“…Where, m number of days, n number of chickens, 2 S sum of square between regression coefficients for each groups and 1 S sum of square of deviation from regression for each groups (KOCABAŞ et al, 1997). Following linear model was used to compare groups for the growth curve parameters of body mass index (AKBAŞ et al, 2006). (3) Where ij Y =growth curve parameters for BMI, µ =overall mean, i α =fixed effect of the group (i=1, 2, 3), ij e =residual error distributed as N ( 2 σ 0, ) Nonlinear model (1) to estimate growth curve parameters and linear model (2) to identify group differences was fit using NCSS statistical package program (HINTZE, 2001).…”