Pregled bibliografske jedinice broj: 575131
Data Analysis Strategies for Reducing the Influence of the Bias in Cross-Cultural Research
Data Analysis Strategies for Reducing the Influence of the Bias in Cross-Cultural Research // Collegium antropologicum, 36 (2012), 1; 31-37 (međunarodna recenzija, članak, znanstveni)
CROSBI ID: 575131 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Data Analysis Strategies for Reducing the Influence of the Bias in Cross-Cultural Research
Autori
Sindik, Joško
Izvornik
Collegium antropologicum (0350-6134) 36
(2012), 1;
31-37
Vrsta, podvrsta i kategorija rada
Radovi u časopisima, članak, znanstveni
Ključne riječi
importing constructs; latent dimensions; measuring instruments
Sažetak
In cross-cultural research, researchers have to adjust the constructs and associated measurement instruments that have been developed in one culture and then imported for use in another culture. Importing concepts from other cultures is often simply reduced to language adjustment of the content in the items of the measurement instruments that define a certain (psychological) construct. In the context of cross-cultural research, test bias can be defined as a generic term for all nuisance factors that threaten the validity of cross-cultural comparisons. Bias can be an indicator that instrument scores based on the same items measure different traits and characteristics across different cultural groups. To reduce construct, method and item bias, the researcher can consider these strategies: (1) simply comparing average results in certain measuring instruments ; (2) comparing only the reliability of certain dimensions of the measurement instruments, applied to the »target« and »source« samples of participants, i.e. from different cultures ; (3) comparing the »framed« factor structure (fixed number of factors) of the measurement instruments, applied to the samples from the »target« and »source« cultures, using explorative factor analysis strategy on separate samples ; (4) comparing the complete constructs (»unframed« factor analysis, i.e. unlimited number of factors) in relation to their best psychometric properties and the possibility of interpreting (best suited to certain cultures, applying explorative strategy of factor analysis) ; or (5) checking the similarity of the constructs in the samples from different cultures (using structural equation modeling approach). Each approach has its advantages and disadvantages. The advantages and lacks of each approach are discussed.
Izvorni jezik
Engleski
Znanstvena područja
Psihologija, Etnologija i antropologija
POVEZANOST RADA
Projekti:
196-1962766-2751 - Populacijska struktura Hrvatske - antropogenetički pristup (Rudan, Pavao, MZOS ) ( CroRIS)
Ustanove:
Institut za antropologiju
Profili:
Joško Sindik
(autor)
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