Pregled bibliografske jedinice broj: 679106
Simple Robust Method for Quasi-Confirmatory Factor Analysis (Three Examples)
Simple Robust Method for Quasi-Confirmatory Factor Analysis (Three Examples) // Collegium antropologicum, 37 (2013), 4; 1071-1080 (međunarodna recenzija, članak, znanstveni)
CROSBI ID: 679106 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Simple Robust Method for Quasi-Confirmatory Factor Analysis (Three Examples)
Autori
Sindik, Joško
Izvornik
Collegium antropologicum (0350-6134) 37
(2013), 4;
1071-1080
Vrsta, podvrsta i kategorija rada
Radovi u časopisima, članak, znanstveni
Ključne riječi
comparing factor structures; exploratory and confirmatory factor analyses
Sažetak
In this article we present a simple robust method named Quasi-Confirmatory Factor Analysis (QCFA), with the purpose of comparing two factor structures, obtained by using exploratory factor analysis (EFA). EFA and CFA (confirmatory factor analysis) approaches, together with other methods that are used in this field, are often used simultaneously in cross-cultural research in testing the possibility of generalizing imported theoretical constructs on different sample of subjects. In the discussions about the matter 'is it better to use EFA or CFA?', it is the most correct to say that each strategy is appropriate for certain research situations. QCFA is conceptually closer to EFA than to CFA, but it gives the exact numerical indicators of the differences, as well as the correlations, between these two factor structures in the final phase of EFAs. The details on the practical application QCFA are presented in three different examples. The advantages and shortcomings of this method are discussed, together with its possible extension.
Izvorni jezik
Engleski
Znanstvena područja
Psihologija, Etnologija i antropologija
POVEZANOST RADA
Ustanove:
Institut za antropologiju
Citiraj ovu publikaciju:
Časopis indeksira:
- Current Contents Connect (CCC)
- Web of Science Core Collection (WoSCC)
- Social Science Citation Index (SSCI)
- SCI-EXP, SSCI i/ili A&HCI
- Scopus
- MEDLINE