Abstract
The data on any aspect of public health, including that on infant mortality, has inbuilt hierarchical structure. Using traditional regression approach in data analysis, i.e., ignoring hierarchical structure, either at micro (individual) or at macro (community) level will be avoiding desired assumption related to independence of records. Accordingly, this may result into distortion in the results due to probable underestimation of standard error of the regression coefficients. To be more specific, an irrelevant covariate may emerge as an important covariate leading to inappropriate public health implications. To overcome this problem, the objective of the present work was to deal with multilevel analysis of the data on infant mortality available under second round of National family Health Survey and notify changes in results under traditional regression analysis that ignores hierarchical structure of data. This method provides more accurate results leading to meaningful public health implications. In addition, estimation of variability at different levels and their covariance are also obtained. The results indicate that the community (e.g., state) level characteristics still have major role regarding infant mortality in India. Further, if computational facilities are available, multilevel analysis may be preferred in dealing with data involving hierarchical structure leading to accurate results having meaningful public health implications.
Highlights
The public health research, the data structures are often hierarchical in nature especially those available at national/state/district/village/household/ individual level
In view of the above mentioned brief facts, if data is of hierarchical structure, there is need to deal with comparatively new approach in data analysis that: 1) takes hierarchical structure into account which makes it possible to incurporate variables from all levels and retain them at their own levels; 2) satisfies assumption of independence and considers total variation which obviously lead to correct analysis and proper interpretation of the data; 3) provides the relative importance of an individual’s characteristics and those of the community in which he/she lives; 4) helps in working out importance of variables across the levels through partition of variance; and 5) facilitates ranking of communities that obviously provides important clues strengthening ensuing public health programs
Under assessment of the community effects on infant mortality in India, the distribution of infant deaths in the relation to various socio-economic and demographic characteristics (Table 1) reveals that those children are more likely to die before celebration of their first birthday; whose mothers/fathers are having low level of education [2.22 (1.90 - 2.60), 1.70 (1.51 - 1.90)]; who come from rural area [1.51 (1.33 - 1.71)]; belong to SC/ST/
Summary
The public health research (e.g., outcomes like child survival; contraceptive adoption), the data structures are often hierarchical in nature especially those available at national/state/district/village/household/ individual level. Assumption of independence of observations cannot be ensured that is basic for the classical statistical techniques and can provide distorted results [1,2]; and 2) the second procedure is to aggregate the individual level variables to the higher level and do the analysis at higher level. Under this approach, all the within group information (variation) are thrown away which may be as much as 80% or 90% of the total variation before we start the analysis. In view of the above mentioned brief facts, if data is of hierarchical structure, there is need to deal with comparatively new approach in data analysis (i.e., hierarchical/multilevel analysis) that: 1) takes hierarchical structure into account which makes it possible to incurporate variables from all levels and retain them at their own levels; 2) satisfies assumption of independence and considers total variation which obviously lead to correct analysis and proper interpretation of the data; 3) provides the relative importance of an individual’s characteristics and those of the community in which he/she lives; 4) helps in working out importance of variables across the levels through partition of variance; and 5) facilitates ranking of communities that obviously provides important clues strengthening ensuing public health programs
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.