The repeated measurement problems on Stated Preference (SP) data can be defined as the treatment of the combination of different variances (heteroscedasticity) and correlation of repeated observations from each individual. However, the repeated measurement problem has been approachedonly either as upward biased t-ratioor as correlation of disturbances according to repeated observations. Purposes of this paper are to consider the correlation in repeated observations from each individual and the heteroscedasticity of repeated observations from SP data and to suggest practical methods against the repeated measurement problem. This paper compares results of these approaches with those of the existing standard methods suggested. To identify the heteroscedasticity of the repeated observations, standard multiomial logit models with different number observations are estimated and then heteroscedasticity tests are accomplished. Correlation in repeated observations from each individual is considered by incorporating a stochastic individual heterogeneity variable. We specify this individual heterogeneity variable to vary following a normal distribution and the random variation effect is forced to be same within individual, but to be normally distributed across people. The SP data from a survey of drivers route choice in response to traffic information are used in this application. This research will contribute to identification of the errors in SP data and will high light the biased parameter estimates in absolute coefficients or coefficients ratios when the repeated SP observations are applied.