Based on a reanalysis of 881 studies in applied behavior analysis, Huitema (1985 Behavioral Assessment7, 107-118) concluded that behavioral data are not autocorrelated. This conclusion was challenged by Suen (in press Behavioral Assessment) from a number of methodological perspectives. One of the most essential criticisms was the inherent high probability of Type II error in Huitema's conclusion. In this paper, the common power analysis method is modified and is suggested as a means to assess the appropriateness of a "no autocorrelation" conclusion.
|Original language||English (US)|
|Number of pages||6|
|State||Published - Mar 1 1987|
All Science Journal Classification (ASJC) codes
- Psychiatry and Mental health