single-case research as an alternative to the increasingly discredited null-hypothesis significance testing tradition in Psychology. In December 2012 he completed a 7-year stint as Head of Department and member of the Executive of the College of Science. He also served six years as Director of Scientific Affairs for the New Zealand Psychological Society (2004–2010), two years as National President of the Association of University Staff (2000–2001), three years on the Board of the NZ Universities Academic Audit Unit (2001–2003), and was President, Division 6 of the International Association of Applied Psychology 2010–2014.|
In his magnum opus, The Behavior of Organisms (Skinner, 1938) Skinner presented an experimental analysis of behavior that eschewed averaging data across subjects. Even at the time, this stance was unusual, and since then has become even more so. Since the 1950's more than 80% of published quantitative empirical research in Psychology has used group mean data analysed by null-hypothesis statistical tests (NHST). Recently in behavior analysis there have been recurrent calls for the more widespread use of averaging (and NHST) and claims that this would make behavior analysis more acceptable to mainstream psychology. So, was Skinner right, or should we resort to conventional data analytic practices? This paper will review the recent calls for change within behavior analysis about averaging. Then it will consider some developments in fields outside of behavior analysis that bear on the question. It will consider implications from biology and natural selection. It will also consider growing criticisms of group averaging from within mainstream psychology, especially from personality research. Finally it will consider the implications of fundamental measurement theory concerned with ergodicity. The conclusion - Skinner was right, and behavior analysis should keep faith with his rejection of group averaging.
|Target Audience: |
Basic and applied researchers and methodologists in behavior analysis and those concerned with the behavior of scientists
|Learning Objectives: At the conclusion of the presentation, participants will be able to: (1) describe typical data analysis techniques in behavior analysis; (2) describe typical data analysis techniques in the psychological sciences; (3) describe compare and contrast the typical data analysis techniques in behavior analysis and the psychological sciences; (4) describe the contexts in which the typical data analysis techniques in behavior analysis and the psychological sciences are appropriate and inappropriate.|