hrvatski

TRICH - Algoritam i program za hijerarhisku faktorsku analizu pod komponentnim modelom

Definiran je algoritam i napisan program za hijerarhijsku faktorsku analizu analizu pod komponentnim modelom skupa kvantitativnih multivarijatno normalno distribuiranih varijabli. Soluicija prvog reda glavnih komponenti reduciranih prema PB-kriteriju (Štalec i Momirović, 1971) transformira se Harris-Kaiserovim orthoblique postupkom, nakon čega se izvodi ekstrakcija faktora drugog i trećeg reda po GK-kriteriju (Guttman ; Kaiser, 1960) redukcije, te orthoblique transformacija dobijenih solucija. Za svaki faktorski nivo određuje se relacija sa inicijalnim skupom varijabli, te relacija među faktorima različitih nivoa. Program je napisan u meta- jeziku SS (Zakrajšek, Štalec i Momirović, 1974) verzija 5.1 i izveden na računalu UNIVAC 1100 Sveučilišnog računskog centra u Zagrebu.

hijerarhijska faktorska analiza ; komponentni model

Rad je citiran (CC, SCI-1984) u članku: Stančić, Vladimir (1984):Some subjective and objective prerequisites to educational integration of handicapped children, International Journal of Rehabilitation Research: September 1984 - Volume 7 - Issue 3 - ppg 273-282. https://journals.lww.com/intjrehabilres/Citatio n/1984/09000/Some_subjective_and_objective_prer equisites_to.2.aspx *** TRICH - an algorithm and program for determination of hierarchical latent structure *** Abstract: An algorithm has been proposed and program designed for determination of hierarchical latent structure of set of multivariate normal distributed variables. The algorithm determines significant principal axes of variable correlation matrix, reduced according to PB- criterion (Štalec, Mmirović, 1971) and transforms them according to Harris- Kaiser orthoblique transformation (Harris. Kaiser, 1964). Extraction of factors of second and third order is determined according to G.K.-criterion (Guttman-Kaiser. 1961). and factors obtained are also transformed by orthobllque transformation. Algorithm determines relations to initial set of variables for each factorial level, and defines relations among factors from dif