A Bi-level Clustering Analysis for Studying About the Sources of Vehicular Pollution in Chennai
The aim of this paper is to study about the awareness among the people in Chennai city, Tamil Nadu, about the causes of pollution. Initially, the k-means clustering method was applied to group variables rather than observations in the design of questionnaires. The first draft of a questionnaire contained more questions than is prudent to ensure a good response rate. When the draft questionnaire is tested on a smaller number of respondents (75 samples), it was observed that the responses to certain groups of questions are highly correlated. Hence, clustering analysis was applied to identify groups of questions that are most predominant in contributing to the reduction in air pollution in Chennai. Thus, the selected questions were used for survey purpose to study the acceptability among different sectors of people. Primary data were collected from 110 people belonging to different sectors of Chennai using questionnaire method. In the second level of cluster analysis, the cluster analysis was carried out to assign observations to groups. These results were further applied to identify the recommendation of suitable transport policies to mitigate vehicular pollution. This method of applying clustering techniques in two levels of the questionnaire analysis has been newly proposed in this paper.
KeywordsPollution Clustering Questionnaire survey Transport policies Data mining
- 2.V.M.H. Borden, Identifying and Analyzing Group Differences, in Intermediate/Advanced Statistics in Institutional Research, ed. by M.A. Coughlin (2005), pp. 132–168Google Scholar
- 4.P. Ewell, M. Boeke, Critical Connections: Linking States’ Unit Record Systems To Track Student Progress (Lumina Foundation for Education, Indianapolis, 2011)Google Scholar