Improving the quality of disaggregated SDG indicators with cluster information for small area estimates

Abstract

The increasing needs for more disaggregated data motivates National Statistical Offices (NSOs) to develop efficient methods for producing official statistics without compromising on quality. In Indonesia, regional autonomy requires that Sustainable Development Goals (SDGs) indicators are available up to the district level. However, several surveys such as the Indonesian Demographic and Health Survey produce estimates up to the provincial level only. This generates gaps in support for district level policies. Small area estimation (SAE) techniques are often considered as alternatives for overcoming this issue. SAE enables more reliable estimation of the small areas by utilizing auxiliary information from other sources. However, the standard SAE approach has limitations in estimating non-sampled areas. This paper introduces an approach to estimating the non-sampled area random effect by utilizing cluster information. This model is demonstrated via the estimation of contraception prevalence rates at district levels in North Sumatera province. The results showed that small area estimates considering cluster information (SAE-cluster) produce more precise estimates than the direct method. The SAE-cluster approach revises the direct estimates upward or downward. This approach has important implications for improving the quality of disaggregated SDGs indicators without increasing cost. The paper was prepared under the kind mentorship of Professor James J. Cochran, Associate Dean for Research, Prof. of Statistics and Operations Research, University of Alabama.

A Primer on Clustering of Forest Management Units for Reliable Design-Based Direct Estimates and Model-Based Small Area Estimation

“…Cluster information can be incorporated into the SAE area-level models either as a categorical variable [66] or as an additional parameter to improve the estimates' efficiency [67,68]. In some studies, employing hierarchical clustering based on covariates has been shown to effectively yield more accurate predictions of small area means.…”

“…This is achieved by considering whether variance components are equal or unequal across different clusters [69]. Lastly, cluster information can be included as an average area random effects value in both unit-level [70] and area-level models [68,71,72] for predicting non-sampled areas.…”

The FAO Data Lab on statistical innovation and the use of big data for the production of international statistics

“…Topic 2 highlights the new skills in statistics needed to handle big data, which implies the emergence of the Data Science field and the importance of democratising the data understanding. In addition, the topic also highlights the Sustainable Development Goals (SDGs) in the context of the data revolution, and an example of this is "Improving the quality of disaggregated SDG indicators with cluster information for small area estimates" by Zulkarnain et al [8]. Topic 3 focuses on data management, such as data acquisition and integration.…”

“…The process for getting data from alternative online sources consists in "web scraping" 7 the information by following these steps: 1) searching for information on the internet, mainly from National Statistical Offices or Ministries websites; 8 2) extracting the data from its original source (e.g., PDF documents, Microsoft Excel files) and transforming it to a common format (a SQL table); 3) standardising the data, with respect to commodity names (converted to the Central Product Classification, CPC 9 ) and administrative area names (converted to GADM 10 ); 4) comparing to FAOSTAT data; 5) saving consistent data to the agricultural production database or analysing discrepancies. Though this process seems relatively straightforward, it is indeed plagued with several difficulties.…”

“…It means programmatically extracting data from websites (from web + scrape). A brief introduction to this topic is available at https://en.wikipedia.org/wiki/Web_scraping 8. This first step is mainly done when a new country in added to the list of countries for which data is required, though it is not always…”