arXiv subjects

Sania Ashraf

Publications and source records attributed to Sania Ashraf.

4 recordsLinked to original sources

arXiv2023

Learning from diversity: jati fractionalization, social expectations and improved sanitation practices in India

Prevalence of open defecation is associated with adverse health effects, detrimental not only for the individual but also the community. Therefore, neighborhood characteristics can influence collective progressive behavior such as improved sanitation practices. This paper uses primary data collected from rural and urban areas of Bihar to study the relationship between jati (sub-castes) level fractionalization within the community and toilet ownership and its usage for defecation. The findings indicate a diversity dividend wherein jati fractionalization is found to improve toilet ownership and usage significantly. While exploring the channels, we find social expectations to play an important role, where individuals from diverse communities tend to believe that there is a higher prevalence of toilet usage within the community. To assess the reasons for the existence of these social expectations, we use data from an egocentric network survey on a sub-sample of the households. The findings reveal that in fractionalized communities, the neighbors with whom our respondents interacted are more likely to be from different jatis. They are also more likely to use toilets and approve of its usage due to health reasons. Discussions about toilets are more common among neighbors from fractionalized communities, which underscore the discernible role of social learning. The inferences drawn from the paper have significant implications for community level behavioral change interventions that aim at reducing open defecation.

econ.GN↗

arXiv2023

Shifting Beliefs Changing Behavior: Experimental Evidence on Sanitation from India

Open defecation, which is linked to poor health outcomes and lower cognitive ability has been widespread in India. Improved sanitation practice generates local health externalities, which implies that the returns to private toilet usage depend on community wide compliance. Therefore, beliefs about others toilet usage behavior may influence both sanitation choices and the perceived value of residing in areas with lower prevalence of open defecation. This paper evaluates a randomized, intervention in peri-urban Tamil Nadu designed to shift social expectations regarding sanitation practice. The intervention combines descriptive norm messaging, community engagement, and information provision to increase perceived prevalence and approval of toilet usage. The findings indicate a significant increase in toilet access in the intervention areas by 8 percentage points and consistent toilet usage by 6 to 7 percentage points relative to control areas. The effects are robust to potential bias due to disruption due to local socio-political protest and the COVID 19 pandemic during the study period. We also find significant increase in empirical expectations (beliefs about the prevalence of toilet usage) and normative expectations (beliefs about its approval). Mediation analysis indicates that the improvement in toilet access and usage is substantially driven by changes in these social expectations. We further observe significant increase in the willingness to pay among respondents for residing in open defecation free surroundings, suggestive of an increased perceived valuation of cleaner sanitation environments because of the intervention. The findings underscore the need for norm-centric interventions to propel change in beliefs and achieve long-term and sustainable sanitation behavior.

econ.GN↗

arXiv2020

Masked Linear Regression for Learning Local Receptive Fields for Facial Expression Synthesis

Compared to facial expression recognition, expression synthesis requires a very high-dimensional mapping. This problem exacerbates with increasing image sizes and limits existing expression synthesis approaches to relatively small images. We observe that facial expressions often constitute sparsely distributed and locally correlated changes from one expression to another. By exploiting this observation, the number of parameters in an expression synthesis model can be significantly reduced. Therefore, we propose a constrained version of ridge regression that exploits the local and sparse structure of facial expressions. We consider this model as masked regression for learning local receptive fields. In contrast to the existing approaches, our proposed model can be efficiently trained on larger image sizes. Experiments using three publicly available datasets demonstrate that our model is significantly better than $\ell_0, \ell_1$ and $\ell_2$-regression, SVD based approaches, and kernelized regression in terms of mean-squared-error, visual quality as well as computational and spatial complexities. The reduction in the number of parameters allows our method to generalize better even after training on smaller datasets. The proposed algorithm is also compared with state-of-the-art GANs including Pix2Pix, CycleGAN, StarGAN and GANimation. These GANs produce photo-realistic results as long as the testing and the training distributions are similar. In contrast, our results demonstrate significant generalization of the proposed algorithm over out-of-dataset human photographs, pencil sketches and even animal faces.

cs.CV↗

arXiv2020

Love Thy Neighbor? Perceived Community Abidance and Private Compliance to COVID-19 Norms in India

Compliance with measures like social distancing, hand-washing and wearing masks have emerged as the dominant strategy to combat health risk from the COVID-19 pandemic. These behaviors are often argued to be pro-social, where one must incur private cost to benefit or protect others. Using self-reported data across India (n=934) through online survey, we assess if changes in perceived community compliance can predict changes in individual compliance behavior, controlling for the potential confounders. We observe statistically significant and positive relationship between the two, even after accounting for omitted variable bias, plausibly allowing us to view the results from a plausible causal lens. Further, we find subsequent lockdowns such as the ones imposed in India, have a detrimental effect on individual compliance though the gains from higher perceived community compliance seems to offset this loss. We also find that sensitization through community can be particularly effective for people with pre-existing co-morbidities. Our findings underscore the need for multi-level behavioral interventions involving local actors and community institutions to sustain private compliance during the pandemic.

econ.GN↗