Abstract
Since 2005, response rates of social surveys in Japan have declined sharply. For the prompt improvement of response rates, it is essential to grasp survey interviewers' actions. This paper proposes a use of visiting records because consequences of visiting patterns can be identified by recording when interviewers visited respondents' home and whom the interviewers met at their initial contact and the following contact(s). An analysis of visiting records of JGSS-2005 and JGSS-2006 clearly showed benefits to grasp interviewers' visiting patterns. Three major findings were 1) respondents' cooperative attitudes and interviewers' persistent visits lead to the improvement of response rates in JGSS-2006 ; 2) an increase on the number of women who live in complex housing resulted in a failure to contact them, which in return lead to low response rates among them ; 3) respondents' family members were continued to be a barrier to contact the respondents, while respondents cooperation rates improved. We argue that the accumulation of visiting record analysis is a key to raise response rates in large-scale surveys.
