Naslov
Što treba znati kada izračunavamo koeficijent korelacije?
(What we need to know when calculating the coefficient of correlation?)

Autori
Udovičić, Martina ; Baždarić Ksenija ; Bilić-Zulle Lidija ; Petrovečki Mladen

Izvornik
Biochemia Medica (1330-0962) 17 (2007), 1; 10-15

Vrsta, podvrsta i kategorija rada
Radovi u časopisima, pregledni rad, znanstveni

Ključne riječi
korelacija; Pearsonov koeficijent; Spearmanov koefiicjent; koeficijent determinacije; pogrješka; statistika
(correlation; Pearson's correlation coefficient; Spearman's correlation coefficient; coefficient of determination; error; statistics)

Sažetak
Abstract: Correlation is a statistical procedure applied to calculate association between two variables. The value of correlation is numerically shown by a coefficient of correlation, most often by Pearson's or Spearman's coefficient, while the significance of the coefficient is expressed by P value. The coefficient of correlation shows the extent to which changes in the value of one variable are correlated to changes in the value of the other. A sign preceding the coefficient of correlation (+ or -) indicates the direction of correlation. The most frequent errors in calculating correlation are related to conditions for calculation, interpretation of the coefficient and correlation significance, high correlation coefficients, assumption of causal relationship, the strength of correlation (coefficient of determination), and comparison of two correlation coefficients.

Izvorni jezik
Hrvatski

Znanstvena područja
Javno zdravstvo i zdravstvena zaštita

POVEZANOST RADA