2011 Volume 38 Issue 1 Pages 13-32
In Japan the “Basic Resident Register (BRR) ” is often used for statistical random sampling in public opinion surveys. The objective of the present paper is to compare two types of area sampling methods, without using the BRR. The first one (RRM) is a random route method, and the second one (DRM) is a method based on a database for a digital residential map. As a result, the percentage of data collected in the case of RRM was very low;and in the case of DRM, there was an uneven distribution in terms of respondents' age (a higher percentage of older respondents). However, when this bias was corrected by according weight to the population proportion, no strong significant difference concerning answer distributions was found. In each survey, the same substantial changes were detected statistically, compared with a past survey data based on the BRR. The findings suggest that these sampling methods are useful when accompanied with a careful analysis in spite of the resulting problems.