2011 Volume 67 Issue 5 Pages 67_I_911-67_I_918
In this paper, a methodology to set the aggregation factor for railway samples in the National Inter-city Travel Survey was developed. At first, the problem with the present methodology was showed ; because of sampling biases, OD traffic volume were not precisely calculated. A new methodology, which solved the problem by maximum likelihood estimation with additional information, was proposed. The new methodology was applied to the sample data taken in the 4th NITS in 2005. The results confirmed that the new methodology can set the aggregation factors which avoid the biases in travel purpose or traveler's attributes.