Interactivity Theory: Analyzing Human Environments Using Linear Prediction Filters
Abstract
Analyses of complex interactive human environments pose analytic difficulties for commonly used methods. Linear prediction filters were selected as a methodology that could realistically reflect the characteristics of an interactive environment and conform to appropriate scientific criteria: empiricism, replication, prediction, and parsimony. Filters were compared to related linear models (e.g., ANOVA, path analysis). Scientific criteria were used to identify weaknesses in interactive applications of traditional linear methods. A multichannel linear prediction filter was derived and integrated with a measurement model, producing time-series factor analysis. The model was applied to data showing a long term cyclical relationship between promotion rates in the U.S. Army and survey measures of company effectiveness.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 1985
- Accession Number
- ADA172066
Entities
People
- Roland J. Hart
- Stephen C. Bradshaw
Organizations
- U.S. Army Research Institute for the Behavioral and Social Sciences