ERIC Number: ED356251
Record Type: Non-Journal
Publication Date: 1992-Nov
Reference Count: N/A
Detecting Point Diffusion Patterns in Two-Dimensional Maps: A Simulation Study.
A way of identifying non-random patterns of effects on a group of individuals as a result of some intervention when a sample of participants is arrayed according to some indices of similarity is presented. The principle of proximal similarity and the concept of pattern matching provide the background for this effort. Major advantages are the treatment of an individual as a unit of analysis, visual representation of the outcome across individuals, and the ability to use the representation in pattern matching. A non-metric multidimensional scaling method is used to generate a person map in which each individual is represented by an individual point (profile analysis). From the developed rectangular profile matrix a square matrix can be constructed to be used as the input for multidimensional scaling. Once the person map is generated with the overlaying of gain scores, the next step is to determine the presence of prototypical individuals on the map. A simulated data set representing the ideal pattern is compared to the map to determine point diffusion patterns. Although more work is needed to establish this method as an alternative to conventional statistical methods, it appears to provide a practical blend of qualitative thinking and quantitative methods. Eight figures illustrate the process. (SLD)
Publication Type: Reports - Evaluative; Speeches/Meeting Papers
Education Level: N/A
Sponsor: National Inst. of Mental Health (DHHS), Bethesda, MD.
Authoring Institution: N/A