Pregled bibliografske jedinice broj: 775654
Random Number Generator Comparisons of Effect Size Measures in One-Way Repeated Measures ANOVA
Random Number Generator Comparisons of Effect Size Measures in One-Way Repeated Measures ANOVA // Proceedings Book / Grgantov, Zoran ; Krstulović, Saša ; Paušić, Jelena ; Bavčević Tonči, Čular, Dražen ; Kezić, Ana ; Miletić, Alen (ur.).
Split: Kineziološki fakultet Sveučilišta u Splitu, 2015. str. 585-594 (predavanje, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)
CROSBI ID: 775654 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Random Number Generator Comparisons of Effect Size Measures in One-Way Repeated Measures ANOVA
Autori
Jelaska, Igor
Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni
Izvornik
Proceedings Book
/ Grgantov, Zoran ; Krstulović, Saša ; Paušić, Jelena ; Bavčević Tonči, Čular, Dražen ; Kezić, Ana ; Miletić, Alen - Split : Kineziološki fakultet Sveučilišta u Splitu, 2015, 585-594
Skup
5th International Scientific Conference Contemporary Kinesiology
Mjesto i datum
Split, Hrvatska, 28.08.2015. - 30.08.2015
Vrsta sudjelovanja
Predavanje
Vrsta recenzije
Međunarodna recenzija
Ključne riječi
methodology; practical significance; meta-analysis; variance decomposition
Sažetak
Goal of this research is to discuss practical and theoretical issues of different effect sizes parameters together with theoretical background of One-Way Repeated Measures ANOVA. In accordance with a goal, random number generator was applied for simulation of N1=500, N2=350, N3=200, N4=150, and N5=100 subject's and k=3, 4, and 5 repeated measurements. Effect size parameters eta squared, epsilon squared, and omega squared were calculated, compared and discussed. Furthermore, effect size is discussed as „unit independent“ measure of how much variability of dependent variable is due to treatment or different experimental conditions effect, as parameter in power analysis while determining the sample size for future studies and finally as parameter which allows meta-analytic conclusions by comparing effect sizes across similar studies. Results and theoretical background clearly indicate that in applied sciences, it is necessary to report effect sizes as a measure of practical significance in designs concerning differences repeated measures or groups simply because values of statistical tests can be significant when the mean differences are so small that they do not have or have very small practical value.
Izvorni jezik
Engleski
Znanstvena područja
Informacijske i komunikacijske znanosti
