Constructing Multilevel Geographic Data Using an Online Survey and Systematic Social Observation

DOI Web Site Web Site 35 References

Bibliographic Information

Other Title
  • インターネット調査と系統的社会観察による地理的マルチレベルデータの構築
  • インターネット チョウサ ト ケイトウテキ シャカイ カンサツ ニ ヨル チリテキ マルチレベルデータ ノ コウチク

Search this article

Description

<p>In the past few decades, multilevel studies, which aim to explain individual outcomes in terms of both individual attributes and contextual exposures, have been rapidly increasing worldwide. For example, health geographers have paid more attention to how a neighborhood environment affects the health of residents. However, such multilevel studies are still relatively limited in Japan. One of the reasons for this is the difficulty in collecting data at the individual and contextual level, both of which are necessary to construct multilevel geographic data. Since 2000, and especially after the enforcement of the Act on the Protection of Personal Information in 2005, traditional surveys for collecting individual data (e.g., door-to-door surveys) and contextual/aggregated data (e.g., the population census) have shown significantly declining response rates. To compensate for the lack of multilevel geographic data, we propose a new method that will combine online surveys to collect individual responses and systematic social observations to measure neighborhood-level characteristics.</p><p>First, we conducted an online survey with the registered members of a survey company, who were aged 20 years or older and living in Bunkyo ward, Tokyo, for the purpose of maximizing the number of respondents. In total, 989 responses were collected from all postal areas (n=20) in the study region, which was considered sufficient for a small-area multilevel analysis. Next, we assessed the geographic patterns of neighborhood characteristics by conducting a systematic social observation of the streetscape (2,718 street segments in Bunkyo ward) using Google Street View to improve efficiency. Finally, we examined the association between habitual walking for leisure among respondents and micro-scale walkability in the neighborhoods by analyzing the multilevel geographic data constructed through the online survey and neighborhood observation. The results of the multilevel logistic regression analysis confirmed a significant positive association between them (odds ratio=1.54, 95% confidence interval: 1.11–2.14), after adjusting for demographic and socioeconomic characteristics.</p><p>These results indicate that constructing multilevel geographic data using online surveys and systematic social observations is useful for advancing multilevel studies as a complement to traditional data collection methods. Given that the traditional methods, including door-to-door surveys and the population census, have become increasingly difficult, the method proposed in this study may contribute to collecting original data and conducting multilevel geographic analysis with neighborhood-level resolution for broader geographic areas.</p>

Journal

References(35)*help

See more

Related Projects

See more

Keywords

Details 詳細情報について

Report a problem

Back to top