Exploring the Possibilities of Mail Survey as an Alternative to Land-Line Survey

Abstract

NHK’s national telephone public opinion surveys have been using “land-line and mobile RDD” method since December 2016. Meanwhile, regional surveys still have no other choice than depending only on land-line numbers (hereinafter referred to as “land-line RDD”) because mobile phone numbers do not identify the areas the users live in.However, it is revealed that the divergence between the sample composition ratios of recent “land-line RDD” surveys, and the Basic Resident Register’s composition ratio is widening, which suggests it is high time to review the effectiveness of this method.With this in mind, the NHK Broadcasting Culture Research Institute adopted the mail survey instead of “land-line RDD” to conduct a public opinion survey in Kobe City and surrounded areas in October 2018. In conducting the survey we attempted to shorten the survey period as much as possible to pursue the highest agility, which is a strong point of the telephone survey. As a result the following findings were obtained.　-Reminders were issued less frequently than usual and the survey period was shortened almost by half, but a comparatively high response rate was maintained.　-The first deadline was brought forward, which allowed speedy collection and contributed to building up a high response rate in the first half of the survey period.　-The sample composition ratio was less divergent from the Basic Resident Resister’s composition ratio, and the bias towards older generations was largely improved.The above findings imply that replacing “land-line RDD” surveys with mail surveys will increase the accuracy of regional surveys unless prompt publication of the results is required.