Methods of Assessing Replicability in Canonical Correlation Analysis (CCA).

Autor/in	King, Jason E.
Titel	Methods of Assessing Replicability in Canonical Correlation Analysis (CCA).
Quelle	(1997), (21 Seiten)Verfügbarkeit
Sprache	englisch
Dokumenttyp	gedruckt; Monographie
Schlagwörter	Correlation; Effect Size; Multivariate Analysis; Robustness (Statistics); Sampling + Suchen Sie Ihr Suchwort? Korrelation; Multivariate Analyse; Widerstandsfähigkeit
Abstract	Theoretical hypotheses generated from data analysis of a single sample should not be advanced until the replicability issue is treated. At least one of three questions usually arises when evaluating the invariance of results obtained from a canonical correlation analysis (CCA): (1) "Will an effect occur in subsequent studies?"; (2) "Will the size of the effect replicate?"; and (3) "Will variables induce similar weightings on replicated variables?" When "external" replication is not feasible, "internal" invariance estimates using only data in hand can yield information about the preceding questions. Approaches to assessing replicability that are discussed are Wherry's correction formula, rotation of function and structure matrices, crossvalidation, and bootstrapping techniques for CCA. Even the bootstrap method is afflicted with the problem common to all the methods of replicability discussed in this paper: results are based solely on data obtained from a single sample. Since this technique creates multiple configurations of the data, it is more robust to the sample-specificity limitation, but it is always advisable to compute multiple invariance estimates. These approaches can aid in determining the replicability of research results, but they should only be used when external resampling is not feasible or practical. (Contains 21 references.) (Author/SLD)
Erfasst von	ERIC (Education Resources Information Center), Washington, DC