Adult community inhabitants, aged 25–50 years, were probabilistically selected from the residential registry in each of four municipalities (two in the Tokyo metropolitan area and two in neighboring prefectures). We intended to invite inner-sprawl urban and suburban regions, to account for variations in structural and social environments. For each municipality, 60 sample units were selected proportionally to the registered population, and systematic sampling12 was conducted for each unit, with oversampling of those aged in their 20s based on the expected lower response rate in this age stratum.
This sampling scheme, rather than a national representative random sampling, allowed us to collect data on individuals with various socioeconomic backgrounds under a homogeneous health, economic, and social policy environment, since each municipality forms a basic unit of local policy administration in Japan. With a sufficient number of samples, the scheme is expected to capture multilevel impact on health and related behaviors based on municipality characteristics and individual socioeconomic conditions. In addition, receiving endorsement from each municipality provides credibility which will help improve the response rate.
We contracted with independent survey agencies to conduct the surveys. Professional surveyors with more than three years of experience in conducting interview-based social surveys were recruited and underwent training sessions specifically to conduct the J-SHINE. The sessions lasted 6 hours for each wave, and included required lectures on the purpose of the survey, communication skills for home visiting and recruitment, contents of the questionnaire, operation of computer-based instruments and physiological measurement, and ethical consideration for confidentiality protection and safety, following training methods in previously established social surveys.13 We excluded poorly skilled surveyors during this training period. We further set up regular review sessions with surveyors during the survey wave in order to monitor their performance and quality of data collected, as well as to provide advice and consultation for troubleshooting.
The wave 1 survey was conducted between July 2010 and February 2011. The trained surveyors made at least five visits to reach the originally selected sample after sending an invitation letter. If they agreed to participate in the study, the participants were asked to provide written informed consent and then choose a convenient means for completing the survey questionnaire. Owing to the complex and contingent nature of the socioeconomic conditions among participants, we chose to use a computer-aided personal instrument (CAPI) to individually customize questionnaire items. The CAPI program was developed on an open-source platform and was accessible via the internet from the participant’s personal computer or on a left-behind laptop computer.
Surveyors provided participants with an ID and password for CAPI access and instructed participants on how to operate the CAPI session. The session was available around the clock and could freely be suspended and resumed at any point for participants’ convenience. Technical support was available by calling the support center. For those who were unfamiliar with computer use, a personal interview with the CAPI was provided. The collected data on left-behind computers were encrypted and sent via e-mail to our main server in the research laboratory. Each participant received a monetary incentive of ¥4000.
Among the participating households in the wave 1 survey, those with a spouse/partner of any age or child aged less than 18 years were invited to participate in supplemental surveys from August to December 2011. The spouse/partner survey asked the spouse/partner of the wave 1 participants to answer corresponding items to the wave 1 survey questionnaire to allow for pairwise comparisons. The child survey collected data on birth history and current conditions of child through the primary caregivers of the child, as well as through the child themselves if they were of school age. When the number of children was over three, the youngest three children were recruited for the survey.
The wave 2 survey was conducted between July and December 2012. The participant recruitment is summarized in Figures1 and 2 . All of the questionnaires and measurement in the wave 2 survey were responded by the wave 1 participants.
This sampling scheme, rather than a national representative random sampling, allowed us to collect data on individuals with various socioeconomic backgrounds under a homogeneous health, economic, and social policy environment, since each municipality forms a basic unit of local policy administration in Japan. With a sufficient number of samples, the scheme is expected to capture multilevel impact on health and related behaviors based on municipality characteristics and individual socioeconomic conditions. In addition, receiving endorsement from each municipality provides credibility which will help improve the response rate.
We contracted with independent survey agencies to conduct the surveys. Professional surveyors with more than three years of experience in conducting interview-based social surveys were recruited and underwent training sessions specifically to conduct the J-SHINE. The sessions lasted 6 hours for each wave, and included required lectures on the purpose of the survey, communication skills for home visiting and recruitment, contents of the questionnaire, operation of computer-based instruments and physiological measurement, and ethical consideration for confidentiality protection and safety, following training methods in previously established social surveys.13 We excluded poorly skilled surveyors during this training period. We further set up regular review sessions with surveyors during the survey wave in order to monitor their performance and quality of data collected, as well as to provide advice and consultation for troubleshooting.
The wave 1 survey was conducted between July 2010 and February 2011. The trained surveyors made at least five visits to reach the originally selected sample after sending an invitation letter. If they agreed to participate in the study, the participants were asked to provide written informed consent and then choose a convenient means for completing the survey questionnaire. Owing to the complex and contingent nature of the socioeconomic conditions among participants, we chose to use a computer-aided personal instrument (CAPI) to individually customize questionnaire items. The CAPI program was developed on an open-source platform and was accessible via the internet from the participant’s personal computer or on a left-behind laptop computer.
Surveyors provided participants with an ID and password for CAPI access and instructed participants on how to operate the CAPI session. The session was available around the clock and could freely be suspended and resumed at any point for participants’ convenience. Technical support was available by calling the support center. For those who were unfamiliar with computer use, a personal interview with the CAPI was provided. The collected data on left-behind computers were encrypted and sent via e-mail to our main server in the research laboratory. Each participant received a monetary incentive of ¥4000.
Among the participating households in the wave 1 survey, those with a spouse/partner of any age or child aged less than 18 years were invited to participate in supplemental surveys from August to December 2011. The spouse/partner survey asked the spouse/partner of the wave 1 participants to answer corresponding items to the wave 1 survey questionnaire to allow for pairwise comparisons. The child survey collected data on birth history and current conditions of child through the primary caregivers of the child, as well as through the child themselves if they were of school age. When the number of children was over three, the youngest three children were recruited for the survey.
The wave 2 survey was conducted between July and December 2012. The participant recruitment is summarized in Figures
Free full text: Click here
