| Summary: | This technical volume is intended to serve as a reference book for researchers who are using constructs or indicators to describe family interaction. In these analyses, each concept {latent construct} is defined by a minimum of 3 {and up to 10} indicators. The volume details the psychometric analyses of each indicator {itemetric study, reliability, distribution, skewness, and kurtosis}. The fit of the indicators to the construct is examined with factor analysis constrained to a single solution. The results of the analyses constitute the preliminary definition of a construct. The process of construct definition is set forth in the papers by Patterson and Bank {1986; in press}. This manual can also serve as a source of further information to researchers who read published articles or books from the Oregon Youth Study {OYS} and need more detailed information on the analyses conducted than can be provided in the space of a book or journal article on theory and results. One of the costs of working with a data set of this magnitude is that the analyses conducted cannot be reported fully in anyone publication, with the result that the interested reader finds it insufficient to replicate the studies. It is hoped that this volume will provide a solid foundation for all who have than a passing interest in the OYS, or in analyses for this type of data more set
|
|---|