Global Issues and Policy Solutions
Chapter 6: New Results for South and Southeast Asia
Having created the Asian samples of children under the age of five matched with their mothers and constructed the dependent and independent variables, the next step of the analysis entailed cross-tabulating each of the nutritional status measures and the independent variables by maternal employment. The objective of the bivariate tabulations was to test whether those variables have a statistically significant bivariate association prior to entering them into the multivariate specifications. These tests provide insight as to patterns across countries in how the prevalence of poor nutritional outcomes may differ for children of non-employed and employed women, and how socio-economic status and household characteristics may differ between these groups. BIVARIATE RESULTS Figure 6.1 reports the bivariate statistics on the prevalence of small birth size, stunting, and wasting among mothers with varying degrees of attachment to the labor force. Given the structure of the survey question about the mother’s employment status, the figure categorizes maternal employment as either (1) not employed at all in the past year, (2) employed in the past year but not now, and (3) currently employed. This three-way categorization provides more specific information on women’s employment activity compared to a binary variable measuring currently employed versus not currently employed or an alternative binary variable measuring employed currently or in the past year versus not employed at all in the past year. As the test of statistical significance for all of the bivariate associations, the Pearson chi-squared test statistic is corrected for clustering at the household level and is...
You are not authenticated to view the full text of this chapter or article.