Study of K（892） mass and width splitting caused by model difference

According to the new K^＊（892）^0 and K^＊（892）^- masses reported by the BELLE experiment and the K^＊（892）^0 mass reported by the FOCUS experiment, mass splitting between neutral and charged K^＊（892） becomes very small. This is significantly different from the current world average values given by the Particle Data Group 2008. We find that there are differences between models used to fit the K^＊（892） decay invariant mass spectra in different measurements and study the model dependence in the measurement of K^＊（892） parameters. We refit the K^＊（892）^0 mass spectra of the BELLE and FOCUS experiments with the formula used by BELLE in fitting K^＊（892）^- to get new mass and width. After refitting, the K^＊（892）^0 mass of the BELLE experiment becomes 1.4 MeV/c^2 larger than the initial value and that of the FOCUS experiment is 1 MeV/c^2 smaller than the initial value. We also fit the spectra of some other experiments to extract the K^＊ （892） parameters using the BELLE K^＊ （892）^- parametrization.